Sie sind auf Seite 1von 391

Economic Analysis of Law in European Legal Scholarship 3

Klaus Mathis
Avishalom Tor Editors

Nudging - Possibilities,
Limitations and
Applications in
European Law and
Economic Analysis of Law in European Legal

Volume 3

Series Editor
Klaus Mathis, University of Lucerne, Switzerland

Editorial Board
Pierluigi Chiassoni, University of Genoa, Italy
Péter Cserne, University of Hull, United Kingdom
Bruno Deffains, University of Paris II - Sorbonne Universities, France
Thomas Eger, University of Hamburg, Germany
Mariusz J. Golecki, University of Łódž, Poland
Andreas Heinemann, University of Zurich, Switzerland
Régis Lanneau, University of Paris West Nanterre La Défense, France
Aurélien Portuese, University of Westminster, United Kingdom
Kai Purnhagen, University of Wageningen and Erasmus University Rotterdam,
The Netherlands
Lucia A. Reisch, Copenhagen Business School, Denmark
Anne-Lise Sibony, University of Liège, Belgium
Endre Stavang, University of Oslo, Norway
The purpose of this book series is to publish high quality volumes in the growing
field of law and economics research in Europe, from a comprehensive theoretical
and practical vantage point. In particular, the series will place great emphasis on
foundational and theoretical aspects of economic analysis of law and on
interdisciplinary approaches in European Legal Scholarship. Following Nobel
laureate Ronald Coase’s famous essay “The Problem of Social Cost” (1960) fifty
years ago law and economics has become the lingua franca of American
jurisprudence. In recent decades, law and economics has also gained widespread
popularity in Europe and its influence on Legal Scholarship is growing significantly.
Therefore, the economic analysis of law in European Legal Scholarship academic
book series illustrates how law and economics is developing in Europe and what
opportunities and problems – both in general and in specific legal fields – are
associated with this approach within the legal traditions of European countries.
Rather than further exploring economic analysis as such, the main focus of this
series lies on the implementation of economic methods in legislation and legal
adjudication from a European perspective. It takes into account the particular
challenges the European legal systems face. Volumes will address law and economics
research in Europe from a critical and comparative viewpoint. The studies in this
series are strong and bold narratives of the development of economic analysis of law
in European Legal Scholarship. Some are suitable for a very broad readership.
Contributions in this series primarily come from scholars in Europe. The purpose is
to provide the next generation of European lawyers with the models and skills
needed to understand and improve the economic analysis of law in their own legal
field. The series includes monographs focusing on specific topics as well as
collections of essays covering specific themes.

More information about this series at

Klaus Mathis • Avishalom Tor

Nudging - Possibilities,
Limitations and Applications
in European Law
and Economics
Klaus Mathis Avishalom Tor
Faculty of Law Notre Dame Law School
University of Lucerne University of Notre Dame
Lucerne, Switzerland Notre Dame. IN, USA
Faculty of Law
University of Haifa
Haifa, Israel

Economic Analysis of Law in European Legal Scholarship

ISBN 978-3-319-29560-2 ISBN 978-3-319-29562-6 (eBook)
DOI 10.1007/978-3-319-29562-6

Library of Congress Control Number: 2016936892

© Springer International Publishing Switzerland 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland

Once again, it was a pleasure to welcome as the Vice-Dean of the Faculty of Law at
the University of Lucerne numerous academics and practitioners from around the
world to the 4th Law and Economics Conference here in Lucerne. The issue of
“Nudging – Theory and Applications” seemed to be the rational consequence of last
year’s issue “Behavioural Law and Economics”, since nudging is about influencing
behaviour. Although nudging has already been part of marketing strategies of pri-
vate enterprises for decades, it is only recently that this instrument has found its way
into the discourse on intelligent regulation. In that context, nudging is about influ-
encing decisions of persons for the sake of objectives of the public welfare without
command or prohibitions. A basic example given by Thaler and Sunstein is – for-
give me – a urinal with the picture of a fly right in its centre, nudging a male person
to try to hit it with the liquid jet and thereby avoiding any unwanted loss of this
liquid material by distraction. Obviously, this serves the common welfare of provid-
ing for a clean environment in the toilet. Furthermore, the instrument works just by
focusing the individual on its own action. It is debatable whether the nudge in this
example refers to the hunting instinct or rather to the play instinct. Surely, the play
instinct is concerned, when some years ago in Germany, I had the chance to try a
nudge replacing in our example the picture of the fly by a little soccer goal with a
soccer ball attached to the crossbar by a little string. Thaler and Sunstein state that
due to the nudge of the fly toilets have become cleaner by 80 %. In Germany, for the
nudge of the soccer goal, I lack any figures. However, we all know, that in 2014
Germany won the world championship in soccer.
To be serious again, both, in the USA and in the UK, the governments have
established administrative ‘Teams on Behavioural Sciences or Behavioural
Insights’. With due delay, in Germany, the Federal Government tries to establish a
similar team. However, in Germany the concept faces strong criticism – citation: ‘is
the gawk the new concept of the civilian in regulation?’ – and likewise in
Switzerland, the well-known NZZ asked, whether we all should now become petty
bourgeois (‘kleine Spiesser’).
For governments the concept of nudging has some obvious advantages. People
are led to make ‘better’ decisions and are pushed towards contributing to the public

vi Foreword

welfare without feeling suppressed that is without blaming the government for
interfering with their rights. When relying on advantages for the individual, regula-
tion might overcome hurdles seen by systems theory for steering different systems,
like economic behaviour being influenced by politics and law. The thought of nudg-
ing is especially attractive where major changes have to be put into effect, like the
energy strategy in Switzerland calling for decarbonization and an exit from nuclear
power in its present shape.
However, a myriad of questions remain:
• What is the underlying picture of the individual under that approach, an enlight-
ened person or a lazy ‘gawk’ looking for convenience? And does a regulatory
approach form persons according to its underlying picture of them?
• Where are the boundaries of nudging when the precautionary principle calls for
immediate action?
• Is nudging acceptable, even if it makes the state use non-transparent strategies of
Furthermore, in literature it has been doubted that nudging is really libertarian:
There is an interesting example from the community of Zurich asking people to
make a choice between energy supply from hydropower or a mix of energy
resources, nudged by the use of colours. A warning red and a calming blue colour
were chosen to nudge people towards the hydropower. The next year the ‘general
mix’ was replaced by ‘nuclear energy mix’ and a year later, the option for choice
has vanished: sorry, there is only blue energy available in Zurich left. This has been
described as ‘nudging on a slippery slope’.
On the contrary, some might regard nudging simply as deregulation in disguise.
I would like to add a question concerning the underlying rational of nudging. Is
it really about paternalism, acting in conformity with the objectives of the individ-
ual in the long run? Some authors differentiate between the objectives of an indi-
vidual in the short run (enjoying sweets or smokes) and in the long run (suffering
from health issues). Thaler and Sunstein state that nudging aims at the common
welfare. In my analysis, intelligent regulation is only interested in nudging for the
sake of common welfare. The individual behaviour only raises the interest of the
regulator, if it might cause more costs to society, like costs for the public health
system or costs for the social welfare system. Interestingly, if that is true, then nudg-
ing shows that ‘better’ behaviour is something outside the thoughts of the so-called
rational homo economicus limited to egoistic-economical views in traditional eco-
nomic science.
Further, the question may be raised how successful nudging really is. As already
mentioned, in their example, Thaler and Sunstein state that after the use of the pic-
ture of the fly, toilets have become cleaner by 80 %. In German public toilets the fly
has vanished, but a pay and voucher system has taken over. So, under the approach
for intelligent regulation, nudging not only has to qualify by results, but by better
results than gained by concurring instruments. This is the threshold by the test of
Foreword vii

As the managing director of the Center for Law and Sustainability at the
University of Lucerne, I am very proud that my distinguished colleague, Klaus
Mathis, has succeeded in bringing together so many big academic elephants – if you
kindly let me exploit the logo of this conference with the big and the small ele-
phants – together in this room at the small University of Lucerne to discuss some of
the many questions raised by the concept of nudging. Obviously, the logo works in
the other direction as well: even a small elephant can nudge the big ones.

Faculty of Law Sebastian Heselhaus

University of Lucerne
Lucerne, Switzerland
December 2015

This anthology, ‘Nudging – Possibilities, Limitations and Applications in European

Law and Economics’, is the result of the 4th Law and Economics Conference held
at the University of Lucerne on 17th to 18th of April 2015. The conference was
organized in partnership with Notre Dame Law School and the Notre Dame Program
on Law and Market Behavior (ND LAMB). The main focus of the conference lay
on European legal questions as presented by European legal scholars. They were
complimented by insights from distinguished scholars from the USA in order to
foster the dialogue between the two different legal cultures. The thematic scope of
this volume spans both the theoretical and practical developments of nudging.
We take this opportunity to thank all those who have contributed to the organiza-
tion of the conference and to the successful completion of this volume. First of all,
we would like to thank Uta Dietrich and Julia Wetzel, MLaw, for their flawless
coordination and organization of the conference. Furthermore, we wish to thank
Gareth Hunt, BSc, and Lynn Gummow, MLaw, for their diligent proofreading. A
special thank goes to the Swiss National Science Foundation (SNSF), the Research
Commission (FoKo) of the University of Lucerne, the Institute lucernaiuris, as well
as the Rectorate of the University of Lucerne for supporting the conference. Finally,
we are grateful to Neil Olivier and Diana Nijenhuijzen at Springer Publishers for
overseeing the publishing process.

Lucerne, Switzerland Klaus Mathis

Notre Dame, IN, USA Avishalom Tor
December 2015


Part I Foundations
1 The Critical and Problematic Role of Bounded Rationality
in Nudging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Avishalom Tor
2 Beneficial and Exploitative Nudges . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Bruno S. Frey and Jana Gallus
3 The Crucial Importance of Interests
in Libertarian Paternalism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Mark D. White
4 Condorcet’s Jury Theorem as a Rational Justification
of Soft Paternalistic Consumer Policies . . . . . . . . . . . . . . . . . . . . . . . . 39
Malte Frederic Dold
5 To What Extent Should the State Protect Human Beings
from Themselves? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Peter G. Kirchschlaeger
6 Nudging Is Judging: The Inevitability of Value Judgments . . . . . . . . 69
Ariel David Steffen

Part II Applications
7 Nudging and the Principle of Proportionality . . . . . . . . . . . . . . . . . . . 93
Mark Schweizer
8 Nudging in Swiss Contract Law? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Klaus Mathis and Philipp Anton Burri
9 Designing Disclosures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Geneviève Helleringer

xii Contents

10 Neutral Third-Party Counselling as Nudge Toward

Safer Financial Products? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Piotr Tereszkiewicz
11 The Potential Use of Visual Packaging Elements as Nudges . . . . . . . . 197
Kai Purnhagen, Erica van Herpen, and Ellen van Kleef
12 Nudging: Information, Choice Architecture and Beyond . . . . . . . . . . 217
Rainer Baisch
13 Nudging and Environmental Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Felix Ekardt and Jutta Wieding
14 Behavioural Regulation in International Trade . . . . . . . . . . . . . . . . . . 263
Georgios Dimitropoulos
15 Nudging in Tax Law? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Mariusz J. Golecki, Marcin Romanowicz,
and Jerzy W. Wojciechowski
16 Nudging as a Tool of Media Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Mira Burri
17 Academic Fraud at Hogwarts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
Rute Saraiva

About the Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379

Rainer Baisch Rechtswissenschaftliche Fakultät, University of Zurich, Zurich,

Mira Burri World Trade Institute, University of Bern, Bern, Switzerland
Philipp Anton Burri Faculty of Law, University of Lucerne, Lucerne, Switzerland
Georgios Dimitropoulos Max Planck Institute Luxembourg for Procedural Law,
Luxembourg, Luxembourg
Malte Frederic Dold Albert-Ludwig University Freiburg, Freiburg i. Br., Germany
Felix Ekardt Forschungsstelle Nachhaltigkeit und Klimapolitik, Leipzig, Germany
Bruno S. Frey Center for Research in Economics, Management and the Arts
(CREMA), University of Zurich, Zurich, Switzerland
Jana Gallus Center for Research in Economics, Management and the Arts
(CREMA), Univeristy of Zurich, Zurich, Switzerland
Mariusz J. Golecki Department of Legal Theory and Philosophy of Law, Faculty
of Law and Administration, University of Łódź, Łódź, Poland
Geneviève Helleringer Essec Business School, Cergy-Pontoise Cedex, France
Institute of European and Comparative Law, Oxford University, Oxford, UK
Peter G. Kirchschlaeger Faculty of Theology, University of Lucerne / Yale
University, Lucerne, Switzerland
Klaus Mathis Faculty of Law, University of Lucerne, Lucerne, Switzerland
Kai Purnhagen Law and Governance Group, Wageningen University, Wageningen,
The Netherlands
Marcin Romanowicz Faculty of Law and Administraion, University of Warsaw,
Warsaw, Poland

xiv Contributors

Rute Saraiva Alamedad de Universidade, University Lisbon Law School, Cidade

Universitária, Lisboa, Portugal
Mark Schweizer Law School, University of St. Gallen, St. Gallen, Switzerland
Ariel David Steffen Faculty of Law, University of Lucerne, Lucerne, Switzerland
Piotr Tereszkiewicz Faculty of Law and Administration, Jagiellonian University
of Cracow, Cracow, Poland
Avishalom Tor Notre Dame Law School, University of Notre Dame, Notre Dame,
Faculty of Law, University of Haifa, Haifa, Israel
Erica van Herpen Marketing and Consumer Behavior Group, Wageningen
University, Wageningen, The Netherlands
Ellen van Kleef Marketing and Consumer Behavior Group, Wageningen
University, Wageningen, The Netherlands
Mark D. White Department of Philosophy, College of Staten Island, Staaten
Island, NY, USA
Jutta Wieding Forschungsstelle Nachhaltigkeit und Klimapolitik, Leipzig,
Jerzy W. Wojciechowski Faculty of Psychology, University of Warsaw, Warsaw,

Cass R. Sunstein and Richard Thaler proposed in their book ‘Nudging – Improving
Decisions About Health, Wealth and Happiness’, that altering people’s behaviour in
a predictable way without forbidding any option or significantly changing economic
incentives can and should be used to help people make better decisions. This has
sparked interdisciplinary debates in academia as well as in the public sphere.
People can be ‘nudged’ in the desired direction, for instance, by means of default
rules. One example for a default rule is the objection clause in the case of organ
donations, according to which every human being is considered to be a donor after
death unless they explicitly opt out. The use of nudges, however, raises a number of
questions: who decides what behaviour is desired? Do nudging and the so-called
libertarian paternalism threaten the autonomy of people? Does the state not pretend
to have knowledge which it does not have, as Friedrich August von Hayek would
argue? Should we not better trust the spontaneous order, even if it will not always
be perfect? At the extreme, might nudging not pave the way to serfdom, to allude to
Hayek again? The present anthology analyses and discusses the issues surrounding
nudging and its use in intelligent legislation, regulation, and policy making more
In Part I the authors discuss the foundations of nudging theory from different
First, Avishalom Tor discusses the Nudge policy framework and draws attention
to the fact that many other policy instruments are equally suited to promoting the
goal encouraging boundedly rational people to better achieve their own ends.
Furthermore, he highlights that the tools of nudging are equally suited to promoting
goals that are excluded by Nudge’s own framework. This problem of fit between
Nudge’s goals and tools causes some of Thaler and Sunstein’s own applications to
breach the boundaries of what they define as legitimate nudges and, more signifi-
cantly, obscures the broader potential of behaviourally informed policies and the
substantial trade-offs involved in their implementation.
Bruno S. Frey and Jana Gallus state that the effectiveness of nudges in raising the
welfare of the population hinges on the policymakers employing them. They argue
that rather than being concerned about policymakers’ incapacity to raise the

xvi Introduction

population’s welfare we should be concerned about their unwillingness to do so. As

a possible solution, they argue that a legal framework for the use of nudges should
be implemented on a constitutional level. Nudging should not be considered as an
innocuous exception to constitutionally based decision-making even if it appears as
beneficial at first sight. According to them, even ‘Liberal Paternalism’ may not be
imposed on the population of a democratic society without its consent.
One of the main criticisms aimed at liberal paternalism is the question of an
individual’s self-interest. In his contribution, Mark D. White argues that self-interest
has been neglected in recent scholarship aimed at defending paternalism by means
of a behavioural approach. In particular, he disputes claims that paternalism of
means can be meaningfully separated from paternalism of ends, and argues that
modern paternalism does not respect people’s true interests but instead adopts a
perfectionist or objective conception of well-being.
In his contribution Malte Frederic Dold revisits the meaningfulness of the
Condorcet Jury Theorem (CJT) and applies it to the recent debate on liberal pater-
nalism and consumer protection. In his paper, he outlines a case for the application
of a Condorcet jury voting procedure in consumer law in order to nudge rational
From a human rights perspective, Peter Kirchschläger argues that nudging could
seem to be attractive if it would lead to a better realization of human rights. However,
he cautions that it is at the heart of human rights to protect the individual from the
collective, respectively from abuse of power by the collective. This therefore raises
the question if the argument that the ends justify the means should apply in order to
justify the use of nudging.
Ariel Steffen argues that nudges always entail ex-ante value judgements. He sees
this as troublesome when viewed from the perspective of a collapsed fact/value
dichotomy as this means that facts and values are always inextricably entangled. As
a result, the normative concept ‘rationality’ is maximized under the guise of it being
a positive concept while ‘autonomy’ is rejected as a legitimate concept for maximi-
zation on grounds of it being normatively laden. He argues for the use of pragmatic
ethics as a basis to make the normative transparent while at the same time not
exclude it from scientific discourse.
In Part II various applications of nudges are illuminated:
From the perspective of the governmental ‘Choice Architects’, Mark Schweizer
examines the potential limits the German Federal Constitutional Court imposes
through their principle of proportionality. While nudges as such generally do not
interfere with fundamental rights, their implementation will often interfere with the
fundamental rights of citizens other than the decision makers and therefore trigger
constitutional scrutiny. He concludes that the proportionality principle does not
compel the use of nudges.
Klaus Mathis and Philipp Anton Burri discuss various interpretations of the func-
tion of non-mandatory contract law along with the requirements with regard to
adequate formulation of a contract. From this basis they then proceed to discuss the
behavioural economic analysis of non-mandatory contract law, especially in view of
using it as a nudging instrument. Non-mandatory law, they argue, could be used to
Introduction xvii

steer behaviour of individuals. They support their thesis by discussing labour law
and surety contract law as two examples.
Moving from contracts to the financial sector, Geneviève Helleringer explores
the psychology of disclosure in the context of advised financial investment deci-
sions. She investigates how information about the financial advisor’s potential con-
flict of interest impacts on the participants’ judgement. The implications of her
findings for public policy are that, contrary to what is commonly suggested, prob-
lems of conflict of interest in the financial investment context may efficiently be
treated by means of disclosures, provided that such disclosures explicit the conse-
quences of the conflict for the decision maker.
Staying in the financial sector, Piotr Tereskiewicz discusses the idea of neutral
counselling as a measure of improving decisions of borrowers when choosing a
mortgage loan. The developments regarding foreign-currency mortgage loans, pop-
ular in Central and Eastern Europe in recent years, illustrate how incorrect mortgage
choices may have detrimental effects on borrowers and their families. Taking a
sceptical approach towards mandated pre-contractual disclosure as a protection
measure for borrowers, the paper advocates neutral counselling of mortgage appli-
cants in case they are offered risky types of mortgages. Drawing on the experience
of an American mortgage counselling programme, a set of recommendations is
made regarding the design of a possible neutral mortgage counselling scheme.
While the counselling solution has imminent limitations, it displays clear potential
towards improving the process of financial decision-making.
Nudges also play an important role in health and nutrition. In their contribution
Kai Purnhagen, Erica van Herpen and Ellen van Kleef argue that in order to protect
consumers in the context of making healthy food choices, internal market regulation
should focus more on pictorial health and nutrition claims and representations. They
back up their claim by research from behavioural sciences. The main focus of atten-
tion of the misleading potential of health and nutrition claims in the EU is on textual
In his contribution, Rainer Baisch discusses the traditional disclosure paradigm
based on the assumption that transparent and effectively processed information will
enable investors to make well-founded investment decisions. In particular, he analy-
ses human flaws with regard to how information is processed and how the provision
of data for consumers of financial products should be further optimized.
Nudges also play a role in both constitutional and environmental law. Felix
Ekardt’s and Jutta Wieding’s contribution addresses nudges in both areas. They dis-
cuss the criticism that nudges are too paternalistic and argue that many forms of
nudging are not paternalistic, because they do not protect an individual from them-
selves, but rather protect others from negative effects of the actions of an individual.
Especially in environmental policies, nudging therefore proves to be legitimate
approach. Unfortunately, they claim that nudging might not be as effective as policy
instrument for environmental issues, as many hope.
Georgios Dimitropoulos’ contribution proposes to recast a big part of interna-
tional trade law as behavioural regulation. The Sanitary and Phytosanitary and
Technical Barriers to Trade Agreements of the WTO include provisions on disclosure
xviii Introduction

of better information, on legal default rules and on debiasing through law that are
similar to the ones proposed in Behavioural Law and Economics literature to nudge
governments towards specific directions. Therefore, he argues that the current inter-
national trade law is already based on using nudges.
To examine nudges from an empirical point of view, Mariusz J. Golecki, Marcin
Romanowicz and Jerzy W. Wojciechowski conduct an experiment on the verifica-
tion of the hybrid based categorization hypothesis in context of the tax law regula-
tion. Their paper concentrates on the possibility and cognitive effectiveness of
nudging in a very narrow field of legislative design rather than on the debate on the
normative implications of this strategy. However, their approach to illuminating the
descriptive aspects of cognitive processes presupposes the debate on the legitimacy
of nudging from the wider normative, legal and moral perspectives. Nevertheless,
this paper provides some very useful insights on cognitive behaviour.
In her contribution, Mira Burri explores the use of nudges in Media Policy. Using
the Internet as an example, she outlines how the use of nudges as a subtle form of
state intervention is effective intervention aimed at maintaining diversity. She argues
that some form of governance is essential if the individual’s rights are to be pro-
tected and views nudges as an effective tool in the move to develop new Media
Finally, the use of nudges in academic fraud is discussed by Rute Saraiva using
the situations or criteria presented by Harry Potter novels. By implying the identifi-
cation of the costs and benefits behind the motivation of cheating and of the risk
taken both in an individual and collective perspective, the decision to cheat can be
nudged in one direction or another. Furthermore, it is necessary to recognize the
biases behind the decision of cheating, including an eventual distortive institutional
framework. Using Hogwarts School of Witchcraft as an example, she proposes a
correct and preventive architecture of choice and nudging behaviours to decrease
academic fraud.
As this anthology highlights, nudges have a wide scope of application. While
many argue that nudges do indeed steer individuals to greater social or personal
welfare, it is also clear that there are many pitfalls and risks as discussed by the
contributors to this anthology. The 17 essays provide startling insights into the mul-
tifaceted debate surrounding the use of nudges in European Law and Economics.

Faculty of Law Klaus Mathis

University of Lucerne
Lucerne, Switzerland
Notre Dame Law School Avishalom Tor
University of Notre Dame
Notre Dame, IN, USA
Faculty of Law
University of Haifa
Haifa, Israel
December 2015
Part I
Chapter 1
The Critical and Problematic Role of Bounded
Rationality in Nudging

Avishalom Tor

Abstract Nudging has become an increasingly popular policy tool on both sides of
the Atlantic, even while scholars and commentators continue to debate its appropri-
ate boundaries, efficacy, and legitimacy. The present chapter outlines a sympathetic,
‘internal’ critique of Nudge’s policy framework (Thaler and Sunstein 2008. Nudge:
Improving decisions about health, wealth, and happiness. New Haven: Yale
University Press) identifying a fundamental problem that has received little atten-
tion to date. Thaler and Sunstein define their key concept as the use of a particular
toolkit (rationally-neutral behavioural interventions) to promote a specific policy
goal (individuals’ self-judged well-being). To remain within the boundaries of these
defining parameters, the chosen toolkit must nudge only by helping boundedly
rational people better achieve their own ends. However, a closer look at this intui-
tively appealing approach reveals a fundamental problem of fit that should trouble
even those who support Nudge’s mission. Namely, while the tools of nudging can
sometimes promote its declared goals, other tools can often do so equally well or
even more effectively. And the tools of nudging are equally suited to promoting
goals that are excluded by Nudge’s own framework. This problem of fit causes some
of Thaler and Sunstein’s own applications to breach the boundaries of what they
define as legitimate nudges and, more significantly, obscures the broader potential
of behaviourally-informed policies and the substantial trade-offs involved in their

1.1 Introduction

The publication of Thaler and Sunstein’s “Nudge: Improving Decisions About

Health, Wealth, and Happiness” [‘Nudge’] in 2008 brought to international, popular
attention the developing academic discourse about behaviourally-informed legal

A. Tor (*)
Notre Dame Law School, University of Notre Dame,
3163 Eck Hall of Law, Notre Dame, IN 46556, USA
Faculty of Law, University of Haifa, Haifa, Israel

© Springer International Publishing Switzerland 2016 3

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_1
4 A. Tor

interventions, which until then took place primarily within the pages of U.S. law
reviews.1 In the years that followed, scholars from law,2 economics,3 political
science,4 philosophy,5 psychology,6 and more have been joined by numerous com-
mentators in discussing the merits and demerits of nudging and related approaches.
Despite the explosion of interest in Nudge, many questions remain about the appro-
priate contours and boundaries of such interventions and their normative justifica-
tion. Scholars have challenged the nudge enterprise on a variety of grounds, from
public choice,7 to welfare,8 autonomy,9 and beyond.10
The present remarks contribute to this conversation by considering an important
set of foundational problems with the policy framework offered by Nudge that have
not received much attention to date. Specifically, to qualify as a nudge according to
Thaler and Sunstein, a policy intervention must use only economically-neutral
behavioural interventions. Moreover, not all economically-neutral behavioural
interventions qualify as nudges, only those employed to improve individuals’ self-
judged well-being. Yet Individuals require such nudging only insofar as they fail to
use the best means to achieve their own ends—that is, only when they fail to act
rationally. Hence, the need for nudging as defined by Thaler and Sunstein arises
only through deviations from rational action and the goal they assign to nudging can
be accomplished only by making people act more rationally (or as if they were more
Upon closer inspection, both the logic and appeal of Nudge thus ultimately rest
on a critical link among its behavioural toolkit, its well-being goal, and rationality.
Once revealed, however, the role of rationality in nudging exposes a fundamental
weakness of its framework that this chapter outlines and begins to explore.

1.2 The Tools and Goals That Define Nudging

The universe of policy interventions that qualify as nudges can be described along
two dimensions: the policy tools that aspiring nudgers legitimately may employ and
the policy goals towards which they legitimately may aim.

E.g., Camerer et al. 2003; Glaeser 2006; Klick and Mitchell 2006; Rachlinski 2006; Sunstein and
Thaler 2003; Zamir 1998.
E.g., Bubb and Pildes 2015; Sibony and Alemanno 2015; Willis 2013.
Bernheim 2009; Spiegler 2015.
Hausman and Welch 2010; Maloberti 2012.
Bovens 2010; Conly 2012; Hausman and Welch 2010; White 2013.
Glaeser 2006; Johnson et al. 2012.
Mitchell 2005; Mongin and Cozic 2014; Rebonato 2014.
Bernheim 2009, Wright and Ginsberg 2012.
Mitchell 2005; Wright and Ginsberg 2012; Rebonato 2014; Baldwin 2015.
Mitchell 2005; Menard 2010.
1 The Critical and Problematic Role of Bounded Rationality in Nudging 5

Thaler and Sunstein define nudges as “any aspect of the choice architecture that
alters people’s behaviour in a predictable way without forbidding any option or
significantly changing their economic incentives.”11 And choice architecture in the
authors’ terminology is synonymous with “the context in which people make
decisions.”12 In other words, any aspect of the context of choice can qualify as a
nudge, so long as it fulfils two additional, cumulative conditions—namely, the pre-
dictable behavioural effect condition and the rational neutrality condition.
The first condition requires nudges to produce a predictable behavioural effect.
Yet to count as nudges, behavioural decision phenomena also must fulfil the second
condition of rational neutrality. The effects they exert on behaviour must not be
achieved by substantially constraining people’s choices or changing their economic
incentives—the only types of intervention that are normatively relevant for a ratio-
nal economic actor.13
Behavioural research has shown how actual human behaviour is shaped by the
limited cognitive resources that individuals have at their disposal and subject to the
effects of motivation and affect. These and similar factors that are irrelevant to the
hypothetical rational actor describe individuals’ ‘bounded rationality’.14 Notably,
the same processes of bounded rationality that lead to deviations from rational
action also offer new tools with which behaviourally-informed approaches to policy
making, such as Nudge, can try to shape the behaviour of boundedly rational
Thaler and Sunstein seek to constrain the tools and goals of nudging further to
allow only ‘libertarian paternalistic’ behavioural interventions. Nudge explains
that: “libertarian . . . simply mean[s] liberty-preserving. . . .”,15 so it is largely cov-
ered already by the definition of nudging as using only rationally-neutral behav-
ioural tools. The second, ‘paternalism’ component, on the other hand, addresses the
goals of nudging. The book understands “a policy [to be] ‘paternalistic’ if it tries to
influence choices in a way that makes choosers better off, as judged by themselves.”16
This definition substantially limits the universe of legitimate nudges, excluding a
great many behavioural interventions that aim to promote social welfare without
regard to the well-being of the nudged individuals or even those traditionally pater-
nalistic policies that seek to make people better off according to policy makers’
Nudge aims to make people better off as they themselves judge the matter. As
long as people choose the best available means to further their ends, there is no
room for nudging. This observation is important because suiting means to ends is a
basic characteristic of rational action.17 Hence, when boundedly rational individuals

Thaler and Sunstein 2008, p. 6.
Thaler and Sunstein 2008, p. 3.
Hausman 1992.
Tor 2002, 2008.
Thaler and Sunstein 2008, p. 5.
Thaler and Sunstein 2008, p. 5.
Posner 2010.
6 A. Tor

fail to suit means to ends they necessarily deviate from the normative model of
Because the opportunity for nudging arises only when actual behaviour deviates
from the standards of rationality, the purpose of nudging can be translated to the
promotion of (some) rational action.18 However, Thaler and Sunstein instead simply
note that “. . . in many cases, individuals make pretty bad decisions—decisions they
would not have made if they had paid full attention and possessed complete infor-
mation, unlimited cognitive abilities, and complete self-control.”19 Because these
decisions must be bad from the actor’s perspective (the only perspective that should
be relevant for nudging) they can be bad only when failing to promote the actor’s
own ends. In Nudge’s terminology, then, bad decisions are synonymous with the
consequences of deviations from rational action and the goal of nudging turns out
to be simply the promotion of more rational behaviour.
Thaler and Sunstein’s statements regarding the definition of nudging can be
summarized as follows: True nudging employs (a) only rationally-neutral behav-
ioural tools and (b) does so only to accomplish the goal of helping individuals act
more rationally in suiting means to ends. The problem with this framework is that
while nudging can be used to address the consequences of bounded rationality, it
is not necessarily always or only the best tool for accomplishing this goal.
Similarly, the tools of nudging can be used equally well and perhaps even more
effectively to promote policy goals other than the improvement of individual ratio-
nality. Nudge’s framework thus suffers from a problem of fit between its selected
tools and chosen goal.
This problem of fit is not only cosmetic, but has significant consequences for the
book’s agenda. For one, the focus on fitting the tools offered by bounded rationality
to help overcome the consequences of bounded rationality limits the scope of
Nudge’s stated ambition and practical relevance. These strictures mean that faithful
nudging cannot use behavioural tools to advance social welfare goals (say, environ-
mental protection) or even traditionally paternalistic goals that do not necessarily
promote individuals self-judged well-being (e.g., increasing seat-belt use or reduc-
ing drug abuse). In the same vein, nudging cannot legitimately involve the use of
non-behavioural means, such as incentives or legal mandates, to address the nega-
tive consequences of bounded rationality. Faithful nudging therefore is highly trun-
cated with respect to both its tools and its goals. Moreover, the book’s framing of
the policy problems it seeks to address and the cures it offers also obscures the
potentially superior fit of other tools to Nudge’s task and of its own behavioural
toolkit to other policy goals.

At best, successful nudging diminishes individuals’ deviations from rationality, helping them to
overcome their errors of judgment or choice and thereby promotes their rationality in fact. Effective
nudging, however, may lead people to avoid the consequences of their bounded rationality and
therefore appear as if they were more rational even without facilitating their rationality in fact, an
important distinction that is outside the scope of the present analysis.
Thaler and Sunstein 2008, p. 5.
1 The Critical and Problematic Role of Bounded Rationality in Nudging 7

1.3 Non-nudge Tools and Extra-nudge Goals

The problem of fit combines with the desire to exploit the tools of nudging more
broadly and address the goals of nudging more effectively to lead the book to cover
much ground that is outside its own framework. Many of the applications it offers,
in fact, end up transgressing its self-imposed boundaries for legitimate nudging. For
present purposes, however, two examples—one concerning non-nudge tools, the
other involving extra-nudge goals—will suffice to illustrate this point.
Some of Nudge’s applications aim to shape behaviour using tools that are not
rationally-neutral and therefore do not qualify as nudges. One example concerns the
power of social norms to shape behaviour. In one case, Thaler and Sunstein (2008)
praise the web page used by the Illinois’ organ donation program to attract prospec-
tive donors. According to their description, the page draws on the power of social
norms and social influence by suggesting that social norms favour organ donations.
But why should people be more inclined to donate organs just because they believe
that such behaviour better comports with the social norm?
The book points to two main reasons for the power of social norms to shape
behaviour: First, social norms may convey meaningful information about what is
right, good, and so on. Second, deviations from social norms may incur social con-
sequences. However, both of these reasons in fact reveal that influencing behaviour
through social norms may exceed the boundaries of legitimate nudging. Norms that
provide meaningful information about appropriate standards influence behaviour by
shaping preferences. Arguably, though, interventions that shape individuals’ prefer-
ences by providing previously unknown social norm information are not rationally
neutral.20 Even more obviously problematic is the second reason, since social norms
that operate by imposing costs are changing people’s incentives and therefore are
also not rationally neutral. A rational actor may well change her behaviour to avoid
incurring the wrath of peers or to obtain their favour. And the costs and benefits of
such social interaction often extend to the economic domain as well (e.g., people
may be reluctant to transact with those who violate social norms).
Much like it affects the tools of nudging, the problem of fit also leads the book to
advocate some extra-nudge goals. Indeed, the organ donation example above
already offers a case on point. Some may consider increasing the rate of organ dona-
tions to be a socially valuable policy goal; others may disagree. Those who consider
organ donations important may also wish to donate their own organs if the opportu-
nity arises. Individuals who oppose organ donations and perhaps even some of
those who are supportive of them in the abstract may not judge their own well-being
to be best served by joining a donor registry. Yet the social-norm nudge described
above may well result in leading some such individuals to join the registry and
thereby make them worse rather than better off.

Although technically, Nudge’s definitions may be read to allow the shaping of preferences, so
long as such interventions do not literally constrain choice or affect economic incentives.
8 A. Tor

Similar problems arise with respect to the use of standard nudge tools towards
extra-nudge goals. One of the paradigmatic nudges in the book involves using
defaults (e.g., of a higher rate of deduction from one’s salary paycheck) to increase
the rate of retirement savings. Even this nudge, however, already aims at an extra-
nudge goal. To wit, some individuals may wish to save more for retirement at the
cost of reducing their present disposable income but others may hold different judg-
ments of how the trade-offs involved impact their well-being. Of course, policy
makers may wish to increase retirement savings to limit the negative social effects
of retirees with limited savings who consume a great deal of public resources. But
this reasonable policy goal is one of social welfare, not a goal of self-judged indi-
vidual well-being. And even policy makers who believe that the retirees who will
have saved more will be better off than if they depended more on public resources
still are implementing a traditional, extra-Nudge paternalistic policy. The default
savings nudge in this case makes retirees better off according to the beliefs of policy
makers. It may even promote the self-judged well-being of some retirees. But the
nudge here cannot be said to promote the well-being of those employees who judge
themselves better off with more disposable income pre-retirement and lower retire-
ment savings.21

1.4 Conclusion

All in all, this brief analysis of book’s framework reveals a problem of fit between
the tools of nudging and its goals. This problem turns out to have implications
beyond mere concerns of analytical clarity. The broader significance of behavioural
interventions for policy making renders efforts to faithfully limit nudging to liber-
tarian paternalistic interventions quite frustrating. It also generates an inexorable
pressure on scholars and policy makers to nudge towards traditional paternalistic
goals or even social welfare ones. Similarly, the focus on nudging as a means for
improving individual rationality obscures the potentially superior fit of more tradi-
tional policy tools to this task.22 This aspect of the problem of fit also exerts its own
additional pressure on Thaler and Sunstein’s (2008) policy prescriptions, thereby
facilitating the inclusion of applications that draw on tools beyond nudging.
The sometimes subtle deviations from Nudge’s self-imposed constraints
described here are significant for at least two reasons. First and most obviously, the
couching of the book’s (and similar) policy recommendations as low-risk nudges
can obscure some of the significant trade-offs involved. For example, a traditionally
paternalistic nudge may require more careful justification than a nudge that truly

Note that this analysis concerns the goals of nudging and stands irrespective of the fact that
rational employees would be indifferent to the savings default and save exactly as much as needed
to achieve their self-judged goals. After all, boundedly rational employees may be nudged by the
default irrespective of their self-judgments of well-being.
Bubb and Pildes 2014.
1 The Critical and Problematic Role of Bounded Rationality in Nudging 9

makes people better off as they themselves judge the matter. And a nudge, that aims
to promote social welfare goals without regard to individual well-being, involves
more complex trade-offs among individuals in society that require close examina-
tion and additional justification.
Second, the book’s implicit agenda of fighting the consequences of bounded
rationality (i.e. those ‘bad decisions’) with the nudging tools offered by bounded
rationality is intuitively appealing. But the effects of the problem of fit also divert
attention from some fundamental questions regarding the Nudge enterprise and its
relationship to rationality and bounded rationality that the book leaves unanswered.
For instance, when should deviations from rational action truly be considered ‘bad
decisions’? Does it matter whether the policies that seek to address such bad deci-
sions only aim to facilitate behaviour that resembles rational action or instead try to
promote the actual rationality of individuals in society? And when are either as if or
actual rationality appropriate goals for policy making?
These questions, both regarding the trade-offs involved in nudging towards vari-
ous policy goals and concerning the role of rationality as a goal for nudging or other
interventions, demand further attention and study. Nudge’s detractors may find here
further arguments against the desirability and feasibility of behaviourally-informed
policy making. Yet these questions should be of even greater interest to those who
identify with the book’s overall ambition. After all, the recognition that we cannot
avoid facing boundedly rational individuals as they are also means that policy mak-
ers must grapple with the complexity and challenges involved in this effort.


Baldwin, Robert. 2015. Nudge: Three degrees of concern. Working paper.

Bernheim, Douglas B. 2009. Behavioural welfare economics. Journal of the European Economic
Association 7: 267–319.
Bovens, Luc. 2010. Nudges and cultural variance: A note on Selinger and Whyte. Knowledge,
Technology, and Policy 23: 483–486.
Bubb, Ryan, and Richard H. Pildes. 2014. How behavioral economics trims its sails and why.
Harvard Law Review 127: 1593–1678.
Camerer, Colin, Samuel Issacharoff, George Loewenstein, Ted O’Donoghue, and Matthew Rabin.
2003. Regulation for conservatives: Behavioral economics and the case for ‘Asymmetric
Paternalism’. University of Pennsylvania Law Review 151: 1211–1254.
Conly, Sarah. 2012. Against autonomy: Justifying coercive paternalism. Cambridge: Cambridge
University Press.
Glaeser, Edward L. 2006. Paternalism and psychology. University of Chicago Law Review 73:
Hausman, Daniel M. 1992. The inexact and separate science of economics. Cambridge: Cambridge
University Press.
Hausman, Daniel M., and Brynn Welch. 2010. Debate: To nudge or not to nudge. Journal of
Political Philosophy 18: 123–136.
Johnson, Eric J., Suzanne B. Shu, Benedict G.C. Dellaert, Craig Fox, Daniel G. Goldstein, Gerald
Haubl, Richard P. Larrick, John W. Payne, Ellen Peters, David Schkade, Brian Wansink, and
Elke U. Weber. 2012. Beyond nudges: Tools of a choice architecture. Marketing Letters 23:
10 A. Tor

Klick, Jonathan, and Gregory Mitchell. 2006. Government regulation of irrationality: Moral and
cognitive hazards. Minnesota Law Review 90: 1620–1663.
Maloberti, Nicolas. (unpublished manuscript). 2012. Libertarian paternalism and limited govern-
ment: A reconsideration.
Menard, Jean-Frederick. 2010. A ‘nudge’ for public health ethics: Libertarian paternalism as a
framework for ethical analysis of public health interventions? Public Health Ethics 3:
Mitchell, Gregory. 2005. Libertarian paternalism is an oxymoron. Northwestern University Law
Review 99: 1245–1277.
Mongin, Philippe, and Mikael Cozic. 2014. Rethinking nudges. Working paper.
Posner, Richard A. 2010. Economic analysis of law, 8th ed. New York: Aspen Publishers.
Rachlinski, Jeffrey J. 2006. Cognitive errors, individual differences, and paternalism. University of
Chicago Law Review 73: 207–229.
Rebonato, Riccardo. 2014. A critical assessment of libertarian paternalism. Journal of Consumer
Policy 37: 357–396.
Sibony, Anne-Lise, and Alberto Alemanno. 2015. Nudge and the law: A European perspective. In
Nudge and the law: A European perspective, ed. Anne-Lise Sibony and Alberto Alemanno,
1–24. Oxford: Hart Publishing.
Spiegler, Ran. 2015. On the equilibrium effects of nudging. Journal of Legal Studies
Sunstein, Cass R., and Richard H. Thaler. 2003. Libertarian paternalism is not an oxymoron.
University of Chicago Law Review 70: 1159–1202.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge: Improving decisions about health, wealth,
and happiness. New Haven: Yale University Press.
Tor, Avishalom. 2002. The fable of entry: Bounded rationality, market discipline, and legal policy.
Michigan Law Review 101: 482–568.
Tor, Avishalom. 2008. The methodology of the behavioral analysis of law. Haifa Law Review 4:
White, Mark D. 2013. The manipulation of choice: Ethics and libertarian paternalism. London:
Palgrave Macmillan.
Willis, Lauren E. 2013. When nudges fail: Slippery defaults. University of Chicago Law Review
80: 1155–1229.
Wright, Joshua D., and Douglas A. Ginsberg. 2012. Behavioral law and economics: Its origins,
fatal flaws, and implications for liberty. Northwestern University Law Review 106:
Zamir, Eyal. 1998. The efficiency of paternalism. Virginia Law Review 84: 229–286.
Chapter 2
Beneficial and Exploitative Nudges

Bruno S. Frey and Jana Gallus

Abstract The effectiveness of nudges in raising the welfare of the population hinges
on the policymakers employing them. A frequent criticism based on a logical incon-
sistency questions policymakers’ immunity from the psychological biases of indi-
viduals that are the very foundation of nudging interventions. We argue that, rather
than being concerned about policymakers’ incapacity to raise the population’s wel-
fare, we should be concerned about their unwillingness to do so. We offer a solution
to this problem. We resort to the constitutional level of decision-making in which
voters are able to determine the procedures or processes by which governments may
resort to nudging. Nudging should not be considered as an innocuous exception to
constitutionally based decision-making. It must be admitted, though, that at first sight
most nudges do seem to be beneficial to people. In a democracy, even ‘Liberal
Paternalism’ may not be imposed on the population without its consent in principle.

2.1 Nudging as a Concept

Nudging is an attractive idea. With low or even zero-cost interventions paternalists

help individuals overcome their limitations and act in their own best interest, as if
they had complete information and the perfect willpower and cognitive abilities of
the homo oeconomicus. The deviations from the standard economic model of ratio-
nality have been empirically identified and extensively analysed in ‘Psychology and
Economics’.1 It has now been securely established that human beings are not always
and in every respect perfectly maximizing their utility under complete information.
They are subject to biases or distortions defined as systematic (i.e. non-random)

‘Psychology and Economics’ is often labelled ‘Behavioural Economics’. We use the former
expression since economics has always dealt with human behaviour. Surveys of this literature can
be found in several books (e.g., Frey 1999, 2001; Kirchgässner 2008; Kahneman 2011) and journal
articles (e.g., Conlisk 1996; Rabin 1998; DellaVigna 2009). Congdon et al. (2011) and DellaVigna
(2009) offer useful albeit slightly different categorizations of psychological biases.
B.S. Frey (*) • J. Gallus
Center for Research in Economics, Management and the Arts (CREMA), University of
Zurich, Zurich, Switzerland

© Springer International Publishing Switzerland 2016 11

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_2
12 B.S. Frey and J. Gallus

deviations from the standard model of rationality. Individuals take shortcuts and
decide on the basis of heuristics.2 Such heuristics work well in most routine
situations,3 but they may also fail and lead to value-destroying outcomes.4
Nudging provides an elegant solution to the limitations of the homo oeco-
nomicus5: Individuals are ‘softly’6 induced, or nudged, to make better decisions.
The government, i.e., politicians and public officials, offer solutions guiding people
in the right utility maximizing direction. Importantly, individual decision makers
are not forced to pursue that option if they choose not to do so. Nudging does not
reduce individuals’ freedom of choice. Hence these policy interventions are often
subsumed under the label of ‘liberal paternalism’.7 The intervention by a ‘social
planner’ is considered to be benign; the ‘choice architects’8 act paternalistically by
designing the environment so that individuals are steered towards welfare promot-
ing decisions. Individuals would arrive at the same decision if they cognitively
reflected them with sufficient information instead of relying on intuitive, or auto-
matic, decision procedures.9
There are other concepts similar to Thaler and Sunstein’s liberal paternalism,
such as ‘asymmetric paternalism’.10 All of them are designed to help people improve
their decisions and come closer to acting according to their own best interest.
Nudging, or liberal paternalism, has become highly popular in economics and far
beyond. The interventions, it is assumed, are cheap if not costless. So even if their
effect may not always be substantial, nudges can be more cost-effective than other
policy interventions. Nudging is convincingly applied in particular in the areas of
retirement savings, organ donation, and health (e.g., healthy diets, vaccination).11
The interventions offered by liberal paternalists immediately stand to reason and
appear totally acceptable. It is, for instance, difficult to reject nudging when applied
to securing a reasonable material standard in old age. It is well known that most
individuals find it difficult to look sufficiently into the future; they are subject to
‘hyperbolic discounting’.12 As a consequence, they are reluctant to take the initiative

See Tversky and Kahneman 1974, Gigerenzer et al. 1999, and, more recently, Camerer 2004.
Gigerenzer, Hertwig and Pachur 2011.
Conlisk 1996, Rabin 1998, and Kahneman 2003.
In contrast, ‘hard paternalism’ as proposed, e.g., by Conly 2012, seeks to induce people to act in
the desired way by forcing them to do so by laws and regulations whose violation is punished.
Although in several cases nudging does not provide a solution to, but rather exploits, behavioral
The idea has been propagated by Sunstein and Thaler 2003, Thaler and Sunstein 2003, 2008, and
Sunstein 2014. See also Binder and Lades 2015. The term ‘liberal’ is used in the classical sense as
discussed by Mill 1859 in his book On Liberty.
A term coined by Thaler and Sunstein 2008.
Although several policy interventions pursue social rather than individual optimality.
Camerer et al. 2003. Related concepts have been proposed by Gruber and Koszegi 2001,
O’Donoghue and Rabin 2003, 2006, Jolls and Sunstein 2006, Bernheim and Rangel 2007.
Excellent discussions are given e.g. by Schnellenbach 2011, 2012, Kirchgässner 2014, Madrian
2014, and Schnellenbach and Schubert 2014.
Laibson 1997.
2 Beneficial and Exploitative Nudges 13

to enter an old age pension scheme. The situation is quite different when the default
option is being enrolled in a pension scheme from which they may opt out if they so
wish. Empirical research clearly suggests that the effect is substantial. Many more
individuals are covered by voluntary old age insurance schemes, of which they
greatly benefit when reaching retirement age.13 Another related bias, the mispredic-
tion of future utility, has been identified and analysed in happiness research.14
People systematically underestimate the future satisfaction they gain from relational
contacts with friends, neighbours and relatives. In contrast, they overestimate the
future utility of material goods such as a new home or car. As people get used to a
new home or car rather quickly, the additional utility (or life satisfaction) gained
from these goods vanishes over time. This effect is less likely to take place in per-
sonal relationships because they can continually generate value.
The popularity of the nudging concept has also been supported by the fact that it
was propagated by two scholars15 from the University of Chicago, whose depart-
ment of economics is well known for its strong allegiance to free markets and its
rejection of government interventions wherever possible.16 The kind of paternalism
implied by nudging seems acceptable as it does not reduce individuals’ freedom but
increases their present and future utility by overcoming the decision biases from
which they suffer. It is also consistent with the insight that a market economy with-
out any government intervention is inconceivable, not only because the government
must set the rules under which the market works but also because there are ‘merit’
goods individuals cannot properly evaluate for many different reasons.17 Liberal
paternalism offers a reasonable way in which government interventions should take
place without violating the basic value of human freedom.
Liberal paternalism has various other favourable features. It is certainly prefer-
able to hard paternalism with its negative sanctions. It is also commendable that
academics seriously focus on real-life issues and suggest policy measures. The
establishment of ‘nudging units’, encouraging evidence-based policy making,
began in the United Kingdom in 2010 and the United States, and has more recently
been followed by countries such as Australia, Canada, Columbia, Denmark, France,
Germany, Israel, Netherlands, New Zealand, Norway, Singapore, South Africa,
Turkey and the European Union (with its planned ‘foresight team’), and even Saudi
Arabia.18 It may however be doubtful whether all of these countries are sufficiently
democratic to use nudges in their soft form.

E.g., Madrian and Shea 2001 and Choi et al. 2004.
Loewenstein et al. 2003 and Frey and Stutzer 2014. For the overestimation of the future utility
provided by material goods, see Lane 1991 and Frank 1999.
Richard Thaler and Cass Sunstein, see ftn 8. (Sunstein is now associated with Harvard University).
Leading members who promoted this ideology are Milton Friedman, George Stigler and Gary
As is well discussed by Kirchgässner 2014. The concept of merit goods was introduced by
Musgrave 1957.
See, e.g., Madrian 2014, p. 665.
14 B.S. Frey and J. Gallus

2.2 Critiques of the Nudging Concept

Despite the wide acceptance of liberal paternalism there is some well-founded criti-
cism to consider. Indeed, there has been a lively scholarly discussion about the
value of the concept.19 Here only two major points of criticism will be
The first point of criticism is that the emphasis of liberal paternalism to suggest
welfare enhancing outcomes is ill-taken as it interferes too much with individuals’
freedom to decide. What should be done is to foster their critical thinking and their
capacity to decide. This ‘Autonomy-Enhancing Paternalism’ strengthens the cogni-
tive aspect of personal decisions, promotes self-empowerment, and seeks to make
people independent of external influences.21 Individuals are helped to overcome
their distortions and to take well-considered choices. This gain in autonomy enables
people to engage in mental strategies, thus de-biasing their decisions.22 The goal is
not to attempt to directly change outcomes but rather to improve the processes used
to reach a decision. Autonomy-Enhancing Paternalism takes a dynamic perspective
and wants to improve the decision-making process.
The second point of criticism notes that the political and bureaucratic decision-
makers are themselves subject to biases and distortions.23 They therefore suggest
options to the individual decision-makers that do not improve their current or future
welfare. This assumption is in line with Public Choice Theory, which has always
argued that public decision-makers are subject to the same distortions as are other
people.24 It could even be argued that they act in a less careful manner because they
decide about other people’s, and not their own, money.25 A glance at reality con-
firms this notion. Take, for example, the decision by the German government to pull
down the Wall, which until 1989 separated the German Federal Republic from the
German Democratic Republic. Only a few years later, the same government had to
newly erect parts of the Wall in order to remind people that there had been this
deadly division between the two countries. Some private persons had better predic-
tive capacity. They anticipated that the Wall would become a tourist attraction and
that it would be worthwhile to maintain it at least in some crucial parts.
However, the second criticism does not necessarily hold. There are certainly
conditions under which public decision-makers have better information and more
extensive advice from experts so that they are better able to overcome biases than

See e.g. the recent critique by Whitman and Rizzo 2015 of the welfare standards used by soft
See e.g. Sugden 2013, Schnellenbach and Schubert 2014.
Binder and Lades 2015, also Hausman and Welch 2010, Mills 2013.
See Larrick 2004.
Glaeser 2006.
See e.g. Mueller 1996, 2003 or Frey 1983.
This argument has e.g. been advanced by Tullock 1965.
2 Beneficial and Exploitative Nudges 15

are more isolated individuals. Moreover, the lengthy administrative and political
process should also help to reduce the most obvious distortions.
These criticisms shall not be further pursued because they have been well dis-
cussed in the literature. Rather, our contribution wants to consider an even more
fundamental aspect, namely, that Liberal Paternalism may turn out to be disadvan-
tageous to individuals, reducing their welfare.

2.3 Nudges from a Political Economy Perspective

The proponents of Liberal Paternalism seem to assume, as a matter of course, that

the public decision-makers want to raise the welfare of the general population. They
are thus following the ‘benevolent dictator approach’, assuming that politicians aim
to maximize social welfare and can force public officials to act accordingly.
This is an assumption that has been strongly criticized since the very beginning
of Public Choice Theory.26 Public decision-makers are as self-interested as other
persons are. Their general goal is to stay in power. In democracies this depends on
being re-elected. Politicians are prepared to produce a political business cycle in
order to reach that goal.27 They actively exploit a particular bias, namely the short
time horizon of the voters, to that effect. Before the election they spend money on
particularly popular projects, or hand out money to the population in the form of
higher old age pensions and higher transfers to lower income groups. Most impor-
tantly, they support strong interest groups by giving them special privileges, for
instance by introducing import restrictions and reducing the taxes for the export
industry, such as the tourist sector. More subtly, they may offer a sympathetic ear to
lobbyists’ concerns about, and ideas for, certain nudging interventions. The costs of
such policies are the more strongly discounted the lower is the probability of staying
in power. Thus, a government faced with a low probability of staying in power is
induced to spend more money shortly before elections because they are unlikely to
bear the negative consequences during the forthcoming election period. In contrast,
when a government expects to be re-elected, its time horizon extends over the next
election period and beyond. Under this condition it can undertake ideological poli-
cies. These preferences may, but need not, be in line with the population’s welfare.
The government may be ready to offer options nudging people into a direction
increasing the population’s welfare if this is in line with the government’s ideology.
Such social welfare enhancing policies may also be in the interest of public officials
if they can thereby demonstrate their competence and their superior knowledge
compared to ordinary people. Their behaviour is also influenced by a professional
attitude inducing them to indeed seek ways to improve the welfare of the popula-
tion. But again, this need not be the case, and there are many instances in which
bureaucracies act in their own interest, in particular with respect to the wages and

Buchanan and Tullock 1962, Frey 1983, Mueller 1997. See also the textbook by Mueller 2003.
E.g. Frey 1978, Tufte 1978, Frey and Schneider 1981.
16 B.S. Frey and J. Gallus

privileges they enjoy. In that case they actively seek to exploit to their own advan-
tage the biases in decision-making they observe in the population.
Liberal Paternalism has been discussed almost exclusively in the context of dem-
ocratic countries where governments depend to some extent, and intermittently, on
the will of the voters. But many countries in the world are governed by authoritarian
rulers or by outright dictators.28 While they constantly claim to rule in the interest of
their people, this is of course not unambiguously the case. They will seek to actively
exploit the distortions in decision-making that members of the population are prone
to. They will offer additional options with the goal of nudging people in the direc-
tion benefiting them and their clientele.
The assumption that nudges will always be used in the interest of the population
is most questionable, if not plainly wrong. It should be noted that private enterprises
have always exploited individual decision distortions to sell products that are not
always in the interest of their consumers or society at large.29 An example are self-
service shops where the cheaper products are put in an inconvenient location
(mostly at the very bottom of the shelf), thereby nudging people to buy the more
expensive products with a higher profit margin for the seller. Another example is
candy bars whose convenient placement at the check-out prompts children to crave
for them.
The next section discusses a specific government intervention with a long tradi-
tion in all countries, awards. These take the form of state orders, medals, decora-
tions, and other distinctions.30

2.4 An Application of Nudging: Awards

Awards are given to individuals (and sometimes to groups) who are deemed to have
performed particularly well.31 They differ in several respects from monetary rewards
such as bonuses, which are more directly linked to specific performance. Accepting
an award nudges the recipient into a special relation of loyalty to the donor. The
respective contract is tacit, incomplete, and difficult or impossible to enforce by the
donors. The givers expect that a recipient of an award talks and behaves in a sup-
portive way towards them. If the recipient violates this tacit contract, this is consid-
ered bad form. Other people would argue that the recipient should not have accepted

According to the widely used measure of political regimes, Polity IV, in 2013 about 40 % of all
countries failed to meet the rather broad criteria of democratic rule. See http://www.systemicpeace.
We owe this point to Margit Osterloh.
Switzerland is an exception.
For the emerging literature on awards in economics see Ginsburgh 2003, Besley 2005, Frey
2005, 2006, 2007, Neckermann and Frey 2008, Malmendier and Tate 2009, Chan et al. 2014, Frey
and Gallus 2014, Neckermann et al. 2014, Gallus 2015, and Gallus and Frey 2015. For the socio-
logical literature see, e.g., Bourdieu 1979, 1985.
2 Beneficial and Exploitative Nudges 17

the award if he or she does not agree with the donor’s ideology and policies. In turn
the donor is expected to support the recipients materially or at least psychically if
they run into trouble. In the case of monetary compensation there is no obligation of
loyalty at all. Even the opposite may be true: payments can easily be used as justifi-
cation to work for an organization that one publicly denounces.
Awards look innocuous at first sight. After all, they are a positive reward and
only offer an option the potential recipient is free to decline. They therefore create
no serious problems in democracies, though it may be considered somewhat arro-
gant to decline an award. Notable examples concerning major awards are Jean- Paul
Sartre’s rejection of the Nobel Prize in Literature in 1964, and Thomas Piketty’s
refusal to accept the Légion d’honneur in 2015. The situation is quite different in an
authoritarian system. A person being offered an award by the ruler puts can find
himself or herself in a most unfortunate situation. If he or she rejects the award, the
ruler gets angry and can claim that his honour and that of the whole state is deni-
grated. If he or she accepts the award this signals a large measure of identification
with the ruler. It is, for example, not well possible to accept the Order of a Hero of
Labour of the Russian Federation (formerly Hero of Socialist Worker) but not to
agree with the ruling ideology and policy.
Authoritarian rulers can exploit this hidden feature of awards by offering an
award to regime critics. This puts the potential recipients in an uncomfortable posi-
tion because both, whether they accept or reject the order, the decision will strongly
reduce their welfare, while benefiting the ruler. In many cases, the potential recipi-
ent is nudged to take the award because offending the ruler imposes too high costs
on him or her. This is an instance in which the added option (being offered an
award) has a welfare decreasing effect on the persons concerned.

2.5 Making Nudges Beneficial

We have argued that nudges are not always in the interest of the population. Unlike
most of the literature on Liberal Paternalism we suggest that governments may
nudge people in a direction that fails to raise their welfare. Even seemingly glorious
bestowals of honours in the form of state orders and decorations can carry substan-
tial costs to the ostensible winners. Indeed, under significant conditions, especially
when the political system is non-democratic, rulers tend to exploit nudges for their
own benefit and to the detriment of the population’s welfare.
In a recent article,32 one of the architects behind the nudging concept admits, “[i]f
we are especially fearful of official mistakes – coming from incompetence or bad
motivations – we will want to minimize the occasions for nudging.” Rather than
minimizing the occasions for nudging, we propose that they should be subjected to

Sunstein 2015 forthcoming, p. 13.
18 B.S. Frey and J. Gallus

constitutional rules33 in which voters are able to decide the procedures or processes
under which governments may resort to nudging.
Nudging should not be accepted as an innocuous exception to constitutionally
based decision-making, even if most applications appear to benefit people. In
democracies, even ‘Liberal Paternalism’ must be based on the population’s consent
in principle.

Acknowledgement We are grateful for the helpful discussions with Margit Osterloh.


Bernheim, Douglas B., and Antonio Rangel. 2007. Behavioral public economics: Welfare and
policy analysis with non-standard decision makers. In Behavioral economics and its applica-
tions, ed. Peter Diamond and Hannu Vartainen, 7–84. Princeton: Princeton University Press.
Besley, Timothy. n.d. Notes on honours. Unpublished working paper. London School of
Bourdieu, Pierre. 1979. La distinction. Critique Sociale du Jugement. Paris: Les Editions de
Bourdieu, Pierre. 1985. The market of symbolic goods. Poetics 14: 13–44.
Brennan, Geoffrey, and James M. Buchanan. 1985. The reason of rules. Constitutional political
economy. Cambridge: Cambridge University Press.
Buchanan, James M. 1965. Constitutional economics. Oxford: Blackwell.
Buchanan, James, and Gordon Tullock. 1962. The calculus of consent: Logical foundations of
constitutional democracy. Ann Arbor: University of Michigan Press.
Camerer, Colin F. 2004. Advances in behavioural economics. Princeton: Princeton University
Camerer, Colin F., Samuel Issacharoff, George Loewenstein, Ted O’Donoghue, and Matthew
Rabin. 2003. Regulation for conservatives: Behavioural and the case for ‘Asymmetric
Paternalism’. University of Pennsylvania Law Review 151: 1211–1254.
Chan, Ho Fai, Bruno S. Frey, Jana Gallus, and Benno Torgler. 2014. Academic honors and perfor-
mance. Labour Economics 31: 188–204.
Choi, James J., David Laibson, Brigitte C. Madrian, and Andrew Metrick. 2004. For better or for
worse: Default effects and 401 (k) savings behavior. In Perspectives on the economics of aging,
ed. David A. Wise, 81–126. Chicago: University of Chicago Press.
Congdon, William J., Jeffrey R. Kling, and Sendhil Mullainathan. 2011. Policy and choice: Public
finance through the lens of behavioral economics. Washington, DC: Brookings Institution
Conlisk, John. 1996. Why bounded rationality? Journal of Economic Literature 34(2): 669–700.
Conly, Sarah. 2012. Against autonomy: Justifying coercive paternalism. Cambridge: Cambridge
University Press.
DellaVigna, Stefano. 2009. Psychology and economics: Evidence from the field. Journal of
Economic Literature 47(2): 315–372.
Frank, Robert. 1999. Luxury fever. New York: Free Press.
Frey, Bruno S. 1978. Politico-economic models and cycles. Journal of Public Economics 9:
Frey, Bruno S. 1983. Democratic economic policy. Oxford: Blackwell.
Frey, Bruno S. 1999. Economics as a science of human behaviour. Boston: Kluwer.

Buchanan 1965, Brennan and Buchanan 1985, Mueller 1997.
2 Beneficial and Exploitative Nudges 19

Frey, Bruno S. 2001. Inspiring economics. Human motivation in political economy. Cheltenham:
Edward Elgar.
Frey, Bruno S. 2005. Knight fever: Towards an economics of awards. CESifo Working Paper No.
Frey, Bruno S. 2006. Giving and receiving awards. Perspectives on Psychological Science 1:
Frey, Bruno S. 2007. Awards as compensation. European Management Review 4: 6–14.
Frey, Bruno S., and Jana Gallus. 2014. Awards are a special kind of signal. CREMA Working
Paper No. 2014–04.
Frey, Bruno S., and Friedrich Schneider. 1981. A politico-economic model of the United Kingdom.
Economic Journal 88: 243–253.
Frey, Bruno S., and Alois Stutzer. 2014. Economic consequences of mispredicting utility. Journal
of Happiness Research 15: 937–950.
Gallus, Jana, and Bruno S. Frey. 2015. Awards: A strategic management perspective. Unpublished
working paper.
Gallus, Jana. 2015. Fostering voluntary contributions to a public good: A large-scale natural field
experiment at wikipedia. Unpublished working paper.
Gigerenzer, Gerd, Peter M. Todd, and ABC Research Group. 1999. Simple heuristics that make us
smart. New York: Oxford University Press.
Gigerenzer, Gerd, Ralph Hertwig, and Thorsten Pachur. 2011. Heuristics: The foundations of
adaptive behavior. New York: Oxford University Press.
Ginsburgh, Victor. 2003. Awards, success and aesthetic quality in the arts. Journal of Economic
Perspectives 17: 99–111.
Glaeser, Edward L. 2006. Paternalism and psychology. University of Chicago Law Review 73:
Gruber, Jonathan, and Botond Koszegi. 2001. Is addiction ‘Rational’? Theory and evidence.
Quarterly Journal of Economics 116: 1261–1303.
Hausman, Daniel M., and Brynn Welch. 2010. Debate: To nudge or not to nudge. Journal of
Political Philosophy 18(1): 123–136.
Jolls, Christine, and Cass C. Sunstein. 2006. Debiasing through law. Journal of Legal Studies 35:
Kahneman, Daniel. 2003. Maps of bounded rationality: Psychology for behavioral economics.
American Economic Review 93(5): 1449–1475.
Kahneman, Daniel. 2011. Thinking, fast and slow. New York: Farrar, Strauss, Giroux.
Kirchgässner, Gebhard. 2008. Homo oeconomicus: The economic model of behaviour and its
applications in economics and other social sciences. Heidelberg: Springer.
Kirchgässner, Gebhard. 2014. Sanfter Paternalismus, Meritorische Güter und der Normative
Individualismus. List Forum für Wirtschafts- und Finanzpolitik Sonderheft 3: 210–238.
Laibson, David. 1997. Golden eggs and hyperbolic discounting. Quarterly Journal of Economics
112(2): 443–477.
Lane, Robert E. 1991. The market experience. Cambridge: Cambridge University Press.
Larrick, Richard P. 2004. Debiasing. In Blackwell handbook of judgment and decision making, ed.
Derek J. Koehler and Nigel Harvey, 316–337. Oxford: Blackwell Publishing.
Loewenstein, George, Ted O’Donoghue, and Matthew Rabin. 2003. Projection bias in the predic-
tion of future utility. Quarterly Journal of Economics 118: 1209–1248.
Madrian, Brigitte C. 2014. Applying insights from behavioral economics to policy design. Annual
Review of Economics 6: 663–688.
Madrian, Brigitte C., and Dennis F. Shea. 2001. The power of suggestion: Inertia in 401 (k) partici-
pation and savings behavior. Quarterly Journal of Economics 116: 1149–1187.
Malmendier, Ulrike, and Geoffrey Tate. 2009. Superstar CEOs. Quarterly Journal of Economics
124: 1593–1638.
Martin, Binder, and Leonhard K. Lades. 2015. Autonomy-enhancing paternalism. Kyklos 68(1):
20 B.S. Frey and J. Gallus

Mill, John Stuart. 1859. On liberty. New York: The Harvard Classics.
Mills, Chris. 2013. Why nudges matter: A reply to goodwin. Politics 33(1): 28–36.
Mueller, Dennis C. (ed.). 1996. Perspectives on public choice. Cambridge: Cambridge University
Mueller, Dennis C. (ed.). 1997. Constitutional democracy. Oxford: Oxford University Press.
Mueller, Dennis C. 2003. Public choice III. Cambridge: Cambridge University Press.
Musgrave, Richard A. 1957. A multiple theory of budget determination. Finanzarchiv NF 24:
Neckermann, Susanne, and Bruno S. Frey. 2008. Awards: A view from psychological economics.
Zeitschrift für Psychologie/Journal of Psychology 216: 198–208.
Neckermann, Susanne, Reto Cueni, and Bruno S. Frey. 2014. Awards at work. Labour Economics
31: 205–217.
O’Donoghue, Ted, and Matthew Rabin. 2003. Studying optimal paternalism, illustrated by a model
of sin taxes. American Economic Review 93: 186–191.
O’Donoghue, Ted, and Matthew Rabin. 2006. Optimal sin taxes. Journal of Public Economics
90(10–11): 1825–1849.
Rabin, Matthew. 1998. Psychology and economics. Journal of Economic Lierature 36(1): 11–46.
Schnellenbach, Jan. 2011. Wohlwollendes Anschubsen: Was ist mit liberalem Paternalismus zu
erreichen und was sind seine Nebenwirkungen? Perspektiven der Wirtschaftspolitik
12: 445–459.
Schnellenbach, Jan. 2012. Nudges and norms: On the political economy of soft paternalism.
European Journal of Political Economy 28: 266–277.
Schnellenbach, Jan. 2014. Individuelle Rationalität als Ziel paternalistischer Intervention: Eine
verfassungsökonomische Perspektive. List Forum für Wirtschafts- und Finanzpolitik
Sonderheft 3: 239–257.
Schnellenbach, Jan, and Christian Schubert. 2014. Behavioral public choice: A survey. Available
at SSRN:
Sugden, Robert. 2013. The behavioural economist and the social planner: To whom should behav-
ioural welfare economics be addressed? Inquiry 56(5): 519–538.
Sunstein, Cass R. 2014. Why nudge? The politics of libertarian paternalism. New Haven: Yale
University Press.
Sunstein, Cass R. 2015. Nudging and choice architecture: Ethical considerations. Yale Journal on
Regulation. Forthcoming.
Sunstein, Cass R., and Richard H. Thaler. 2003. Libertarian paternalism is not an oxymoron.
University of Chicago Law Review 70(4): 1159–1202.
Thaler, Richard H., and Cass R. Sunstein. 2003. Libertarian paternalism. American Economic
Review 93(2): 175–179.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge – improving decisions about health, wealth
and happiness. London: Penguin Books.
Tufte, Edward T. 1978. The political control of the economy. Princeton: Princeton University
Tullock, Gordon. 1965. The politics of bureaucracy. Washington, DC: Public Affairs Press.
Tversky, Amos, and Daniel Kahneman. 1974. Judgment under uncertainty: Heuristics and biases.
Science 185(4157): 1124–1131.
Whitman, Douglas G., and Mario J. Rizzo. 2015. The problematic welfare standards of behavioral
paternalism. Review of Psychology 6: 409– 425.
Chapter 3
The Crucial Importance of Interests
in Libertarian Paternalism

Mark D. White

Abstract There has been a resurgence of late in support of paternalism, due

primarily to recent findings in behavioural science that cast doubt on the ability of
individuals to make choices reliably in their best interests. The most popular form
of paternalism currently is libertarian paternalism (or ‘nudge’), which advocates
subtle changes in the presentation or framing of options to steer persons in the direc-
tion of better choices. A key issue with libertarian paternalism (as well as more
coercive forms of paternalism that share similar goals) is the nature of people’s
interests and the state of policymakers’ knowledge about them. In this chapter, I
argue that this issue has been neglected in recent scholarship defending paternalism
along behavioural grounds. In particular, I dispute claims that paternalism of means
can be meaningfully separated from paternalism of ends, and argue that modern
paternalism does not respect people’s true interests but instead adopts a perfection-
ist or objective conception of well-being.

3.1 Introduction

Stemming from the pioneering work of Herbert Simon on bounded rationality and
the work of Daniel Kahneman, Amos Tversky, and colleagues on cognitive biases
and heuristics, behavioural economics has revolutionized the way economists and
policymakers think about choice.1 No more can agents be assumed simply to maxi-
mize their self-interest with perfect willpower and calculative precision. The model
of homo economicus should now be regarded as a Platonic ideal much like perfect
competition, an abstract starting point for discussion, a cipher to be fleshed out with
realistic details and flaws appropriate to the description and prediction of agents,
firms or individuals, in the real world.

Simon 1955; Kahneman et al. 1982. For accessible summaries, see Kahneman 2011 and Angner
M.D. White (*)
Department of Philosophy, College of Staten Island, Staten Island, NY, USA

© Springer International Publishing Switzerland 2016 21

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_3
22 M.D. White

But behavioural economics did not remain a concern merely of positive, descrip-
tive economics for long. Researchers, especially those in the law, soon saw the
potential for legal reform based on the insights of behavioural economics; the field
of behavioural law and economics studied not only how to improve compliance
with laws and the framing of evidence during trials, but also how to improve the
well-being of individuals.2 Once scholars could determine precisely how and why
people were making errors in their decision-making, they could design choice inter-
ventions to help steer them toward making better choices to further their own inter-
ests – choice interventions that often made use of the same cognitive biases and
heuristics that motivated their use in the first place. These interventions became
known as nudges, as in the title of Richard Thaler and Cass Sunstein’s bestselling
book that laid out the reasons why nudges should be used and gave examples of
their demonstrated and potential success.3 Since the publication of Nudge, a litera-
ture has quickly grown, with scholars providing arguments for and against the use
of nudges in different situations and contexts.4
This chapter will focus on one aspect of the debate over nudges and paternalism
that I feel has largely been neglected: a person’s interests and what we know about
them. After briefly summarizing problems with nudges along epistemic, ethical, and
practical lines, I will describe the nature of interests, which I maintain are complex,
multifaceted, and subjective. After detailing the background of this conception of
interests, I will then examine two recent defences of paternalism by Cass Sunstein
and Sarah Conly and two emphases they share: the importance of the distinction
between means paternalism and ends paternalism, and the rejection of perfectionism
or objective theories of well-being.5 I will argue that the nature of interests renders
both of these positions problematic and leaves paternalism open to traditional criti-
cisms that date back at least to John Stuart Mill.

3.2 Problems with Nudges

The problems with nudges can be split into three interrelated areas: epistemics, eth-
ics, and practical effects. The epistemic problem is the focus of this chapter, but it
is magnified by the other two, so in this section I will introduce all three briefly.6
The epistemic problem starts with the claim that economists and policymakers
who design nudges are steering people’s choices in their own interests. For instance,
automatic enrolment of new employees in retirement programs helps people save

For instance, see Jolls et al. 1998; Korobkin and Ulen 2000; Sunstein 2000; Parisi and Smith
Thaler and Sunstein 2008.
See White 2013, pp. 173–178, for a modest list of relevant literature.
Sunstein 2014 and Conly 2013.
For more detail and references, see White 2013; for similar arguments, see Saint-Paul 2011 and
Rebonato 2012.
3 The Crucial Importance of Interests in Libertarian Paternalism 23

more for retirement, and better design of cafeterias helps people eat better and have
better health outcomes. Both of these nudges, however, are designed to promote one
specific interest which is not necessary the interest that individuals are trying to
promote at any given time, and may not be among the interests they were acting
toward when they made choices that seem ‘mistaken’ to those designing the nudges.
A new employee may choose not to enrol in a retirement program because she
wants to have more money for a down payment on a home, to support her parents,
or simply to enjoy her youth in a new city. A person who chooses the huge slice of
chocolate cake in the cafeteria may be doing so to celebrate a colleague’s birthday,
to flirt with the person at the dessert counter, or simply because he likes chocolate
cake, each in full knowledge of the adverse health consequences. People have any
number of interests they are acting toward at any given time, interests that are com-
plex, multifaceted, and, most important, subjective – many times not fully known
even to the person making choices, much less an outside observer. Designers of
nudges do not and indeed cannot know these interests, and by necessity choose
interests toward which they steer people. Regardless of the intentions of the policy-
makers, they are imposing interests on people who have their own and then using
nudges to steer their decisions in those directions. I refer to this as the epistemic
problem because it focuses on the policymakers’ lack of information regarding the
true interests of the people whose interests they claim to further.
The ethical problem starts with the epistemic problem – specifically, that nudge
designers do not have knowledge of individuals’ true interests and therefore impose
other interests in which they influence choices – and connects it to the right of indi-
viduals to make choices in their own interests rather than those chosen for them.
One of the bedrock principles of a liberal society is that each person may pursue his
or her own idea of the good life, provided no one interferes wrongfully with others
doing the same. In steering people’s choices away from their true interests, however,
nudges deny people this self-directedness. Supporters argue that people are free to
make decisions other than the ones into which they are nudged, but the subtle
psychological manipulation that makes nudges so effective belies this claim. Even if
a person has a good reason not to enrol in a retirement program, the same tendencies
toward laziness that make people susceptible to default choices will lead him to
enrol if that is the default. Finally, everyone makes bad choices that do not further
his or her interests; however, the only one who knows if a particular choice is bad is
the person himself or herself, who knows whether that choice furthered his or her
interests. Because an outside observer and policymaker does not have knowledge of
these interests, it is impossible for such a person to know whether someone else has
made a bad choice (absent an admission by that person).
Finally, practical problems also result from the epistemic problem. Simply put,
nudges cannot do what they promise – to improve people’s choices in their own
interests – if policymakers do not know these interests and impose others to substi-
tute for them. Nudges are certainly effective in modifying behaviour, but there is no
way for policymakers to know if those behavioural modifications are in people’s
true interests. For instance, Sunstein and Thaler cite the success rate of their Save
More Tomorrow plan that, by default, signs new employees up for retirement plans
24 M.D. White

and then automatically increases their contributions over time.7 But they are mea-
suring success by the effectiveness of the nudge in changing behaviour, not to what
extent those new employees’ interests are advanced by the higher savings rate –
unless we assume that savings is their preeminent interest. Other possible practical
effects of nudges include long-term adverse influences on decision-making itself:
instead of acting to improve the process of choice, nudges simply attempt to elicit
the ‘correct’ choices, and the way in which they do this, by subverting rational
deliberative processes, may impair these processes in the long run.

3.3 Interests

As seen above, the ethical and practical problems stem from the epistemic problem:
the fact that policymakers do not know the true interests of the people they are
nudging, because actual interests are complex, multifaceted, and subjective.
While behavioural economists have made tremendously valuable contributions
to understanding human behaviour in all its imperfection, it clings to mainstream
economists’ myopic focus on a simplistic form of utility. There, utility is nothing
but an index of preference satisfaction, in which preferences themselves are merely
a formal ranking of options with no necessary psychological basis. In other words,
a preference for A over B need not be based on desire, happiness, or need, but
simply on the assumption that, all other things the same, one would choose A or B,
for whatever reason. This makes choice theory substantively empty and therefore
adaptable to a wide range of psychological motivations, from pure self-interest to
global altruism, or from interdependent utility with a loved one to profound hatred
for humanity. However, in practice economists implicitly assume that preferences
are based on a single criterion, so the utility resulting from assessing preference
satisfaction resembles a more substantive utility, such as happiness or well-being;
a proxy thereof, such as wealth; or a component thereof, such as health.
Behavioural economists accept this basis for economic choice in preferences
oriented toward a single goal and then question the way we pursue these preferences
given limitations on calculative rationality and willpower. Methodologically, this is
understandable: it is much easier to complicate just one aspect of a model while
leaving the rest unchanged. And certainly, valuable insights have been gained
from this approach. However, when behavioural insights regarding the process of
decision-making are incorporated into policymaking, the limited view of motivation
brought with it becomes much more problematic, because it lends a similarly
myopic focus to policymaking in turn.
Human beings, however, do not have a single goal or end that they act to further
(or ‘maximize’), but rather a range of interests that are complex, multifaceted, and
ultimately subjective. I define interests as “everything that matters to a person,
everything that is of concern to him or her, and everything that motivates his or her

Thaler and Sunstein 2008, ch. 6.
3 The Crucial Importance of Interests in Libertarian Paternalism 25

decisions, choices, and actions – in other words, everything he or she is interested

in.”8 This conception of interests transcends simpler ones that focus on one goal,
especially as imagined in typical economic models.
Let’s begin with the standard economic assumption of narrow self-interest,
which most economists realize is merely a heuristic assumption to simplify model-
ling. While this is adequate to describe some decisions, such as which dressing to
have on one’s salad, our self-interest, our own happiness or well-being, does not
exhaust all of the reasons and concerns that determine our decisions in broader
contexts. Most people also care about the happiness or well-being of others, includ-
ing family and friends, perhaps neighbours and co-workers, and maybe even disad-
vantaged persons in our communities or around the world. Many sacrifice significant
portions of their own happiness or well-being to benefit others; terminologically,
this altruistic interest could be rolled into one’s self-interest or happiness, but an
important distinction would be lost.
But the well-being of oneself and others still does not exhaust everything that
people care about and that motivates their choices. Personal principles and societal
ideals, both important qualitative influences on decision-making, must also be
included. Principles such as honesty and fairness constrain choices that further
well-being; in many cases, people will not lie or cheat even when it would be to
their advantage in material or financial terms. In fact, experiments have shown this
time and again, and (ironically) is a hallmark of behavioural economics.9 People are
also concerned with more general societal ideals such as justice, equality, and fair-
ness, for which they sacrifice not only their time and money but also their freedom
and lives. Again, some would collapse all of these disparate and distinct interests
into the term ‘self-interest’, but this does an inestimable disservice to understanding
and appreciating the diversity and range of human motivations and concerns.
Not only are people’s interests multifaceted, including own well-being, others’
well-being, principles, and ideals, but these are also combined in complex ways,
ever changing based on circumstance and context. Some interests take precedence
at some times and others at other times, but the combination is never predetermined
or predictable. Consider the last time you chose where to go for dinner: your pri-
mary concern may have been how good the food would taste, how healthy it would
be, the cost of the meal, the ambience of the restaurant, or how crowded it would be.
Whichever interest took precedence, the others were still relevant, and perhaps a
different one would take precedence as circumstances changed – for instance, if you
heard that an old friend was joining you and you wanted a quiet place to catch up.
Most of this is automatic or unconscious, of course; agents are never perfectly
consciously aware of all of their interests and how they are combined at any given
time, nor could they report them accurately if asked. But this not deny that behaviour
is motivated by these myriad interests, acting together through the filter of intuition
and judgment.

White 2014, p. 93.
See, for instance, Fehr and Schmidt 2006.
26 M.D. White

To reiterate, the complexity of interests stands in stark contrast to the typical

economic assumption, usually made in both mainstream and behavioural econom-
ics, of a single interest such as happiness, wealth, or health – the same assumption
that grounds nudge policies, leading to the epistemic problem at the root of the
enterprise. Again, it must be acknowledged that people themselves are rarely con-
sciously aware of their own interests or what will advance them. The psychological
literature on people’s consistently flawed estimates of their future well-being is
voluminous.10 But this does not imply that policymakers know them any better, or
that they are justified in imposing their own idea of interests on them. Rather, this
should reinforce the position that individuals should be left to make decisions in
their interests, despite any problems in instrumental reasoning towards them.

3.4 Interests and Paternalism

Defenders of nudges as well as more direct forms of paternalism maintain that they
are not questioning people’s interests but are instead merely trying to help them
make better decisions in the interests they have, referencing the original distinction
between soft paternalism, that respects interests, and hard paternalism, that imposes
others.11 In recent work on paternalism, both Cass Sunstein (advocating libertarian
paternalism) and Sarah Conly (arguing for coercive paternalism) use the distinction
between means and ends to support their soft paternalism, and also argue against
perfectionist or objective theories of well-being that would endorse imposing exter-
nal interests on people when designing paternalistic interventions (in the spirit of
hard paternalism). I will discuss these two positions in turn, arguing in response that
the epistemic problem obscures the distinction between means and ends to the point
of irrelevance, and that it also implies that any paternalistic interventions by neces-
sity impose interests based on perfectionist or objective conceptions of interests.
(Although Conly advocates in favour of coercive paternalism over nudges, her
arguments regarding means, ends, and interests parallel those of Sunstein, so I will
discuss them together).

3.4.1 Means Paternalism and Ends Paternalism

To clarify the non-judgmental nature of their brands of paternalism, both Sunstein

and Conly distinguish between means paternalism and ends paternalism, stressing
that they are advocating for the former but not the latter. They both begin their

See Loewenstein and Schkade 1999; Gilbert and Wilson 2000; Haybron 2008, ch. 10.
Feinberg 1986, pp. 12–16. The terms soft and hard have also been used to distinguish between
libertarian and coercive paternalism, while either of these could, in theory, be soft or hard accord-
ing to the traditional usage. On this, on this point, see also Conly 2013, pp. 5–6.
3 The Crucial Importance of Interests in Libertarian Paternalism 27

recent books by recounting the findings of behavioural researchers regarding cogni-

tive biases and dysfunctions that often cause people’s decisions to deviate from
their interests, and use this to justify paternalistic intervention into the process of
decision-making.12 In this sense, paternalists claim to be merely helping people to
further their ends by making better choices in pursuit of them, as opposed to steer-
ing their choices toward different ends altogether.
Both Sunstein and Conly also question John Stuart Mill’s famous passage from
On Liberty in which he asserts the individual’s self-knowledge (which Sunstein
calls Mill’s Epistemic Argument):
But neither one person, nor any number of persons, is warranted in saying to another human
creature of ripe years, that he shall not do with his life for his own benefit what he chooses
to do with it. … [W]ith respect to his own feelings and circumstances, the most ordinary
man or woman has means of knowledge immeasurably surpassing those that can be pos-
sessed by any one else. The interference of society to overrule his judgment and purposes
in what only regards himself, must be grounded on general presumptions; which may be
altogether wrong, and even if right, are as likely as not to be misapplied to individual cases,
by persons no better acquainted with the circumstances of such cases than those are who
look at them merely from without.13

Both Sunstein and Conly challenge Mill based on the aforementioned cognitive
flaws identified by behavioural researchers: Sunstein writes that “behavioural find-
ings are creating serious problems for the Epistemic Argument, because they show
that people make a lot of mistakes, some of which can prove extremely damaging,”
and Conly writes that “it is too simplistic” in that it ignores the distinction between
means and ends, as well as people’s inability to reliably choose the best means to
their end.14 Both scholars, however, interpret Mill’s statements in terms of choosing
the means and are too quick to dismiss the deeper issue of individuals being in the
best position to know his or her own ends and interests. It is unquestionable that
people make bad choices, but absent information regarding their interests, no one
can know which choices are bad except the people making them.
Put another way, emphasizing the distinction between means and ends in this
context begs the question regarding the nature and content of interests themselves.
If policymakers had information regarding people’s true interests, they may be in a
position to offer assistance, whether in the form of nudges or more coercive means,
to people to advance those interests. People who reveal that they want help manag-
ing their money can go to financial advisors, and those who want to lose weight can
join groups such as Weight Watchers; politics aside, the government could offer
such services to people as well. But these assume revelation of interests through
choice, and without such revelation, policymakers have no way to know people’s
true interests or whether choices made by individuals are effective means to promote
their ends. Sunstein writes that the “government might well accept people’s ends but
conclude that their choices will not promote those ends,” but the government does

Sunstein 2014, ch. 1; Conly 2013, pp. 20–23.
Mill 1859, ch. 4.
Sunstein 2014, p. 8; Conly 2013, p. 52.
28 M.D. White

not have enough information regarding people’s ends to accept them (setting aside
the issue of whether the government is in a place to ‘accept’ people’s ends anyway).15
Defenders of paternalism (in both libertarian and coercive forms) are correct to
emphasize the decision-making failures pointed out by psychologists and behav-
ioural economists that cause individuals to make choices that fail to advance their
interests. But the general possibility, or even likelihood, that people make mistakes
in decision-making does not imply that they are making mistakes in any given case,
much less that an outside observer can determine whether a given choice was mis-
taken, because that determination requires knowledge of the person’s interests
toward which he or she made the decision. Given that interests are complex, multi-
faceted, and subjective, policymakers have no way to know them, and therefore
they cannot know whether or not a person’s chosen means furthers his or her true
interests. By interfering with a person’s choice of means, policymakers have no
way to know if they are helping persons further their true interests or steering them
toward interests that are presumed and imposed by policymakers when designing
the intervention. People may make bad choices all the time, of course, but the only
ones who know they are mistakes are the people making them because only they
know why they made them.
In practice, then, means paternalism is indistinguishable from ends paternalism
because the distinction itself assumes knowledge of people’s ends, knowledge that
is unavailable to the policymaker (and perhaps to people themselves). Sunstein does
acknowledge this possibility, stating outright that even relatively benign nudges
such as fuel economy labels not only influence the choice of means but may nudge
interests themselves away from other aspects of an automobile purchase, “insofar as
it isolates fuel economy, rather than other imaginable features of cars, for compul-
sory display.”16 But while he recognizes the possibility of ends paternalism, he mini-
mizes its importance: “there may be a form of ends paternalism, but it is likely to be
of a very modest kind, perhaps so modest that we do not to worry much.”17 Given
the impossibility of distinguishing between means and ends absent knowledge of
ends, however, the degree of ends paternalism may be much higher (and not modest
at all). Sunstein approaches this position when he discusses how to define ends: if
ends are very general, such as health or happiness, then most interventions will
affect means only, but if ends are defined very narrowly, then “there may be no such
thing as means paternalism.”18 I would argue that the complex and multifaceted
nature of interests implies a high degree of specificity to the imposed ones, which is
further enhanced by the subjectivity of true interests, and if proper deference were
granted to them by policymakers, nearly all paternalistic interventions would prop-
erly be regarded as affecting ends and not merely means.

Sunstein 2014, p. 61.
Sunstein 2014, pp. 66–67; see also pp. 91–92.
Sunstein 2014, p. 69.
Sunstein 2014, p. 70.
3 The Crucial Importance of Interests in Libertarian Paternalism 29

The complex, multifaceted, and subjective nature of interests also implies het-
erogeneity of interests among individuals, which is widely acknowledged yet not
appreciated sufficiently. For instance, Sunstein writes that “people are highly
diverse in terms of their tastes, their values, and their situations. One size may not
fit all.”19 He also approximates the definition of interests given in this paper when
he writes that people have concerns reflecting altruism or social responsibility, as
well as different conceptions of their own well-being or happiness, on which “they
strike their own balance; different people will choose differently.”20 However, he
uses these inclusive and pluralistic conceptions of well-being to defend libertarian
paternalism, which offers more choice than do coercive interventions, rather than to
question his paternalist enterprise altogether, which in his examples relies on simple
single interests such as health or savings.
Writing for a philosophical rather than popular audience, Conly is more elabo-
rate when discussing the plurality of possible interests and its implications for pater-
nalistic interventions, which “may be downright harmful to some individuals – it
may prevent them from achieving well-thought-out ends. Obviously, different peo-
ple have some different goals and needs.”21 She responds to that problem by assert-
ing that many ends, such as health and prosperity, are common to most people – all
legislation is general by nature, after all, with unavoidably disparate impacts on
various individuals – and that, as Sunstein also maintains, cost-benefit analysis must
be applied to decide on the effectiveness of any particular intervention.22 But a one-
size-fits-all approach of either type of paternalism then defeats the purpose of
advancing people’s own ends by substituting general, simplistic ends chosen by the
policymaker, and, in its aggregation of interests, resembles welfarism more than
Conly provides a fine example of the multiple interests that may lie behind a
choice when she mentions the purchase of lottery tickets, which she does twice in
her book. The first time she characterizes them as a counterexample to learning
from mistakes, imprudent choices that persist over time. But this judgment is based
on the questionable but universally implicit assumption among academics that most
people play the lottery as part of a considered long-term financial investment plan:
“Buying a lottery ticket every week doesn’t teach people that it isn’t a good use of

Sunstein 2014, p. 17; see also pp. 96–100.
Sunstein 2014, p. 74.
Conly 2013, p. 64.
Conly 2013, pp. 63–66; Sunstein 2014, pp. 18–19.
See Korobken 2009. As noted above, defenders of nudge and stronger forms of paternalism often
argue that the question of justification is an empirical one depending on the cost and benefit of
specific interventions. But this takes us back yet again to the impossibility of knowing individuals’
interests, which are presumably what are to be advanced, and measurement of which would repre-
sent the benefits of paternalistic interventions. Without knowledge of actual interests, however,
advocates of paternalism are left only with the interests on which the interventions are based (or
the behaviour modified in those interests), which places a heavy thumb on the scale in favour of
30 M.D. White

their money.”24 The second time, she acknowledges that there may be other reasons
to play the lottery, such as enjoyment, in her defence of means paternalism over
ends paternalism: “If my subjective end is happiness, and I think playing the lottery
will promote that, not because the suspense gives me some evanescent pleasure, but
because I really think I have a reasonable chance of winning, I am mistaken about
my means.”25 Her argument seems to be that if people play the lottery for the thrill
(quite reasonably priced at a dollar), they may be making a reasonable choice, but
if they are hoping for riches, they are not. This is valid, but while she likely knows
her own subjective ends when assessing the lottery, she does not know those of
anyone else. If policymakers were to intervene paternalistically to limit lottery play
in the interest of people’s finances, they would be more likely be influencing lottery
players’ enjoyment – which would have an impact on people’s actual ends while
trying to improve the choice of means toward others.
Conly is careful to make a distinction between transitory desires and preferences
and more stable, long-term goals and desires, and argues that paternalism should act
to block the former when they do not correspond to the latter. But without knowl-
edge of what people’s true long-term interests are, this distinction is of no use to
crafting policy that respects those interests. It is impossible for a policymaker to
know when choices over short-term means are trivial or an integral part of advanc-
ing deeper interests. When suggesting areas of life in which paternalism may not be
effective, Conly mentions choices regarding marriage or career, in which even the
smallest choices are deeply personal and linked to the larger enterprise itself.
Because the means and ends in such situations are more difficult to identify, the
prudent option is to leave these choices to the decision-maker, despite the ever-
present chance of bad choices. This is a valuable insight, but one I would argue
applies not only to choices about marriage and career but all choices in life; to sepa-
rate these areas out requires a judgment call about the relative importance of differ-
ent choices, which is itself an external opinion about interests. Conly writes that:
requiring that people save more than they now do, that they don’t amass huge debts through
avaricious credit schemes, that they don’t buy cigarettes, that they don’t eat things with
likely lethal effects, doesn’t seem to interfere with the basic life choices rights are intended
to protect.26

These may not seem like important choices to an outside observer, who cannot
judge what a person regards as an important choice to his or her interests. A person
may have any number of reasons to save, borrow, smoke, and eat to an extent or
degree that another person may think imprudent based on his or her judgment on
what is best. But this judgment cannot be made without imposing interests on the
person making the choices, which modern paternalists deny that they do.
Modern paternalists’ focus on long-term interests as opposed to short-term desires
recalls the concept of ideal or rational preferences favoured by many economists

Conly 2013, p. 27; on this point, see also LeGrand and New 2015, pp. 147–150.
Conly 2013, p. 43.
Conly 2013, p. 65.
3 The Crucial Importance of Interests in Libertarian Paternalism 31

and philosophers who study personal welfare and well-being.27 Because of the vari-
ous cognitive biases and dysfunctions as well as informational problems to which
we are subject, some economists and philosophers argue that agents’ preferences
may not reflect their true interests and should be replaced in the policymaker’s tool-
box by ‘rational preferences’, the ones “he would have if he had all the relevant
factual information, always reasoned with the greatest possible care, and were in a
state of mind most conducive to rational choice.”28 Sunstein and Thaler invoke this
concept when they argue that the bad decisions people make may be “decisions that
they would change if they had complete information, unlimited cognitive abilities,
and no lack of self-control.”29 Conly echoes this line of thinking when she describes
the point of paternalism as “to help the person what achieve what in the long run he
wants, and what he would want now if he were not a flawed thinker.”30
But the construction of these rational preferences also, by necessity, involves
external judgment, being essentially counterfactual and hypothetical, and bearing
no necessary relation to a person’s actual interests. As Robert Sugden asks:
How, without making normative judgments, do we determine what counts as complete
information, unlimited cognition, or complete willpower? Even if we can specify what it
would mean to have these supernatural powers, how do we discover how some ordinary
human being would act if he were somehow to acquire them?31

Such idealized preferences may, in theory, counter faults with actual preferences,
but have little tie to actual interests, given their judgment-laden nature, and once
again defeat the stated purpose of paternalists to advance the true ends of people
affected by interventions.

3.4.2 Perfectionist or Objective Theories of Well-Being

The practical irrelevance of the distinction between means and ends, given the lack
of knowledge on the part of policymakers regarding people’s true interests, is symp-
tomatic of the problem with modern paternalism’s claim to respect those interests
and merely improve people’s decision-making toward them. This theme is empha-
sized throughout recent work, from Sunstein and Thaler’s choice of the term ‘liber-
tarian paternalism’ to Conly’s description of her coercive paternalism as
‘benevolent’, both corresponding to soft paternalism in the original sense of the
term. In their recent work, both Sunstein and Conly flatly reject perfectionist or
objective theories of welfare. Sunstein is clear that, “though paternalists might have
any number of views about what would make people’s lives go well,” he is “inter-
ested in defending paternalists who respect choosers’ own views about their ends,

On the theory of rational preferences and their relationship to well-being, see Griffin 1986, chs.
1 and 2.
Harsanyi 1982, p. 55.
Sunstein and Thaler 2001, p. 1162.
Conly 2013, p. 36.
Sugden 2008, p. 232.
32 M.D. White

and who seek to increase the likelihood that their decisions will promote those
ends.”32 Conly writes that what she advocates “is not a paternalism about ultimate
ends; that is, I do not argue that there are objectively good ends, or objectively ratio-
nal ends, or ends objectively valuable in any way, which everyone should be made
to pursue.”33 Instead, it “promotes the satisfaction of people’s long-term desires,”
recalling the discussion of rational preferences above.34
This position is untenable, however, given the nature of interests: if interests are
subjective and therefore externally unknowable, then paternalists cannot intention-
ally and effectively advance them. Either paternalists do claim to have knowledge
of subjective interests, or they are implicitly imposing their own idea of interests,
which is more consistent with an objective theory of welfare. I argue that while they
argue the former, they are in fact engaging in the latter.
Both Sunstein (with and without Thaler) and Conly, in defence of their respec-
tive brands of paternalism, assume a number of central interests common to all. In
fact, Blumenthal-Barby notes that, in most of examples of nudges in the literature,
“there is a sense of an underlying perfectionist standard of the good, namely health
and wealth, to which the various nudges direct the masses.”35 Sunstein and Thaler
famously give examples of promoting health by rearranging items in a cafeteria and
savings by enrolling new employees in retirement programs by default, both inter-
ests they implicitly take to be universal and important. In his most recent book,
Sunstein writes that policymakers can improve people’s lives by implementing
nudges “that give health, wealth, and well-being the benefit of the doubt.”36 Conly
writes, “I would argue that we (most of us) have a stable desire to be healthy and
prosperous, and furthermore have a relatively clear idea of what constitutes a satis-
factory degree of health and prosperity.”37 Both health and prosperity are vague and
common interests, to be sure, but neither of them corresponds precisely to how
individuals understand them or balance them with other interests. I agree that each
of us has a relatively clear idea of what health and wealth mean to us, but that is an
individualized conception that is not available to the policymaker.38

Sunstein 2014, p. 75.
Conly 2013, p. 43.
Conly 2013, p. 50; see pp. 102–112 for a thorough discussion of perfectionism that never con-
fronts the issue of the substitution of interests. See also Cholbi 2013, who defends a Kantian
paternalism predicated explicitly on the interests of the individuals subject to it: “what triggers
justifiable paternalism is not the nature of the good person pursues but the exogenous imprudence
with which she pursues it” (p. 124).
Blumenthal-Barby 2013, p. 180.
Sunstein 2014, p. 23.
Conly 2013, p. 124.
Blumenthal-Barby recognizes the epistemic problem in principle, but agrees with Conly regard-
ing its practical irrelevance: “while differences in preferences exist to be sure, there is much less
variety than one might think, and much more data about what really does (and does not) make
better off” (2013, p. 187). But this assumes that people’s interests consist solely of (or can be
described completely) by well-being, and that defining and measuring well-being is not subject to
same epistemic problems described above; on both, see White 2014.
3 The Crucial Importance of Interests in Libertarian Paternalism 33

One of the examples of paternalism Sunstein discusses in his 2014 book is fuel
efficiency standards designed to save consumers money on gas.39 He considers the
possibility that people do not want to buy fuel-efficient cars, and that fuel-economy
standards may be a form of paternalism that elevates one value above all others, but
then concludes that consumers simply “may not give sufficient attention to the long-
term or aggregate costs.”40 He cites empirical research that supports his point that
consumers do not purchase energy-efficient products despite the cost savings, but
this interpretation again assumes that cost savings is the only or the most important
factor when buying these products. Sunstein is careful to make qualifications, such
as when he writes “if, for example, consumers want to make a sensible trade-off
between up-front costs and long-term fuel costs.”41 I argue that this is a big ‘if,’ and
itself contains assumptions regarding what is a sensible trade-off between current
and future costs. Furthermore, cost is certainly an important factor to many people,
but there is no reason to believe that it overwhelms other considerations such as
aesthetics or additional features, which are often ignored when assuming what
people’s true interests are.
In her lengthy and detailed discussion of paternalistic interventions in health
care, Conly assesses several paternalist interventions, such as portion size control
and cigarette bans, along four criteria, including whether they advance people’s
interests. In each case, however, she defines the interests for people. For instance, in
the case of New York City’s trans-fat ban, she cites the primary interest advanced
as “staying alive, and being able to function normally while we are alive,” as “some-
thing we want.”42 This is certainly a common interest, but it is both trivial and vague:
of course, most all people want to stay alive and be able to function, but this does
not mean they avoid all risks to that interest, and they balance their interest in con-
tinued life with their interest in truly living. These objections aside, longevity may
be a very reasonable, prudent, and common interest, as is health in general (cited
elsewhere), but it is not the only interest of any person (even self-professed ‘health
nuts’), much less an entire population. It is an interest defined and assessed by
someone other than the people affected, and in this way resembles an objective
theory of good as described as Derek Parfit: “According to this theory, certain
things are good or bad for people, whether or not these people would want to have
the good things, or to avoid the bad things.”43
Paternalists may argue that people should take an interest in their health based
on value judgments backed by scientific evidence; as Colin Camerer and his
colleagues point out, “health and food regulations are heavily informed by scientific
understanding… and by a widespread belief among professionals that average
folks require information, prodding, and often regulation to improve their health

Sunstein 2014, pp. 42–44.
Sunstein 2014, p. 43.
Sunstein 2014, p. 62.
Conly 2013, p. 153.
Parfit 1984, p. 499.
34 M.D. White

and diet.”44 But this is to take an objective view of interests: the fact is that some
people do not take a strong interest in their health, despite Conly’s characterization
of it as a ‘dearly held interest’, so scientific evidence regarding diet, exercise, and
sleep are of little import to advancing their true (subjective) interests.45 Conly argues
that health is at least an integral instrumental interest, necessary for the promotion
of other (intrinsic) interests. Once again, this may be a reasonable assumption but is
hardly necessarily true, and many people knowingly and willingly sacrifice aspects
of their own health, as medical professionals would characterize it, to pursue other
ends. Furthermore, among people who do take an active interest in their health, foci
differ, with some emphasizing weight loss, others strength, and yet others sugar
intake. ‘Health’ is much too vague a term to impose on people as an interest in
which paternalists will influence decisions.46
Conly’s analysis is more elaborate, however, and does not restrict itself to this
single interest. For instance, in considering bans on using food stamps for soda,
she lists other interests, such as “being found conventionally attractive, and social
acceptance generally,” which lessens the “psychological burden of being unattract-
ive by conventional standards”; in assessing portion size regulation, she lists as an
interest avoiding the stigma and (in the case of women) diminished job opportuni-
ties stemming from obesity.47 She also considers interests that may be negatively
affected at the same time that health (and attractiveness by conventional standards)
is promoted, such as “loss of enjoyment and loss of status,” but while she takes these
very seriously, at the same time she frames these interests as ‘costs’ rather than par-
allel concerns of equal status to health in the pantheon of subjective interests. While
she admirably goes to great length to list many possible interests that might speak
against the positive effect of an intervention, this also reveals the multifaceted nature
of interests and neglects the complex ways in which they combine and interact –
most important, in ways that are subjective and unknowable to the outside observer.
With regard to smoking, Conly asks (and answers), “Do smokers get sufficient ben-
efit from smoking to make this a rational choice, given their ends? It seems likely
that it is not.”48 While this conclusion may be amenable to many people, neverthe-
less it involves an arbitrary external judgment regarding not only the interests of
smokers but also how they balance and combine them. As much care as policymak-
ers might take in trying to specify every interest that might be affected, they can
never be certain that they have identified all of them or considered them in the same
proportions or priorities as assigned by people themselves. Because of this lack of
information, policymakers who choose to influence decisions must impose interests
that, no matter how reasonable and prudent, are nevertheless not the subjective
interests of people themselves, and instead correspond to objective theories of
well-being or interests.

Camerer et al. 2003, p. 1222.
Conly 2013, p. 169.
See White 2016.
Conly 2013, p. 156 and p. 164.
Conly 2013, p. 170.
3 The Crucial Importance of Interests in Libertarian Paternalism 35

In some cases, such as marriage, Conly explicitly admits that “we just don’t
know enough” to intervene paternalistically, because (as described earlier) the
choices themselves are such an intrinsic part of the enterprise itself that the distinc-
tion between means and ends is blurred. But she maintains that less significant deci-
sions are different, and she uses the example of selecting food as an example. “I
don’t particularly value making nutritional food choices per se; I want to enjoy
eating and I want to have food that doesn’t do me harm, and someone else can do
the legwork as to how best to do that. Romance, however, is obviously different.”49
I disagree that this difference is obvious for everyone: her attitudes towards her food
choices are not necessarily anyone else’s, and may be closer to how she regards
marriage (which, to some, may be more instrumental). As Claire Hill writes, mod-
ern paternalists
sometimes speak as though they have access to the knowledge of what people really want
apart from what they choose. This position is ultimately untenable… As convenient and
tempting as it may be to extrapolate from our own introspection that others want what we
do, or should, want, we simply have no access to others’ beliefs and desires.50

My argument is that in most all cases of choice, policymakers simply don’t know
enough about interests, including the relative importance of different kinds of inter-
ests, to justify interfering with decision-making. Given this lack of information
regarding people’s true interests, policymakers are very unlikely to further those
interests by promoting others of their own choosing, regardless of how prudent and
reasonable they may seem. Instead, policymakers would best promote people’s true
interests by giving them the benefit of the doubt regarding their own interests and
how they choose to pursue them.51
To summarize, any interests that are not embodied specifically in individuals’
choices, but rather are chosen by policymakers, are presumed interests that reflect,
by definition, an objective theory of interests. As Dan Brock writes:
Paternalistic interference involves the claim of one person to know better what is good for
another person than that other person him- or herself does. It involves the substitution by
the paternalistic interferer of his or her conception of what is good for another for that
other’s own conception of his or her good. If this involves a claim to know the objectively
correct conception of another’s good—what ultimate values and aims define another
competent individual’s good, independent of whether that other accepts them—then it is
ethically problematic.52

Conly 2013, p. 184; see pp. pp. 183–186 in general.
Hill 2007, p. 448.
Note that I have not mentioned the effect of paternalism on autonomy understood as an intrinsic
value; I made my argument in this chapter on the purely welfarist grounds favored by Sunstein and
Conly. For a defense of the importance of an intrinsically valuable sense of autonomy—that is,
autonomy not understood simply as a component of well-being—see White 2013, ch. 7, and refer-
ences therein (especially on p. 170 n. 11). (Autonomy is discussed in both Sunstein 2014, ch. 7,
and Conly 2013, ch. 1; both are critical of autonomy as an intrinsic value but incorporate it in an
instrumental sense in their proposed cost-benefit analysis).
Brock 1988, p. 559.
36 M.D. White

Advocates of nudge or paternalism in general cannot maintain that they are respecting
subjective interests while imposing external, objective ones. It would be much
simpler, and would avoid many conceptual inconsistencies, were paternalists to
embrace explicitly the objective theories of interests that they implicitly use, “to admit
that choice environments are designed to direct people towards generalizable good”
and not each person’s true interests.53

3.5 Conclusion

The distinction between means and ends paternalism and the denial of objective
theories of well-being ground the approach of modest paternalists, as exemplified
by Cass Sunstein and Sarah Conly in key quotations from their recent work. Sunstein
writes that the reasoning behind his version of paternalism is that “government does
not believe that people’s choices will promote their welfare, and it is taking steps to
influence or alter choices for their own good.”54 Likewise, Conly writes that
The reason for intervention is that we don’t trust you to choose rightly. We are taking away
freedom of choice in these cases because we don’t think people will choose well them-
selves. We don’t think preserving your autonomy, your freedom to act based on your own
decision, is worth the costs, in part because your decision making is done so badly that your
freedom is used very poorly.55

Although the tone of these two statements is very different, they share the same
element of judgment: ‘the government believes’ people’s choices promote their
interests, and ‘we don’t think’ people can be trusted to make these choice for them-
selves. But this judgment presumes knowledge of people’s true ends and interests
that is unavailable to the paternalist.
Of course, people do judge some of their own decisions to be bad, and they often
take steps to improve these choices. At the end of the day, only a person knows
whether his or her choice in a given situation furthers the interests he or she has at
that time. As Mill wrote, in the context of voluntary risk-taking, “no one but the
person himself can judge of the sufficiency of the motive which may prompt him to
incur the risk.”56 Without this knowledge, the prudence of chosen means cannot be
assessed, and the distinction between means and ends is rendered meaningless for
the purposes of designing paternalistic interventions.
Furthermore, the complex, multifaceted, and subjective nature of interests poses
serious problems for paternalists who claim to be promoting people’s true interests.
Conly writes that “interference is justified on paternalistic grounds only when it
reflects individuals’ actual values, not the values we might like them to have.”57

Blumenthal-Barby 2013, p. 181.
Sunstein 2014, p. 54.
Conly 2013, pp. 179–180.
Mill 1859, ch 5.
Conly 2013, p. 150.
3 The Crucial Importance of Interests in Libertarian Paternalism 37

Paternalists need not be imposing values that they ‘might like’ people to have, but
nonetheless they are necessarily imposing values as implied by their lack of infor-
mation, and this substitution of interests introduces perfectionist or objective values
into the practice of paternalism, whether libertarian or otherwise. Paternalists could
avoid this conundrum by embracing the objective theories of welfare that inform
their theory and practice, although it would diminish the attractive claim of modern
paternalism to be helping people further their own interests. As I have argued, this
claim is false; why it is so readily accepted by so many is a question worth


Angner, Erik. 2012. A course in behavioral economics. New York: Palgrave Macmillan.
Blumenthal-Barby, J.S. 2013. Choice architecture: A mechanism for improving decisions while
preserving liberty? In Paternalism: Theory and practice, ed. Christian Coons and Michael
Weber, 178–196. Oxford: Oxford University Press.
Brock, Dan W. 1988. Paternalism and autonomy. Ethics 98: 550–565.
Camerer, Colin, Samuel Issacharoff, George Loewenstein, Ted O’Donoghue, and Matthew Rabin.
2003. Regulation for conservatives: Behavioral economics and the case for ‘Asymmetric
Paternalism’. University of Pennsylvania Law Review 151: 1211–1254.
Cholbi, Michael. 2013. Kantian paternalism and suicide intervention. In Paternalism: Theory and
practice, ed. Christian Coons and Michael Weber, 115–133. Oxford: Oxford University Press.
Conly, Sarah. 2013. Against autonomy: Justifying coercive paternalism. Cambridge: Cambridge
University Press.
Fehr, Ernst, and Klaus M. Schmidt. 2006. The economics of fairness, reciprocity and altruism—
Experimental evidence and new theories. In Handbook of the economics of giving, altruism
and reciprocity, vol. I, ed. Serge-Christophe Kolm and Jean Mercier Ythier, 615–691.
Dordrecht: Elsevier.
Feinberg, Joel. 1986. Harm to self. Oxford: Oxford University Press.
Gilbert, Daniel T., and Timothy D. Wilson. 2000. Miswanting: Some problems in the forecasting
of future affection states. In Feeling and thinking: The role of affect in social cognition, ed.
Joseph P. Forgas, 178–197. Cambridge: Cambridge University Press.
Griffin, James. 1986. Well-being: Its meaning, measurement, and moral importance. Oxford:
Oxford University Press.
Harsanyi, John. 1982. Morality and the theory of rational behavior. In Utilitarianism and beyond,
ed. Amartya Sen and Bernard Williams, 39–62. Cambridge: Cambridge University Press.
Haybron, Daniel. 2008. The pursuit of unhappiness: The elusive psychology of well-being. Oxford:
Oxford University Press.
Hill, Claire A. 2007. Anti-anti-anti-paternalism. NYU Journal of Law & Liberty 2: 444–454.
Jolls, Christine, Cass Sunstein, and Richard Thaler. 1998. A behavioral approach to law and eco-
nomics. Stanford Law Review 50: 1471–1550.
Kahneman, Daniel. 2011. Thinking, fast and slow. New York: Farrar, Straus and Giroux.
Kahneman, Daniel, Amos Tversky, and Paul Slovic (eds.). 1982. Judgment under uncertainty:
Heuristics and biases. Cambridge: Cambridge University Press.
Korobkin, Russell. 2009. Libertarian welfarism. California Law Review 97: 1651–1685.
Korobkin, Russell B., and Thomas S. Ulen. 2000. Law and behavioral science: Removing the
rationality assumption from law and economics. California Law Review 88: 1051–1144.
Le Grand, Julian, and Bill New. 2015. Government paternalism: Nanny state or helpful friend?
Princeton: Princeton University Press.
38 M.D. White

Loewenstein, George, and David Schkade. 1999. Wouldn’t it be nice? Predicting future feelings.
In Well-being: The foundations of hedonic psychology, ed. Daniel Kahneman, Ed Diener, and
Norbert Schwarz, 85–105. New York: Russell Sage Foundation.
Mill, John Stuart. 1859. On liberty. At
Parfit, Derek. 1984. Reasons and persons. Oxford: Oxford University Press.
Parisi, Francesco, and Vernon L. Smith (eds.). 2005. The law and economics of irrational behav-
ior. Stanford: Stanford University Press.
Rebonato, Riccardo. 2012. Taking liberties: A critical examination of libertarian paternalism.
New York: Palgrave Macmillan.
Saint-Paul, Gilles. 2011. The tyranny of utility: Behavioral social science and the rise of paternal-
ism. Princeton: Princeton University.
Simon, Herbert. 1955. A behavioral model of rational choice. Quarterly Journal of Economics 69:
Sugden, Robert. 2008. Why incoherent preferences do not justify paternalism. Constitutional
Political Economy 19: 226–248.
Sunstein, Cass R. (ed.). 2000. Behavioral law & economics. Cambridge: Cambridge University
Sunstein, Cass R. 2014. Why nudge? The politics of libertarian paternalism. New Haven: Yale
University Press.
Sunstein, Cass R., and Richard H. Thaler. 2001. Libertarian paternalism is not an oxymoron.
University of Chicago Law Review 70: 1159–1202.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge: Improving decisions about health, wealth,
and happiness. New Haven: Yale University Press.
White, Mark D. 2013. The manipulation of choice: Ethics and libertarian paternalism. New York:
Palgrave Macmillan.
White, Mark D. 2014. The illusion of well-being: Economic policymaking based on respect and
responsiveness. New York: Palgrave Macmillan.
White, Mark D. 2016. Bad medicine: Does the unique nature of health care decisions justify
nudges? In Nudging health: Health law and behavioral economics, ed. I. Glenn Cohen, Holly
F. Lynch, and Christopher T. Robertson. Baltimore: Johns Hopkins University Press (in press).
Chapter 4
Condorcet’s Jury Theorem as a Rational
Justification of Soft Paternalistic Consumer
A Philosophical Note

Malte Frederic Dold

Abstract  The aim of this note is to revisit the meaningfulness of the Condorcet
Jury Theorem (CJT) and apply it to the recent debate on liberal paternalism and
consumer protection. The CJT consists of two parts, (a) stating that a jury of experts
is always more competent than a single expert given a certain level of competence,
and (b) asserting that for large juries, the collective competence approaches infalli-
bility. This note argues that these insights suggest the application of a Condorcet
jury voting procedure in case of nudging boundedly rational consumers. The note
proposes a simple calculus for finding an optimal jury size and advocates consum-
ers’ meta-preferences as the jury’s evaluative dimension for designing soft paternal-
istic policies.

4.1  Introduction

The individual at a point in time is assumed to be both a farsighted planner and a myopic

Consumers often do not make rational decisions, either due to natural constraints
(uncertainty about the future), a lack of cognitive ability (slow or wrong information
processing) or systematic decision biases (loss aversion, anchoring and framing
effects, status quo bias, overconfidence, etc.). Especially recurrent systematic deci-
sion biases are ubiquitous and often lead to serious economic and psychological
welfare losses.2 Evidence from behavioural and experimental economics show that

 Thaler and Shefrin 1981, p. 39.


 Kahneman 2011.

M.F. Dold, M.A. (*)

Albert-Ludwig University Freiburg, Freiburg i. Br., Germany

© Springer International Publishing Switzerland 2016 39

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_4
40 M.F. Dold

our preferences are often unstable, context-dependent and inconsistent with our
own long-term preferences.3 Shortly, we depict incoherent preferences. Being in the
role of consumers, we, for example, often overrate our forecasting abilities and
underestimate the likelihood of contingent charges such as overdraft fees for bank
accounts, lagged payment fees for credit cards, minibar charges in a hotel rooms or
roaming charges for international mobile calls.4 In such cases, the ex-ante overcon-
fidence in our own consuming abilities repeatedly leads to severe ex post welfare
losses. Therefore, consumers’ cognitive limitations and psychological biases are
(besides the classic market failures) another potential source of market
One possible way of tackling the problem of incoherent preferences and system-
atic decision biases was introduced by Thaler and Sunstein (2003, 2008).5 Although
not fundamentally new, their approach of libertarian paternalism (which in the lit-
erature sometimes also appears under headings like soft, asymmetrical or new
paternalism) argues for a purposefully designed choice framing, so called ‘nudg-
ing’, in cases of boundedly rational individuals. Optimally, these nudges are non-­
monetary incentives in choice situations that de-bias consumers decisions in a
predictable and desired way. Thaler and Sunstein think of advising a choice archi-
tect who is in charge of framing the decisions of the individuals that could be “any-
one who must design plans for others, from human resource directors to bureaucrats
to kings.”6 They argue in favour of such an interference with individual decision
making when the shaping of the choice architecture is in the ‘best interest’ of the
affected individuals, and the freedom of choice is preserved, i.e. individuals have
the liberty to opt out of the recommended choices at no or only little costs. Thus,
libertarian paternalism takes the individual’s own subjective welfare as the basis for
regulatory recommendation: soft paternalistic policies shall be designed with the
intention that the framings imposed on each individual improves the welfare of that
individual as judged by their own lights. Therefore, the definition of nudging explic-
itly excludes legislation or interventions that are dedicated at regulating external
effects by directly altering relative prices, i.e. economic (monetary) incentives,
through taxes, subsidies or bans.
Thaler and Sunstein’s position evoked a debate on how far policy makers shall
go in shaping the choice situation of other people. One of the critics’ main argu-
ments is that choice architects, who are in charge of soft paternalist policymaking,
might be prone to the very same systematic decision errors, biased beliefs and cog-
nitive flaws like the consumers themselves, just in a more complex way and on a
higher organizational level.7 Above all, due to the Hayekian ‘knowledge problem’,
policy-making agents cannot know what is in the best interest of the affected indi-

 Camerer et al. 2004; Rabin 2002.
 Armstrong and Vickers 2012.
 Sunstein and Thaler 2003; Thaler and Sunstein 2003, 2008.
 Sunstein and Thaler 2003, p. 1190.
 Glaeser 2006.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 41

viduals, rather the individuals themselves should know their preferences far better
than any third party does.8
The following note will mitigate this argument against the concept of nudging by
pointing to the implications of the Condorcet Jury Theorem (CJT). Given we have
a situation of uncertainty and there is a true welfare maximizing means for an indi-
vidual. Then, the CJT states that (under certain conditions) a jury of experts, whose
decision procedure is simple majority voting without any group deliberation, is
more competent in finding the welfare maximizing means than the respective indi-
vidual herself. So the CJT defends Thaler and Sunstein’s notion of liberal paternal-
ism against the critique of the knowledge problem and enriches their approach in
two ways: (a) it can be taken to normatively justify soft paternalistic interventions
and (b) it gives a hint who should be in the position to make suggestions about
paternalistic consumer protection policies, namely, a jury of experts.
Most of the recent work on the Jury Theorem only deals with its formal robust-
ness when altering the theoretical assumptions but not with the very fundamental
question of the jury size.9 Since this is a crucial issue for practical implementation,
this note offers some first reasoning on the optimal jury size when considering
expected transaction costs in form of individual decision error costs and monetary
compensation for the jurors.
The structure of the note is as follows: After presenting in part 2 the knowledge
problem as a widespread point of criticism against the notion of soft paternalism,
part 3 introduces the Condorcet Jury that can sidestep most of the difficulties ema-
nating from the knowledge problem. Part 4 then discusses a simple model for find-
ing the optimal jury size and advocates educative nudges in order to minimize
expected social costs. Before concluding, part 5 proposes consumers’ meta-­
preferences as the jury’s evaluative dimension in order to rebut some possible

4.2  Libertarian Paternalism and the Knowledge Problem

In the aftermath of Thaler and Sunstein’s articles, many authors have raised objec-
tions to the use of behavioural economics to normatively justify paternalism. From
a philosophical point of view, some critics complain that libertarian paternalism
circumvents attempts of rational persuasion in that it does not appeal to critical
thinking of consumers but simply ‘shapes’ their choices in a predefined direction.10
Furthermore, many critics believe that libertarian paternalism lacks a clear welfare
criterion that a paternalistic policymaker could follow to define the desired choice
direction.11 Some critics also question the political legitimacy of nudging.

 Rizzo and Whitman 2009; Schnellenbach 2012; Sugden 2008.
 Kaniovski 2010; Kaniovski and Zaigraev 2011; Peleg and Zamir 2012.
 Hausman and Welch 2010.
 Grüne-Yanoff 2009, 2012; Mitchell 2005.
42 M.F. Dold

They argue that libertarian paternalistic policies shift the responsibility of policy-
makers to the level of individual decision-making. Therefore, nudging can be seen
as a form of ‘obfuscation policy’ that circumvents democratic control.12
Taking a traditional economic viewpoint, critics doubt the efficiency-enhancing
character of soft paternalistic interventions since, in the long run, such interventions
would hinder learning processes and reinforce the status quo.13
One of the most serious points of criticism, however, concerns the problem of
knowledge collection or information aggregation. Critics argue that Thaler and
Sunstein do not sufficiently answer the question who should be in the position to
frame individual decisions and how these policymakers should have access to the
knowledge needed to implement soft paternalist policies.14
For instance, Sugden argues that Thaler and Sunstein’s libertarian paternalism
presupposes a wise social planner with the ability to collect information about indi-
viduals’ true points of views that reflect “complete information, unlimited cognitive
abilities, and no lack of self-control.”15 So the planner needs to know about the
individuals actual preferences to identify what stimulates her welfare and what does
not. Then, guided by that information, she would be able to install the right decision
framework to promote the individual good from a neutral point of view – in Smithian
terms: the view of an impartial spectator. Yet, Sugden argues that such a planner
notoriously will fail since no person or organizational body can distinguish between
a person’s expressed preferences that potentially could be biased and her true desires
behind.16 Similarly, Rizzo and Whitman think that policymakers do not have access
to the kind of knowledge needed to impartially implement welfare-­enhancing pater-
nalist policies17 and quote Hayek that such kind of knowledge does usually not
“[…] exist[s] in concentrated or integrated form, but solely as the dispersed bits of
incomplete and frequently contradictory knowledge which all the separate individu-
als possess.”18
In line with Hayek, most of these authors argue for the market as the best welfare
creating mechanism since it establishes conditions under which the necessary infor-
mation is more likely to be mobilized and used. By tacitly approving market trans-
actions, individuals give their rational consent and hereby decentrally maximize
overall welfare, which no social planner would be capable of achieving.
However, in putting up this principle of mutual advantage, these critics of liber-
tarian paternalism depict an inconsistency in their argument. They believe in welfare
maximizing (or enhancing) decisions of individuals on markets although substantial
findings of behavioural economics demonstrate that consumers on markets repeat-
edly and systematically behave in boundedly rational ways due to cognitive biases

 Holler 2015.
 Glaeser 2006; Schnellenbach 2012.
 Rizzo and Whitman 2009; Schnellenbach 2012; Sugden 2008.
 Sunstein and Thaler 2003, p. 1162.
 Sugden 2008.
 Rizzo and Whitman 2009.
 Hayek 1945, p. 519.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 43

(in form of loss aversion, anchoring and framing effects, overconfidence, etc.). This
bounded rationality leads to severe individual and collective welfare losses.19 So
given this empirical evidence, consumer sovereignty and market transactions alone
should not be the methodological utensils of welfare economics since they cannot
answer the question of good, i.e. welfare-enhancing, policies.
In what follows, this note will show that the CJT provides the missing link to
circumvent the knowledge problem in paternalistic welfare economics while
acknowledging the insights of behavioural economic theory. If its conditions hold,
the Condorcet jury is a tool to approach the knowledge of a benevolent wise planner
that can nourish decentral market transactions of the individuals with soft paternal-
istic nudges to improve the welfare of individuals by their own standards. Following
a Condorcet jury procedure, individuals might be steered away from erroneous con-
sumption decisions while preserving their individual freedom of choice.

4.3  The Condorcet Jury as a Wise Social Planner

The Condorcet Jury Theorem (CJT) goes back to Marquis de Condorcet and his
seminal work Essai sur l’application de l’analyse á la probabilité des décisions
rendues á la pluralité des voix from 1785. Therein Condorcet argues for a calculus –
later called the Condorcet Jury Theorem – that provides a proper guarantee for
majority decisions of a parliament or an assembly being correct. In its classic inter-
pretation the CJT is applied to find a common social or moral good,20 however in
this note the new focal point is the quest for ‘the right’ means to achieve individual
welfare in consumption decisions. In line with the general logic of libertarian
paternalism,21 this note considers a nudge always to be means-oriented, it does not
attempt to alter individuals’ ends but rather helps them to order their short-term
preferences to achieve their given long-term goals.
In order to grasp the framework of the CJT when it comes to consumption, let us
make the following assumptions: There is a group of consumers in which everyone
is striving for her own welfare and there are no external effects of consuming goods.
Let us simplify and say all individuals are identical and have the same welfare goal
Π which shall be a long-term goal like life satisfaction, health, or material welfare.
Now there is a finite set of consumption options C = {c1 ,,c2 ,,…} to achieve Π. These
options are means to reach the welfare goal and could include daily activities such
as spending, saving, insuring, fasting, eating, etc. This set C is the same for all the
individuals, i.e., it is collectively congruent. The exact identity of the subset C* ⊂ C
that represents the optimum alternatives is unknown to the individuals. Although
the individuals have clear preferences over Π, they themselves are unsure about
which subset of C fits best their long-term welfare goal.

 Camerer et al. 2003.
 See, for one, Grofman and Feld 1988.
 Sunstein 2014, p. 61.
44 M.F. Dold

Let us now consider a group of experts N = {1, 2,… n} and call them ‘the jury’.
The members of this jury also face uncertainty about the best means to bring about
Π, yet each juror can rank any pair of means out of the set of consumption options
C with respect to what she believes are their respective effectiveness for fulfilling Π
(criterion of the possibility to rank-order of CJT). With respect to the choice between
any pair of alternative means, every juror i ∈ N has a probability pi to be correct in
picking the one option that promotes good Π relatively better. With respect to each
juror’s competence, we assume, pi > 0.5 so that a juror is better than a fair coin flip
to pick the better option (criterion of competence of CJT). Each member of the jury
is equally competent, i.e. p = pi for all i in N (criterion of homogeneity of CJT).
The jury’s decision procedure is a simple majority rule, i.e., the jury decides
between any two alternatives by means of a majority vote. There is no group com-
munication or deliberation among the jurors in order to avoid correlation between
the individual votes of the jurors, i.e., the probability that any juror i, i ≠ j, votes for
the better alternative is independent from j’s voting behaviour (criterion of indepen-
dence of CJT). In addition, there is no strategic voting behaviour of the jurors, i.e.,
each juror votes sincerely in the sense that she reveals her true expert knowledge
about the matter and not her personal preferences when deciding about the better
alternative (criterion of truthfulness of CJT). Moreover, it exists the possibility of a
‘correct’ ordering of means (for example, from best to worst in efficiently bringing
about Π), i.e. the jury votes upon matters that contain truth-value (criterion of truth-­
aptness of CJT). In order to make this possible, the ultimate welfare goal Π of the
individuals must first be signaled by the individual to the jury. Based on these sig-
nalled evaluative parameters, the jury then can vote on the consumption means to
best reach welfare goal Π. Hence, the jury’s decision is about practical wisdom
about the right means to bring about a given end (in Kantian terms: the jury mem-
bers express a hypothetical imperative in their votes).
If these conditions hold, the CJT states that any jury of odd number of jurors is
more likely to select the correct consumption options C* from C than any single
juror, i.e. the probability that a jury majority may vote for the relatively better alter-
native exceeds pi (non-asymptotic theorem) and the likelihood of picking the correct
alternative tends to 1 as the size of the jury approaches infinity (asymptotic theo-
rem). So, if there exists a welfare goal that can be signalled to a jury and the jury
knows about the available but unsecure consumption options, the outcome of a jury
majority voting process is an effective way of identifying the best means of maxi-
mizing individual welfare. That is, under our assumptions, the same as collective
Following Grofman, Owen and Feld the CJT calculus can be formally illustrated
as follows.22 Let PN be the probabilistic accuracy that the majority of a jury of size
N picks the better means for our group’s welfare goal Π in a pairwise comparison of
alternatives. If we assume N to be odd, simple majority m is ( N +1) / 2 . Let p be the

 Grofman et al. 1983.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 45

individual accuracy level of the jurors. Then, if the jurors vote independently, fol-
lowing the binomial distribution, the jury’s probabilistic accuracy is

h =m
PN = ∑   ( p)h (1 − p) N −h .
N  h 

If p > 0.5, then the majority’s probabilistic accuracy is

lim PN → 1.
N →∞
And if p < 0.5, the majority’s probabilistic accuracy is

lim PN → 0,
N →∞
while when p = 0.5, then, of course,

PN = 0.5.

This shows that for the same jurors’ accuracy level p, and p > 0.5, a jury decision
following majority voting is better than a single juror’s decision, i.e. PN > pi for all
N greater than 1. The outcome of the majority vote gets better if one adds more
members to the jury since Nlim PN → 1. Yet, if the accuracy level of the jurors is
p < 0.5, i.e. any single jury is more likely to vote incorrectly, then it generally holds
that adding more jurors makes things worse since Nlim PN → 0 . In this case the
optimal jury would consist of a single member.
The same logic can be applied to the quest for the true best nudging policy. Let
us assume that there is a finite set of possible nudging policies P = {p1 ,,p2 ,,…} .
Such a nudging policy pi depicts a non-monetary incentive structure that makes it
more likely for an individual to choose a certain consumption bundle from the set of
consumption options C. We further assume the jury faces uncertainty about the best
p* ∈ P that brings about the optimum subset C* most effectively. Then, if each juror
can rank any pair of means out of P with respect to their effectiveness for realizing
C*, and every juror has a probability pi > 0.5 to be correct, then the same CJT calcu-
lus holds. This time, the best consumption bundle C* would be the given end, and
the right nudge policy pi the identifiable means. Although this quest for the right
nudging policy might require some more expertise on behalf of the jury about how
individuals react to incentive structures, there is no fundamental argument against
applying the depicted (instrumental) calculus in this question, too.
Even when relaxing some assumptions, the main implications of the CJT still
hold. For one, if we assume the jurors to have different competence levels (i.e. a
varying p), but average accuracy level is p > 0.5 , and judgmental competence is
normally distributed, then the probability of the jury making the correct choice
46 M.F. Dold

under majority rule still increases to 1 as the group gets large.23 Further, Kirstein and
Wangenheim illustrate that situations exist in which the individual jurors’ accuracy
level is smaller than 0.5 but the jury decision generates a higher expected welfare
than an individual decision.24 Kaniovski demonstrates that allowing for correlation
between the jurors voting behaviour, i.e. dropping the criterion of independence,
leads to a trade-off so that the probability of the jury voting in favour of the better
option increases with the jurors’ competency but decreases with positive correlation
among their competencies.25
The limitation of the CJT to binary choice settings is much less important than it
might first appear. In situations with more than two alternatives, the decision proce-
dures can be decomposed into sequences of pairwise choices without any loss of
information.26 However, in any case, the criterion of truthfulness has to hold, other-
wise the main results of the theorem would not necessarily follow anymore.27
For the purpose of this note, let us consider the CJT in its very restrictive version
where the jury is homogeneous and there is no correlation since the main implica-
tions of the CJT also hold under heterogeneous jury competencies.28 Accepting
these assumptions, the convergence of the majority’s probabilistic accuracy PN is
strong, e.g. if the individual probability of the jury members is only p = 0.7 (which
might be a reasonable assumption since we deal with experts) and there are n = 11
jurors, then there is a probability PN larger than 0.9 that the jury picks the correct
alternative under majority vote.
First experimental studies of simple majority jury decision rules support these
theoretical predictions in the laboratory. Guarnaschelli, McKelvey and Palfrey find
strong experimental evidence for the effectiveness of small juries to find ‘the truth’
(in their study they call it ‘the signal’). If pi = 0.7, n = 6 and there is no group delib-
eration, then ‘[u]nder majority rule, the subjects voted the same direction as their
signals more than 94 % of the time.’29 Another experimental study by Ladha, Miller
and Oppenheimer illustrates that judgmental accuracies of majority rule are empiri-
cally robust and real juries do even better than predicted by the theoretical results of
the CJT. They also show that groups do clearly better than single individuals, even
in an experimental setting that was designed to maximize the advantages of uninfor-
mative voting by individuals.30
Given these results, the CJT has found its way back into academic discussion
mainly to normatively justify majority decision procedures within representative

 Grofman et al. 1983.
 Kirstein and Wangenheim 2010.
 Kaniovski 2010. See also Berg 1993, Ladha 1993 or Peleg and Zamir 2012 for an in-depth dis-
cussion of the criterion of independence.
 Berg 1996.
 Feddersen and Pesendorfer 1998.
 Grofman et al. 1983; Berg 1996.
 Guarnaschelli et al. 2000, p. 413.
 Ladha et al. 1996.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 47

democratic institutions.31 However, as already indicated in the illustration of the

CJT calculus, its insights can also be applied to questions of consumer protection
and give reason to consider jury decisions in cases of boundedly rational individu-
als. Apart from hard regulatory effort, consumer protection policies are typically
designed to nudge consumers so that they are better-off. However, such policies
often lack any clear welfare criteria and, above all, miss any evaluative procedure
how to find the right decision framings for welfare-enhancing consumption. Now
following the implications of CJT, policymakers shall acknowledge that there are
different long-term welfare criteria Π, which they do not decide upon, but which can
be signalled to them by the consumers. Then, a Condorcet jury may be a good
evaluative instrument to filter available consumption options C how to best reach
this given welfare goal Π and to identify an appropriate nudging policy p* subse-
quently, such that the individuals are more likely to consume C*.
So, in practice, the Condorcet jury voting procedure would be three-staged when
it comes to consumer protection: First of all, consumers have to signal their long-­
term welfare goals to the jury. Secondly, the jury has to elect the best short-term
consumption bundle by majority rule and, thirdly, the jury votes upon the best
choice architecture to frame individual behaviour given the optimal consumption
bundle derived in step two, again following majority voting. In both of the latter
stages, the Condorcet voting procedure is applicable since the underlying evaluative
dimensions, i.e. the welfare goal Π and the consumption bundle C*, are fixed and the
assessing issue is truth-apt inasmuch as we deal with questions of practical wisdom.
Given uncertainty, at both the second and the third stage, a jury might be the better
judge than the individuals themselves if the jurors all have a certain level of exper-
tise with respect to the underlying normative dimensions. Then, it follows that –
even in a situation in which a single consumer might have a higher competence than
a single member of the jury – it is reasonable to entrust the framing of the decision
to a jury since the jury’s majority accuracy level will be mostly higher than the one
of the single individual.
Consequently, the CJT can be considered as a rational justification of soft pater-
nalistic framings of consumption decisions. The Condorcet jury approximates a
rational self in identifying the best consumption bundle and, by means of the nudg-
ing policy, rules out problems of bounded willpower or present bias consumption.
Therefore, the CJT delivers a counterargument to the critique that a soft paternalis-
tic planer is prone to the very same decision biases as the nudged individual

4.4  The Optimal Jury Size and Educative Nudges

Given the theoretical insights of the CJT, it would be optimal to establish large
juries following the maxim: the bigger, the better. The more experts there are, the
likelier a jury would approach infallibility. However, it does not need a lot of sceptic

 Black 1958; Grofman and Feld 1988; List and Goodin 2001; Miller 1986.
48 M.F. Dold

to see that this would be an impractical suggestion. Self-evidently, experts cause

economic costs. It is very unlikely that they will give their information pro bono.
Installing a jury will cause transaction costs since the jurors want to be compensated
monetarily for their expertise and their time effort. Accordingly, a bigger group of
jurors N will cause higher costs for the boundedly rational individuals who have an
interest in self-regulatory nudging. On the other hand, if we had a bigger jury, the
economic costs from decision flaws could be reduced since a larger jury is more
likely to install the right kind of decision framework to nudge the individuals to a
welfare maximizing behaviour. Considering these aspects, the costs of decision
flaws and the transaction costs, we can model both expenses for the individuals
depending on the size of the jury of n experts.32
Let Φ denote the monetary damage of an individual’s decision errors (‘the error
costs’), whereas a given parameter α indicates the cognitive ability of the respective
individual, i.e. a higher α shows that the individual has a decent expertise in making
the right consumption decision with respect to her long-term welfare goal and so
reduces the likelihood of the occurrence of individual decision errors. The probabil-
ity of the jury members PN to vote correctly is given, homogeneous and significantly
higher than 0.5.
Since the likelihood of decision flaws decreases with an increase of jury mem-

bers, Φ (n;2 α) is a probability function decreasing in n, so < 0 . Let us further
dΦ dn
assume < 0 , indicating the positive but decreasing effects resulting from the
dn 2
right kind of nudges implemented by a larger jury. This seems to be a realistic
assumption since a larger jury approaches a better result but does so in an asymp-
totic way (see the asymptotic theorem above).
Let Ψ denote the transaction costs of the jury experts (‘the jury costs’) and β a
parameter that indicates the monetary aspiration level of the experts.
Straightforwardly, let us assume Ψ (n; β) to be strictly increasing and linear in n,

with Ψ (0) = 0, reflecting the basic assumption of a homogeneous jury, i.e. >0
dΨ dn
and = const . For simplification, we assume that there are no other costs for the
individuals apart from error costs and the jury costs. Consequently, we can sum up
these two forms of costs to get the total costs for the group of individuals, denoting
it as Θ:

Θ (n; α ; β ) = Φ (n; α ) +Ψ (n; β ).

The competence effect of a larger jury is depicted in the falling decision error curve
Φ (n; a) and the increasing costs of a larger jury is shown by the ascending jury cost
function Ψ (n; β). The total cost curve Θ is obtained by summing vertically the two

 This idea is based on Calabresi’s seminal work The Costs of Accidents (1970) which deals with
efficiency analysis of tort law. Therein, Calabresi argues that the goal of tort law is the minimiza-
tion of total expected accident costs, which include both the expected costs of the accidents and the
spent costs for avoiding the accidents.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 49

functions Φ (n; α) and Ψ (n; β) at every level of jury size n. Although the negative
effects of decision errors would diminish when installing larger juries, the jury costs
would increase naturally. So there is a trade-off between the error costs emanating
from decision flaws and transaction costs emanating from a larger jury. At small
jury sizes, there were almost no transaction costs, but a high likelihood for the
occurrence of decision errors. Because the total cost curve Θ (n; α; β) is U-shaped,
there is a cost minimizing team size of n individuals, denoted n*, which follows
from the first-order condition:

dΨ dΦ
=− .
dn dn

So at the optimal team size n*, the marginal costs of one more jury member equals
the marginal benefits of a reduction in one monetary unit of decision errors. Facing
this trade-off, individuals will have to accept a certain threshold of decision error
costs when designing the optimal size of a jury. The first-best solution in a
transaction-­cost free world would be at n → ∞ , but considering the positive trans-
action costs in form of monetary compensation for jurors’ information, n* turns out
to be the second-best solution in the set of achievable jury sizes.
This very stylized discussion shall illustrate the following. First, since the costs
arising from decision errors and the jury costs will differ with the individuals cogni-
tive abilities α and the jurors’ monetary aspiration level β, there is not one optimal
jury size for all soft paternalistic issues. Secondly, these parameters α and β indicate
regulating screws, i.e. the political scope, for reducing the total costs.
On the one hand, one can try to get a juror’s information at a lower per unit cost,
e.g. by subsidizing the jury’s work, so that β decreases. Then the jury cost function
Ψ (n; β) turns to the right and the optimal jury size n* gets bigger, but the total social
costs Θ are reduced in the optimum. Yet, this approach would be inefficient over
time since the subsidies themselves have to be constantly financed somehow (e.g.
by taxing individuals), and, more importantly, the error proneness of the individuals
would stay the same.
Therefore, a better – i.e. dynamically more efficient – solution for reducing the
total costs would be to directly improve the consumers’ cognitive abilities α. If α
increases, the error cost function Φ (n; α) turns to the left and both the jury size n*
and the social costs Θ get smaller so that the need for a jury and the overall social
costs are gradually reduced.
In order to increase α, the jury can implement educative nudges.33 Contrary to
mere manipulative nudging, where individuals are unaware of the nudging process,
educative nudges (in form of feedbacks or reminders) promote learning and address
the information level of the consumers directly. Consider, for example, decisions in
which consumers have difficulty linking their actions with the later outcomes. Here,
the implementation of a feedback nudge can provide better information about the
link between actions and outcomes. Such a feedback nudge would be educative in

 Sunstein 2015, p. 18.
50 M.F. Dold

the sense that it depicts an investment in consumers’ ‘stock of knowledge’ about the
consequences of their own behaviour.34 In a meta-analysis, Darby (2006) shows that
providing consumers with feedback on home energy consumption (e.g., a real-time
energy use display monitor) reduces energy consumption by 5–15 %.35 Such feed-
back information make consumers rethink their short-term preferences. By pointing
to the medium- or long-term choice consequences, these educative nudges debias
individuals’ choices. Naturally, there are costs emerging from implementing these
educative nudges in the first place. Yet, this form of nudging is more efficient than
constantly subsidizing the jury’s work, since they reduce the error costs of the con-
sumers over time by increasing the consumer’s cognitive ability α gradually.
Another good example for an educative nudge is a fair trade label for clothing.36
This labelling informs consumers about the labour conditions in the production pro-
cess, makes the channel of distribution more transparent and explains the composi-
tion of the end prices to the consumers. In doing so, such a nudge goes beyond the
mere price signal and informs the consumer about the shrouded attributes of the
product.37 Consumers are able to learn something about the product in their act of
consumption. This ‘moment of education’ makes it more likely that individuals
align their preferences with the true (i.e. long-term) value they attach to that good.
So educative nudges bring actual individuals closer to the hypothetical version of
themselves that has fully informed preferences about the world.38 When nudging
consumers in this way, they themselves increase their knowledge and understand-
ing of decision biases (in this case, for example, impulsive buying of low-cost prod-
ucts), so that they can make better choices for themselves (in this case, for example,
buying products that meet their moral standards).

4.5  R
 ebutting Some Critique: Thinking in Hierarchical
Preference Structures

Let us sum up: If the welfare criterion Π can be signalled to a group of jurors (stage
1), effective means to reach Π might be identified using the Condorcet jury proce-
dure (stage 2). Based on the found consumption bundle the jury can frame the indi-
viduals’ choice in such a way that it is easier and more likely for the individuals to
pick the most effective consumption option(s) with regard to their long-term welfare
goals (stage 3). The optimal jury size will then depend on the individuals’ cognitive
abilities and the jurors’ monetary aspiration level. In order to reduce overall costs,
educative nudges might be efficient since consumers can make better choices for

 Gigerenzer 2014.
 Darby 2006.
 For example in form of the well-known FLO International’s Fairtrade certification.
 Loureiro and Lotade 2005.
 The individual with “complete information, unlimited cognitive abilities, and no lack of will-
power” is the normative benchmark individual for Thaler and Sunstein 2003, p. 176.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 51

themselves, which in turn reduces the need for the jury over time. So far, so good.
Now critics might struggle with the idea that individuals are able to express their
welfare criterion in the first place. How should consumers be able to signal their
long-term goal Π to the jury, if they have contradictory and context-dependent pref-
erences as demonstrated by various behavioural and experimental economic
A way out of this dilemma is a richer understanding of human reasoning in hier-
archical preference structures.39 In Thaler and Shefrin’s terms: “The individual at a
point in time is assumed to be both a farsighted planner and a myopic doer.”40
On the one hand, in daily consumption decisions individuals are myopic doers.
They often express outcome-oriented desires that can be contradictory and prone to
many decision biases resulting from, what Kahneman calls, System 1 thinking.41
This is an automatic, emotional and mostly intuitive way of thinking that requires
only little cognitive engagement, is driven by impulsive desires and marked by
direct responses to incentives in the environment. On the other hand, in long-term
considerations individuals are farsighted planners. They can name stable, non-­
contradictory, long-term wishes like health, wealth, or happiness.42 These consider-
ations result from System 2 thinking, which is an effortful, logical and conscious
way of making decisions based on knowledge about facts and values. This kind of
thinking portrays our values and intentions, which requires focused cognitive effort.
In line with Kahneman’s distinction of two cognitive modes, one can relate short-­
term preferences to System 1 and long-term preferences to System 2 thinking.
Kahneman’s conception explains why daily preferences are rather myopic and
prone to decision biases and long-term preferences are rather thoughtful, stable and
well-defined. Myopic preferences can be called first-order preferences and long-­
term preferences can be named second-order or meta-preferences to indicate the
hierarchical relation between these two types of preferences. Since the latter ones
are reflective and well-defined they should – from a normative point of view – over-
rule the first-level ones.
An example for a conflicting hierarchical preference structure is a ‘shopaholic’
person who maximizes her short-term, myopic utility by impulsively buying all
sorts of new goods to get a thrill of the consuming act but whose second-order,
long-term preference is such that she would in fact prefer to not surrender to her
addiction and smooth her consumption path over time.43 Another example for a
conflicting hierarchical preference structure would be a person who despite her
long-term wish to lose weight, still buys the chocolate bar at the checkout in the
supermarket and thus violates her meta-preference for a healthy lifestyle.

 Frankfurt 1971; Thaler and Shefrin 1981.
 Thaler and Shefrin 1981, p. 39.
 Kahneman 2011.
 Strack and Deutsch 2004.
 Mueller et al. 2010.
52 M.F. Dold

Of course, such consuming decisions can also reflect a conscious act by an indi-
vidual who values short-term pleasures in life more highly than their health.44
However, if it reflects a non-deliberate, impulsive act that is accompanied by feel-
ings of ex post regret, the individual might be better off, if she had listened to her
long-term wishes since they more likely reflect her true intentions. Some econo-
mists deny this fundamental, normative gap between values or intentions on the one
hand, and affective, impulsive behaviour on the other hand.45 They argue that espe-
cially System 1 behaviour would depict the true costs and benefits of decisions and
long-term preferences would be systematically biased towards all sorts of utopian
wishes. However, after compulsive System 1 consuming acts many people express
severe ex post regrets for what they have done.46 So there is an actual (and often
material) welfare loss people are suffering from due to present biased decisions
emanating from System 1 thinking. Since second-order preferences reflect System
2 thinking that is consistent over time, there is a further normative argument that
they should have authority. Further, their level of information is higher and the ex-­
ante estimation of costs and benefits is more in line with the factual ex post
So Kirchgässner rightly states that
[…] the fact that human persons, reflecting on themselves, usually take on the long-run
perspective rather provides evidence for the long-run preferences, as does the fact that
individuals often heavily underestimate the costs of the actions that are guided by their
short-run preferences. Otherwise it is difficult to understand why people regret their own
earlier behaviour when foreseeable consequences become obvious that were not taken into
account before.48

Depending on the choice context, individuals depict different, contradictory

short-run preferences. Referring to one of these preferences does not make sense
when searching for the true preference signals for designing the right nudge. Short-­
run preferences imply a multitude of possible nudges. It would never be possible for
a jury to pick one of these and name it superior to all the others if they had no supe-
rior criterion to judge the right preference.49
Yet, if we accept the fact that individuals have stable long-term preferences
resulting from System 2 thinking, we can also assume that individuals are able to
signal them coherently to a jury of experts.50 So considering the fact that individuals
do have non-contradictory meta-preferences is a way of handling the problem of
incoherent short-term preferences that is puzzling for some critics of libertarian

 On the importance of pleasure, see Dolan 2014.
 Schnellenbach 2012.
 Mueller et al. 2010. Sunstein 2015.
 Sunstein 2015.
 Kirchgässner 2014, p. 15.
 Kirchgässner 2013.
 For empirical arguments backing the existence of stable long-term preferences, see: Kahneman
2003, or Strack and Deutsch 2004.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 53

Consumers might express their meta-preferences in reflective situations that acti-

vate deliberate System 2 thinking, e.g. in polls, through questionnaires, or in public
debates.51 Based on the expressed meta-preference (e.g. for ethical clothing) the
jurors are in charge of aligning the available short-term consume options (e.g. low-­
cost shirt vs. locally produced shirt vs. oversea fair trade shirt) by means of nudges
(e.g. by fair trade labelling) with the individuals’ long-term goals.
Having in mind this distinction of preference expression, the notion of libertarian
paternalism is still compatible with methodological individualism. The individual
herself is the one who autonomously expresses her long-term preferences, it is only
the jury who implements the right incentive structure so that the individual is capa-
ble of aligning her short-term preferences with her long-term welfare goals. The
individual herself stays the bearer of preferences and can always opt out at no costs
from the incentivized consumption bundle. The jury, on the other hand, is commit-
ted to sincerely voting on means and therefore does not interfere with the realm of
individual values. Applying the CJT respects consumer sovereignty over long-term
preferences, and helps consumers to behave rationally in the short run by giving and
framing information. These framings of the individual choice architecture can be
understood as individual self-imposed rules established via the rational means-end
calculation within the Condorcet jury. In its voting procedure, the jury focuses on
means (i.e. choice framings to circumvent biases), not ends (i.e. individuals’ long-­
term preferences).
In general, besides labelling, a jury could think of a wide variety of nudging poli-
cies to make individuals’ choices more likely to fit their long-term welfare goals:
providing palatable information about what others are doing (‘social norm feed-
back’), telling risk narratives to accompany dangerous products, changing a passive
default rule to an active choosing scenario, including cooling-off periods before
making complex market transactions, or altering habitual and expected product
placements.52 This non-inclusive list of examples gives a first hint regarding how a
jury might effectively frame decisions by accepting that System 1 thinking is domi-
nant in consumption decisions, but System 2 intentions need to be executed through
the right choice architecture to fulfil the individual’s actual long-term welfare goals.
Of course, identifying the right nudges requires expertise on the part of the jurors
about how an individual reacts to modifications of incentives in her consumption
environment. In this regard, recent, neuro-economic and psychological research
might support the hypothesis that experts are quite good in empathizing with the
thinking of third-party individuals.53 Moreover, as illustrated above, this empathiz-
ing process does not need to be ideal. It is completely sufficient, if the members of
a jury do it in a way that their single probability to make a correct assessment stays
larger than 0.5. Taken these two points together, we should not bother too much
about the ‘empathizing problem’ in the debate.

 For an in-depth discussion on how to identify meta-preferences, see Beshears et al. (2008). The
authors identify six approaches that jointly contribute to the identification of, what they call, nor-
mative preferences: structural estimation, active decisions, asymptotic choice, aggregated revealed
preferences, reported preferences and informed preferences.
 Camerer et al. 2003; Thaler and Sunstein 2008.
 Fehr and Singer 2005; Singer and Hein 2012.
54 M.F. Dold

4.6  Concluding Remarks

Consumers make daily choices under uncertainty that are subject to systematic
decision biases. Based on this observation, Thaler and Sunstein have proposed an
effective framing of the choice architecture in order to help individuals to make bet-
ter decisions.54 However, in their notion of soft paternalism they make no suggestion
who should be in the position to design the choice architecture and they also do not
give an answer to the knowledge problem many critics associate with it. This note
has mitigated both points of criticism by considering implications of the Condorcet
Jury Theorem. The theorem offers majority voting as an effective means for electing
the right nudging policies to frame individuals’ consumption options. Considering
the costs arising from decision errors and the jury implementation, this note has
shown, on a very general level, how to compute the optimal jury size and advocated
meta-preferences as a welfare criterion upon which the jury can base its calculus for
educative nudges. In times when many are against experts or regulatory measures,
and mainly favour the plain outcome of markets, this note argues for jury panels to
improve the choice architectures of consumers within markets.
This note has discussed the theorem and the respective issue of the optimal jury
size in a stylized way. The assumptions of the original version of the Jury Theorem
are restrictive and assume independence among the jurors. Some authors have
shown that dependence among jurors need not make the Condorcet Jury irrelevant,
only negative correlation between the voters is crucial for its applicability.55
However, as Holler and Leroch rightly point out,56 negative correlation is a ques-
tionable condition since empirical evidence suggests that opinions within juries
often tend to move towards extreme views.57 This alludes to the need for further
studies on the issue of the right institutional design to reach the outcome of an unbi-
ased Condorcet jury. Here, the combination of economic thinking in incentive
structures and psychological work about judicial juries could lead to further, essen-
tial suggestions for unbiased Condorcet juries in the field of soft paternalistic
First experimental evidence on Condorcet juries supports the proposition of the
CJT and displays that the main results of the theorem hold in the laboratory.58 Other
recent experimental studies underpin the view that it is often more rational if exter-
nal third parties make decisions when uncertainty or risk is involved.59 In this con-
text, further empirical and experimental studies should be conducted, especially on
the question whether a group of jurors is able to show enough empathy to e­ ffectively

 Thaler and Sunstein 2003, 2008.
 Kaniovski 2010.
 Holler and Leroch 2010.
 See also Sunstein 1999.
 Guarnaschelli et al. 2000; Ladha 1996.
 Brock 2011.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 55

nudge the decisions of consumers in a desired direction while preserving the latters’
freedom of choice.
The effectiveness of educatively nudging real-life consumption decisions needs
to be further assessed in field studies. Many traditional approaches in consumer
protection depend on making appeals to System 2 thinking.60 Based on providing
reasonable information, the goal is to change beliefs and attitudes by explaining the
prospects of future benefits. Still, at the most, these efforts have been modestly
effective in changing consuming patterns.61 Therefore, when implementing educa-
tive nudges one might also address System 1 thinking, since then information is
much easier processed and translated into action. This means educative nudges
should be accompanied by sound emotional framings like graphic warnings or
reminders62 and effective default rules.63
Allowing for heterogonous consumers, the discussed jury nudging procedure
does not have to be the gold standard in every branch of consumer protection. Here,
Camerer et al. give some powerful ideas when a jury might be a good way to go.
According to the authors, a nudging policy should only be applied if its imple-
mented incentive structure works asymmetrical, i.e. those can be helped who are
prone to decision biases but the rational consumers must not be disturbed or fined in
their consumption patterns. By considering this rule of thumb, one is able to circum-
vent the problem that nudging might only help the boundedly rational consumers
but relatively weaken the position of the wise ones.64 One should also be very cau-
tious to democratically legitimize juries by the affected consumers to not undermine
their normative and factual force.65 Above all, expert decision procedures should
always be transparent and open to public scrutiny.

Acknowledgement  I thank Manfred J. Holler for an in-depth discussion on the draft of the paper
and I am grateful for the valuable comments of Christian Schubert, Jan Schnellenbach and the
participants of the 4th Law and Economics Conference in Lucerne on Nudging. All errors, of
course, remain mine.


Armstrong, Mark, and John Vickers. 2012. Consumer protection and contingent charges. Journal
of Economic Literature 50(2): 477–493.
Berg, Sven. 1993. Condorcet’s jury theorem, dependency among jurors. Social Choice and Welfare
10(1): 87–95.

 Marteau et al. 2011.
 WHO 2008; Willis 2011.
 Jolls 2013.
 Johnson et al. 2002.
 Armstrong and Vickers 2012.
 Dold and Holler 2015.
56 M.F. Dold

Berg, Sven. 1996. Condorcet’s jury theorem and the reliability of majority voting. Group Decision
and Negotiation 5(3): 229–238.
Beshears, John, James J. Choi, David Laibson, and Brigitte C. Madrian. 2008. How are prefer-
ences revealed. Journal of Public Economics 92(8): 1787–1794.
Black, Duncan. 1958. The theory of committees and elections. Cambridge: Cambridge University
Brock, J. Michelle, Andreas Lange, and Y. Ozbay Erkut. 2011. Dictating the risk -experimental
evidence on giving in risky environments. Working papers 139, European Bank for
Reconstruction and Development.
Calabresi, Guido. 1970. The cost of accidents: A legal and economic analysis. Yale: Yale University
Camerer, Colin, Samuel Issacharoff, George Loewenstein, Ted O’Donoghue, and Matthew Rabin.
2003. Regulation for conservatives: Behavioural economics and the case for ‘asymmetric
paternalism’. University of Pennsylvania Law Review 151: 1211–1254.
Camerer, Colin, George Loewenstein, and Matthew Rabin. 2004. Behavioural economics: Past,
present, future. In Advances in behavioural economics, ed. C.F. Camerer, G. Loewenstein, and
M. Rabin, 3–51. Princeton: Princeton University Press.
Condorcet, Marquis de. 1785. Essai sur l’application de l’analyse a la probability des decisions
rendues a la plurality des voix. Paris: De l’imprimerie royale. Translated in 1976 to ‘Essay on
the Application of Mathematics to the Theory of Decision-Making’. In Condorcet: Selected
Writings, ed. K. M. Baker. Indianapolis: Bobbs-Merril.
Darby, Sarah. 2006. The effectiveness of feedback on energy consumption. A Review for DEFRA
of the Literature on Metering, Billing and direct Displays 486.
Dolan, Paul. 2014. Happiness by design: Change what you do, not how you think. New York:
Dold, Malte, and Manfred J. Holler. 2015. Nudging, responsibility, and the Condorcet paradox
(paper in progress).
Feddersen, Timothy, and Wolfgang Pesendorfer. 1998. Convicting the innocent: The inferiority of
unanimous jury verdicts under strategic voting. American Political Science Review 92: 23–35.
Fehr, Ernst, and Tanja Singer. 2005. The neuroeconomics of mind reading and empathy. American
Economic Review, AEA Papers and Proceedings 95(2):340–345.
Frankfurt, H.G. 1971. Freedom of the will and the concept of a person. Journal of Philosophy 68:
Gigerenzer, Gerd. 2014. Risk savvy: How to make good decisions. New York: Penguin.
Glaeser, Edward L. 2006. Paternalism and psychology. University of Chicago Law Review 73:
Grofman, Bernard, and Scott L. Feld. 1988. Rousseaus general will: A condorcetian perspective.
The American Political Science Review 82(2): 567–576.
Grofman, Bernard, Guillermo Owen, and Scott L. Feld. 1983. Thirteen theorems in search of the
truth. Theory and Decision 15: 261–278.
Grüne-Yanoff, Till. 2009. Welfare notions for soft paternalism, Papers on economics and evolu-
tion, vol. 0917. Jena: Evolutionary Economics Group, MPI.
Grüne-Yanoff, Till. 2012. Old wine in new casks: Libertarian paternalism still violates liberal
principles. Social Choice and Welfare 38: 635–645.
Guarnaschelli, Serena, Richard D. McKelvey, and Thomas R. Palfrey. 2000. An experimental
study of jury decision rules. American Political Science Review 94: 407–423.
Hausman, Daniel M., and Brynn Welch. 2010. Debate: To nudge or not to nudge. Journal of
Political Philosophy 18(1): 123–136.
Hayek, Friedrich A. 1945. The use of knowledge in society. American Economic Review 35:
Holler, M.J., and M.A. Leroch. 2010. Jury on stage: A common law play. European Journal of Law
and Economics 30(2): 89–110.
4  Condorcet’s Jury Theorem as a Rational Justification of Soft Paternalistic… 57

Holler, Manfred J. 2015. Paternalism, gamification, or art: An introductory note. Homo

Oeconomicus 32(2). (forthcoming).
Johnson, Eric J., Steven Bellman, and Gerald L. Lohse. 2002. Defaults, framing and privacy: Why
opting in-opting out. Marketing Letters 13(1): 5–15.
Jolls, Christine. 2013. Product warnings, debiasing, and free speech: The case of tobacco regula-
tion. Journal of Institutional and Theoretical Economics 169(1): 53–78.
Kahneman, Daniel. 2003. Maps of bounded rationality: Psychology for behavioural economics.
The American Economic Review 93(5): 1449–1475.
Kahneman, Daniel. 2011. Thinking, fast and slow. New York: Macmillan.
Kaniovski, Serguei. 2010. Aggregation of correlated votes and Condorcet’s jury theorem. Theory
and Decision 69(3): 453–468.
Kaniovski, Serguei, and Alexander Zaigraev. 2011. Optimal jury design for homogeneous juries
with correlated votes. Theory and Decision 71(4): 439–459.
Kirchgässner, Gebhard. 2013. Sanfter Paternalismus, meritorische Güter und der normative
Individualismus. In Grundfragen der Ökonomik: Grenzen der Konsumentensouveränität, ed.
M. Held, G. Kubon-Gilke, and R. Sturn, 41–62. Marburg: Metropolis-Verlag.
Kirchgässner, Gebhard. 2014. Soft paternalism, merit goods, and normative individualism. CESifo
working paper no. 4688.
Kirstein, Roland, and Georg von Wangenheim. 2010. A generalized Condorcet jury theorem with
two independent probabilities of error (No. 11-2010). Joint discussion paper series in
Ladha, Krishna K. 1993. Condorcet’s jury theorem in light of de Finetti’s theorem: Majority-rule
voting with -correlated votes. Social Choice and Welfare 10: 69–86.
Ladha, Krishna K., Gary J. Miller, and Joe Oppenheimer. 1996. Information aggregation by major-
ity rule: Theory and experiments. Mimeo, University of Maryland.
List, Christian, and Robert E. Goodin. 2001. Epistemic democracy: Generalizing the Condorcet
jury theorem. Journal of Political Philosophy 9(3): 277–306.
Loureiro, Maria L., and Justus Lotade. 2005. Do fair trade and eco-labels in coffee wake up the
consumer conscience? Ecological Economics 53(1): 129–138.
Marteau, Theresa M., David Ogilvie, Martin Roland, Marc Suhrcke, and Michael P. Kelly. 2011.
Judging nudging: Can nudging improve population health? BMJ 342: 263–265.
Miller, Nicholas R. 1986. Information, electorates, and democracy: Some extensions and interpre-
tations of the Condorcet jury theorem. In Information pooling and group decision making, ed.
B. Grofman and G. Owen, 173–192. Greenwich: JAI Press.
Mitchell, Greogory. 2005. Libertarian paternalism is an oxymoron. Northwestern University Law
Review 99(3): 1245–1277.
Mueller, Astrid, James E. Mitchell, Ross D. Crosby, Olaf Gefeller, Ronald J. Faber, Alexandra
Martin, and Martina de Zwaan. 2010. Estimated prevalence of compulsive buying in Germany
and its association with sociodemographic characteristics and depressive symptoms. Psychiatry
Research 180(2): 137–142.
Peleg, Bezalel, and Shmuel Zamir. 2012. Extending the Condorcet jury theorem to a general
dependent jury. Social Choice and Welfare 39(1): 91–125.
Rabin, Matthew. 2002. A perspective on psychology and economics. European Economic Review
46: 657–685.
Rizzo, Mario J. and Douglas G. Whitman. 2009. The knowledge problem of the new paternalism.
Brigham Young University Law Review 905: 103–161.
Schnellenbach, Jan. 2012. Nudges and norms: On the political economy of soft paternalism.
European Journal of Political Economy 28(2): 266–277.
Singer, Tania, and Grit Hein. 2012. Human empathy through the lens of psychology and social
neuroscience. In The primate mind: Built to connect with other minds, ed. F.B.M. de Waal and
P.F. Ferrari, 158–174. Cambridge, MA: Harvard University Press.
Strack, Fritz, and Roland Deutsch. 2004. Reflective and impulsive determinants of social behav-
iour. Personality and Social Psychology Review 8(3): 220–247.
58 M.F. Dold

Sugden, Robert. 2008. Why incoherent preferences do not justify paternalism. Constitutional
Political Economy 19(3): 226–248.
Sunstein, Cass R. 1999. The law of group polarization. University of Chicago: John M. Ohlin Law
& Economics Working Paper No. 91.
Sunstein, Cass R. 2014. Why nudge? The politics of libertarian paternalism. New Haven: Yale
University Press.
Sunstein, Cass R. 2015. Nudging and choice architecture: Ethical considerations. Yale Journal on
Regulation, forthcoming.
Sunstein, Cass R., and Richard H. Thaler. 2003. Libertarian paternalism is not an oxymoron. The
University of Chicago Law Review 70(4): 1159–1202.
Thaler, Richard H., and Hersh M. Shefrin. 1981. An economic theory of self-control. The Journal
of Political Economy 89: 392–406.
Thaler, Richard H., and Cass R. Sunstein. 2003. Libertarian paternalism. American Economic
Review 93(2): 175–179.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge: Improving decisions about health, wealth,
and happiness. New Haven: Yale University Press.
WHO. 2008. Behaviour change strategies and health: the role of health systems. Regional
Committee for Europe.
Willis, Lauren E. 2011. The financial education fallacy. The American Economic Review 101:
Chapter 5
To What Extent Should the State Protect
Human Beings from Themselves?
An Analysis from a Human Rights Perspective

Peter G. Kirchschlaeger

Abstract Nudging could seem to be attractive from a human rights perspective if

this nudging would lead to a realization of human rights. At the same time it is pre-
cisely part of the nucleus of human rights to protect the individual from heteronomy,
respectively from abuse of power by a collective. In other words, human rights
protect all the powerless from the powerful. Human beings enjoy human rights as
individuals, independent of any collective. However, human rights struggle also
with the particular interests of states and non-state actors. Furthermore, from a
human rights perspective, there is a necessity to respect the autonomy of an indi-
vidual in order to remain coherent with the core concept of human rights. This
necessity to respect the autonomy of an individual goes as far as that human rights
themselves need a moral justification because autonomy entails knowing why one’s
freedom should be restricted by somebody or something. Therefore the questions
arise if and how this aspect could be combined with the above-mentioned argumen-
tation that the end would justify the means, and if and how nudging could be justi-
fied. In the following paper these issues will be addressed and it will be attempted
to develop a solution to the question as to what extent the state should protect human
beings from themselves.

5.1 Introduction

Cass R. Sunstein and Richard Thaler suggest in “Nudge – Improving Decisions

About Health, Wealth, and Happiness”1 that the challenges of today’s world could
be better met if states would nudge humans via libertarian paternalism by correcting
cognitive biases. From a human rights perspective, this could seem to be attractive

Sunstein and Thaler 2008.
P.G. Kirchschlaeger (*)
Faculty of Theology, University of Lucerne / Yale University,
Frohburgstrasse, 6002 Lucerne, Switzerland

© Springer International Publishing Switzerland 2016 59

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_5
60 P.G. Kirchschlaeger

if this nudging would lead to the realization of human rights. The latter would be
extremely welcome:
On the one hand, nudging by the states protecting human beings from them-
selves could be understood as part of the responsibility of the states for the imple-
mentation of human rights.
On the other, it belongs precisely to the nucleus of human rights to protect the
individual from heteronomy, respectively from abuse of power by a collective. In
other words, human rights protect all the powerless from the powerful. The univer-
sality of human rights implies that human rights apply to all human beings and are
therefore universal. Human beings enjoy human rights as individuals, independent
of any collective. All human beings are holders of human rights, irrespective of
what they do, where they come from, where they live, and which nationality they
have or which community they belong to. Human rights struggle exactly, however,
with the particular interests of states and non-state actors, e.g. states limiting human
rights of their citizens by claiming the priority of their sovereignty over the univer-
sality of human rights. Human rights violations e.g. by traditions, cultures, civiliza-
tions, religions, or value systems are often allegedly justified by cultural and
religious leaders concerned about a possible loss of institutional power and influ-
ence and usually not by core elements of the traditional, cultural, or religious ‘truth’
shared by the communities.
Furthermore, from a human rights perspective, there is a necessity to respect the
autonomy of an individual in order to remain coherent with the core concept of
human rights. This necessity to respect the autonomy of an individual even goes so
far as to require that human rights themselves need a moral justification as auton-
omy entails knowing why one’s freedom should be restricted by somebody or
something including human rights. Therefore the questions arise if and how this
aspect could be combined with the above-mentioned argumentation that the end
would justify the means, and if and how nudging could be justified.
In the following paper these issues will be addressed. It will be attempted to
develop a solution to the question to what extent the state should protect human
beings from themselves.

5.2 The Realization of Human Rights

Human rights protect every human in the essential elements and spheres of human
existence needed for survival and for a life as a human being. They do not represent
a high ideal but a minimal standard and are therefore fundamental. Actually it seems
to be a shame that this minimal standard is sometimes called an ideal or even a
utopia2 because human rights build a rather minimalistic approach. As they protect
every human in exactly the same way, they are universal and egalitarian – belonging
to every human regardless of skin colour, nationality, political convictions, religious

Moyn 2010.
5 To What Extent Should the State Protect Human Beings from Themselves? 61

beliefs, social standing, gender, sexual orientation, age,… Human rights are indi-
vidual rights because humans are holders of human rights as individuals. As one
does not have to fulfil any requirement (e.g. having a certain income, paying taxes,
respecting the law) for the entitlement to human rights, human rights are categori-
cal. Furthermore, human rights are enforceable. The enforceability in a legal system
embraces the implementation of human rights, which must be possible and realized
in order to guarantee every human this minimal standard for survival and for a life
as a human. Human rights embrace four dimensions: legal, historical, political, and
moral dimension.3
Although there is a generally recognizable positive tendency of acceptance of
human rights by states and non-state actors, an increasing international institution-
alization for the protection of human rights, a progress in the mechanisms for moni-
toring human rights performance by states to respect the universality of human
rights, and some contributions by the corporate world in the area of human rights,
at the same time it has to be stated that the implementation of human rights is not
yet where it should be. The vast majority of humans still suffer violations of their
human rights. Human rights are still a minority phenomenon, with only a minority
of humans enjoying full or partial human rights. Difficulties in implementing human
rights raise the question as to whether human rights are achievable at all, if they are
not just abstract ideals and if some of them should be given up.
The status quo of the implementation of human rights does not keep up with the
universality of human rights. Of course this does not mean that human rights are not
universal because the universality of human rights can be justified in the moral
dimension of human rights4 and therefore is still valid. But it indicates that there
exists urgency and necessity to enhance the implementation of human rights as law
which is not fully implemented and respected and, in reality, could lose its concrete
significance and ultimately its legal impact – both leading to its decomposition. This
is hard to imagine in the case of human rights because human rights are a ‘Monument’
which cannot be overseen and passed without taking notice of it, but the necessity
of a better realization of human rights must be addressed.5
Beyond that, the human rights discourse entails criticisms challenging the uni-
versality of human rights which can be categorized according to Georg Lohmann in
three groups6: The third challenge to the universality of human rights represents a
critical relativism based on scepticism related to the small potential of realization of
human rights and differences within this potential between the three categories of
human rights. This criticism introduces the failure of implementation of certain
human rights as a reason for cancelling them out of the human rights-catalogue.
Again, of course it can be justified why this cancellation cannot be pursued, but the
idea itself should already be taken as a serious warning sign and as indication of
existing problems.

See Kirchschlaeger 2013b.
See Kirchschlaeger 2013a.
See Joas 2011, p. 280.
Lohmann 2008, pp. 218–228.
62 P.G. Kirchschlaeger

A positive change in favour of human rights lies especially in the hands of

those – be it state or non-state actors – with greater power and influence. Therefore
one could argue that in this case the end – the realization of human rights – would
justify the means – nudging of humans by states.

5.3 Human Rights Protecting the Individual as Part

of a Community

This nudging though would have to respect human rights because human rights
protect all humans as individuals from all violations of essential elements and
spheres of human existence necessary for survival and for a life as a human, and
they are categorical. The latter means that – as mentioned above – no one needs to
fulfil any conditions to be entitled to human rights (being human is enough). Every
human is entitled with human rights which means that the nudging by a state must
respect human rights.
The bearers of human rights are individuals, not collectives which however does
not make them ‘individualistic’. In distinction from ‘individualistic’ indicating that
the individual rules, human rights are individual rights which embrace correspond-
ing duties as they are not exclusive rights but human rights. As one shares them with
all other humans, these rights come with the corresponding duties to contribute to
the realization of human rights of all other humans. Human rights and correspond-
ing duties form an asymmetric relationship because due to the categorical nature of
human rights mentioned above, entitlement to human rights does not depend on
fulfilment of the corresponding duties. “These rights are not forfeited by wrong
behaviour. Men and women do not have to prove themselves worthy of being granted
human rights.”7 In addition, one must consider Article 29 of the Universal
Declaration of Human Rights of 1948 as well: “Everyone has duties to the commu-
nity in which alone the free and full development of his personality is possible.” It
states that the individual as holder of human rights is embedded in a collective
which she/he needs for her/his own development and for which she/he should take
care as well. Finally rights possess always a social rather than an ‘individualistic’
component, as it has to be part of a social system because without at least a second
individual there would be no need for law.
But is it even legitimate for human rights to entail corresponding duties and
obligations? Is it justified that all humans are holders of human rights which can
limit one’s personal freedom e. g. because every right-holder needs to respect the
human rights of all other humans as well? Do the specific human rights that con-
cretely limit an individual’s actions possess any legitimacy? These questions show
the necessity of a justification of human rights as well.

Wolbert 2003, p. 176.
5 To What Extent Should the State Protect Human Beings from Themselves? 63

5.4 Autonomy: The Nucleus of Human Rights

Human rights need a moral justification in order to prove their legitimacy and in
order to remain coherent with their own core concept of the autonomy of the indi-
vidual because autonomy embraces the claim to know the reason why one’s free-
dom should be restricted by human rights. These challenges lead to the question
how human rights can be justified. Every human merits a justification why she/he is
a right-holder and also a duty-bearer of human rights as the responsibility corre-
sponding to human rights to respect the human rights of all other humans limits her/
his freedom.8 Robert Alexy ties the existence of human rights exclusively to the
possibility of their justification.9 Emphasizing the status quo of human rights as a
historical, political, and legal consensus enjoying global acceptance does not satisfy
as a justification either due to its descriptive and not normative nature or due to the
particular origin of historical, political, and legal consensus in general.
The relevance of this question how to justify human rights grows even further
when e.g. there are attempts to exclude a specific group of humans from human
rights in general or from some rights, when human rights in general are neglected,
when some rights are denied, or when nudging by a state could violate specific
human rights or disrespect human rights in general. Facing these realities, reasons
justifying human rights are necessary. An ethical model of justification of human
rights which is based on the principle of vulnerability contributes to meet this chal-
lenge and shows that human rights can be justified.10 The justification-model based
on the principle of vulnerability11 represents an approach ex negativo and starts with
a human being’s self-awareness of his/her own vulnerability. The awareness-
building of one’s own vulnerability is a self-recognition-process, whose empirical
correctness is not relevant. It is crucial though that humans are willing to do some-
thing about this awareness, namely to protect themselves from vulnerability or to
find a reasonable way to deal with it. During this awareness-building-process a
human being recognizes ex negativo the “first-person-perspective”.12 Human beings
understand that they make and interpret this basic anthropological situation of vul-
nerability as a subject (meaning as the first person singular) who is acting, deciding,
and suffering, and is living as a human being. “By acting and suffering, human
beings experience themselves as living beings, which are not just living like all
other beings but which live by living their own lives. To relate to themselves, to act
neither compulsively nor arbitrarily but to be guided by reasons and to pursue
freely-chosen purposes constitutes the life form which connects them with all
human beings as their own kind. This life form makes them vulnerable as the self-
relation which is inherent in this life form depends on fundamental conditions for

See Kirchschlaeger 2007, pp. 55–64.
Alexy 1998, pp. 244–264.
See Kirchschlaeger 2013a.
See Kirchschlaeger 2013a, pp. 231–335.
Runggaldier 2003, pp. 143–221.
64 P.G. Kirchschlaeger

realization”.13 During the awareness-building-process of one’s own vulnerability

and of one’s own “first-person-perspective”, human beings put themselves in a rela-
tionship with the self and with all other human beings, recognizing their own vul-
nerability and the vulnerability of all other human beings. They become equal in
terms of vulnerability; they make themselves equal. Human beings share this
During the awareness-building-process of their own vulnerability and of their
own “first-person-perspective”, human beings understand that they share not only
the vulnerability with other humans but also the individual “first-person-perspective”
on their individual vulnerability and on the vulnerability of all humans, and the
individual self-relation that every human is the subject of his/her life. They recog-
nize not only the equality of all humans in their vulnerability, but also in the
“first-person-perspective” and “self-relation”. They realize that the “first-person-
perspective” and “self-relation” are prerequisites for human life.
Based on this entire multi-layer awareness-building-process leading to recogni-
tion of the vulnerability of the “first-person-perspective” and the “self-relation”
shared with all human beings, a human is prepared to entitle all humans and him/
herself to human rights in order to reduce vulnerability and to actively balance and
compensate in the case of a transformation from vulnerability to violations.
Beyond the ethical justification of human rights based on the principle of vulner-
ability, human rights protect the autonomy of the individuals also by specific human
rights enabling the participation of the individual in political opinion-forming and
decision-making-processes14 which could be at risk because of nudging by states, e.
g. with article 19 (right to freedom of opinion and expression), with article 20 (right
to freedom of peaceful assembly and association), and with article 21 of the
Universal Declaration of Human Rights of 1948, also called the ‘democracy
1. Everyone has the right to take part in the government of his country, directly or
through freely chosen representatives.
2. Everyone has the right of equal access to public service in his country.
3. The will of the people shall be the basis of the authority of government; this shall
be expressed in periodic and genuine elections, which shall be by universal and
equal suffrage and shall be held by secret vote or by equivalent free voting
Democracy can also be seen as the institutional expression of the respect of the
autonomy of the individual giving the individual the possibility to participate in the
opinion-forming- and decision-making-process of the legal system she/he is living
in as a citizen. Human rights and democracy go hand in hand as democracy is the
political system which is embodying the autonomy of the individual inherent within
the idea and concept of human rights. Therefore the respect of human rights is part
of a democratic system. Consequently a democracy must integrate mechanisms
which ensure that human rights are respected respectively which exclude elements

Honnefelder 2012, pp. 171–172.
See Kirchschlaeger 2014, pp. 112–125.
5 To What Extent Should the State Protect Human Beings from Themselves? 65

violating human rights regarding the access to democratic opinion-forming- and

decision-making processes and in the way these processes are taking place, and that
the possibility of a democratic decision violating human rights is impossible from
the start. Human rights are the frame of reference for a democracy. Different forms
to guarantee the respect of human rights within a democratic system are known,
such as the Constitutional or the Supreme Court and need to be established in
democracies without such institutional measures in coherence with human rights
and in order to further the realization of human rights.
Human rights may open a global horizon but they start locally at the same time.
In 1958 Eleanor Roosevelt said, on the occasion of the tenth anniversary of the
Universal Declaration of Human Rights of 1948:
Where, after all, do universal human rights begin? In small places, close to home – so close
and so small that they cannot be seen on any maps of the world. Yet they are the world of
the individual person; the neighbourhood he lives in; the school or college he attends; the
factory, farm or office where he works. Such are the places where every man, woman and
child seeks equal justice, equal opportunity, equal dignity without discrimination. Unless
these rights have meaning there, they have little meaning anywhere. Without concerned
citizen action to uphold them close to home, we shall look in vain for progress in the larger

5.5 Conclusion

From a human rights perspective, scepticism towards the approach of nudging

remains high, firstly as long as nudging crosses the legal and ethical frame of refer-
ence of human rights and secondly as long the state does not protect humans from
human rights violations and does not contribute to the realization of human rights.
The fundament for the first reason for scepticism entails the protection of autonomy
by human rights, the guarantee of participation by the individual in opinion-
forming- and decision-making-processes, and the character of human rights as indi-
vidual rights, not as rights of the collective. Nudging can end up in contradicting or
violating some of these elements and therefore it can be criticized from a human
rights perspective.
The basis for the second reason embodies the possibility that nudging contrib-
utes to the stopping and prevention of human rights violations and to the realization
of human rights. Nudging is justified from a human rights perspective if nudging
furthers the realization of human rights. Even in that case thought nudging must
respect human rights because of the principle of indivisibility of human rights which
entails that the entire human rights-catalogue must be respected, implemented, and
realized. All specific human rights must go hand in hand.
The concerns which provoke the call for nudging of humans by states would be
addressed from a human rights perspective, thirdly, with – if even possible – nudg-
ing within the legal and ethical frame of reference of human rights only and exclu-
sively if it serves the realization of human rights.
66 P.G. Kirchschlaeger

Fourthly, the response to these concerns would be the emphasis of the autonomy
of the individual protected by human rights.
Fifthly, these concerns would be answered by pointing out that the chal-
lenges provoking the call for nudging of humans by states could be met by over-
coming the naïve assumption that every human being is born democratic and with
the skills and competencies to participate in a democratic opinion-forming- and
decision-making-process, and by furthering human rights education. Thomas
Hammarberg emphasizes: “Educating citizens in their human rights creates an
informed society which in turn strengthens democracy”.15 In the occasion of an
expert-seminar during the preparation-process of the UN Declaration on Human
Rights Education and Training in Marrakech 2009,16 Navanethem Pillay, UN-High
Commissioner for Human Rights, outlined the expectations in human rights educa-
tion: “Human Rights Education is essential for the prevention of human rights
abuses, the promotion of non-discrimination, equality and sustainable development,
and the enhancement of people’s participation in democratic decision making pro-
cesses”. The fundamental role of human rights education is to empower citizens to
defend their own rights and those of others. “This empowerment constitutes an
important investment for the future, aimed at achieving a just society in which all
human rights of all persons are valued and respected”.17 The idea of ‘empower-
ment’ means the capability to determine one’s own present and future with self-
confidence and awareness of one’s own rights and – in solidarity – of the rights of
all humans and to participate actively in political processes. The UN Declaration on
Human Rights Education and Training (Article 2/2) is defining human rights educa-
tion and training as:
Human rights education and training comprises all educational, training, information,
awareness-raising and learning activities aimed at promoting universal respect for and
observance of all human rights and fundamental freedoms and thus contributing, inter alia,
to the prevention of human rights violations and abuses by providing persons with knowl-
edge, skills and understanding and developing their attitudes and behaviours, to empower
them to contribute to the building and promotion of a universal culture of human rights.
(a) Education about human rights, which includes providing knowledge and understand-
ing of human rights norms and principles, the values that underpin them and the mech-
anisms for their protection;
(b) Education through human rights, which includes learning and teaching in a way that
respects the rights of both educators and learners;
(c) Education for human rights, which includes empowering persons to enjoy and exercise
their rights and to respect and uphold the rights of others.
From a human rights perspective, it is necessary to choose alternatives to nudg-
ing of humans by states if nudging is violating human rights. And there are alterna-
tives to nudging.

Hammarberg 2008.
The author has contributed as a consultative expert to the development of the UN Declaration on
Human Rights Education and Training during the entire preparation process of the Declaration.
DeMello 2004, p. 3.
5 To What Extent Should the State Protect Human Beings from Themselves? 67


Alexy, Robert. 1998. Die Institutionalisierung der Menschenrechte im demokratischen

Verfassungsstaat. In Philosophie der Menschenrechte, ed. Stefan Gosepath and Georg
Lohmann, 244–264. Frankfurt a. Main: Suhrkamp Verlag.
De Mello Vieira, Sérgio. 2004. Foreword. ABC teaching human rights, ed. UN. New York:
Hammarberg, Thomas. 2008. Viewpoint 06/10/2008. Strasbourg. Online:
Honnefelder, Ludger. 2012. Theologische und metaphysische Menschenrechtsbegründungen. In
Menschenrechte: Ein interdisziplinäres Handbuch, ed. Arnd Pollmann and Georg Lohmann,
171–178. Stuttgart: J.B. Metzler.
Joas, Hans. 2011. Die Sakralität der Person. Eine neue Genealogie der Menschenrechte. Frankfurt
am Main: Suhrkamp.
Kirchschläger, Peter G. 2007. Brauchen die Menschenrechte eine (moralische) Begruendung? In
Menschenrechte und Kinder, Internationales Menschenrechtsforum Luzern (IHRF) Bd. IV, ed.
Peter G. Kirchschlaeger, Thomas Kirchschlaeger, Andrea Belliger, and David J. Krieger,
55–64. Bern: Staempfli Verlag.
Kirchschläger, Peter G. 2013a. Wie können Menschenrechte begründet werden? Ein für religiöse
und säkulare Menschenrechtskonzeptionen anschlussfähiger Ansatz, ReligionsRecht im Dialog
Bd. 15. Münster: LIT.
Kirchschläger, Peter G. 2013b. Die Multidimensionalität der Menschenrechte – Chance oder
Gefahr für den universellen Menschenrechtsschutz? MenschenRechtsMagazin 18(2): 77–95.
Kirchschläger, Peter G. 2014. The relation between democracy and human rights. In Globalistics
and globalization studies: Aspects & dimensions of global views, ed. Leonid E. Grinin, Ilya
V. Ilyin, and Andrey V. Korotayev, 112–125. Volgograd: Uchitel Publishing House.
Lohmann, Georg. 2008. Zu einer relativen Begründung der Universalisierung der Menschenrechte.
In Gelten Menschenrechte universal? Begruendungen und Infragestellungen, ed. Günter
Nooke, Georg Lohmann, and Gerhard Wahlers, 218–228. Freiburg i. B: Herder.
Moyn, Samuel. 2010. The last Utopia. Human rights in history. Cambridge: Harvard University
Runggaldier, Edmund. 2003. Deutung menschlicherGrunderfahrungen im Hinblick auf unser
Selbst. In Unser Selbst – Identität im Wandel neuraler Prozesse, ed. Günter Rager, Josef
Quitterer, and Edmund Runggaldier, 143–221. Paderborn: Schöningh.
Sunstein, Cass R., and Richard Thaler. 2008. Nudge – Improving decisions about health, wealth,
and happiness. New Haven: Yale University Press.
Wolbert, Werner. 2003. Menschenwürde, Menschenrechte und Theologie. Salzburger Theologische
Zeitschrift 7: 161–179.
Chapter 6
Nudging Is Judging: The Inevitability of Value
Consequences of the Collapse of the Fact/Value
Dichotomy for Behavioural Law and Economics

Ariel David Steffen

Abstract This article shows that there can never be nudging without judging.
Specifically, the essay analyses the impact of the collapse of the fact/value dichot-
omy on rational choice and behavioural law and economics. Not only does a nudge
require an ex-ante value judgment, but facts and values are almost always inextrica-
bly entangled. Thus, the real problem does not lie in “getting the facts right” (i.e.
judging people as neutrally as possible), but in the so-called “facts” themselves,
which already contain value judgments. In the context of rational choice and behav-
ioural law and economics, the term ‘rational’ is at the same time used in its positive
meaning (how things are) and in its normative meaning (how things ought to be). As
a result, researchers in BLE engage in both positive and normative science. Instead
of accepting the normative and making it explicit, however, it is often shunned by
researchers. In doing so, an unscientific double standard is maintained with the
explicit positive science in the fore and the implicit normative science looming in
the background. As a result, the normative concept ‘rationality’ is maximized under
the guise of it being a positive concept while ‘autonomy’ is rejected as a legitimate
concept for maximization on grounds of it being normatively laden. Pragmatic eth-
ics could serve as a basis to make the normative transparent while at the same time
not exclude it from scientific discourse.

A shortened, more concise version of this essay is scheduled to be published in German at Nomos
Verlag in 2016.
A.D. Steffen, Ph.D. Candidate (*)
Faculty of Law, University of Lucerne, Frohburgstrasse 3, 6002 Lucerne, Switzerland

© Springer International Publishing Switzerland 2016 69

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_6
70 A.D. Steffen

6.1 Introduction

It was not until what Vivian Walsh calls the “second phase” of the “revival of clas-
sical [economic] theory during the twentieth century”, heralded by Amartya Sen in
the late 1960s1 and developing into the Capabilities Approach during the 1980s, that
the simplified view of the unmitigated self-interested and unboundedly rational
homo oeconomicus maximising his or her individual utility came under attack.2 The
“first phase” was characterized by a simplistic analysis of Adam Smith’s An Inquiry
into the Nature and Causes of the Wealth of Nations stripping it of its ethical back-
ground.3 As Sen succinctly puts it:
Indeed, it is precisely the narrowing of the broad Smithian view of human beings, in mod-
ern economics, that can be seen as one of the major deficiencies of modern economic the-
ory. This impoverishment is closely related to the distancing of economics from ethics.4

The wrong and narrow view that Smithian economics can be distilled down to
the principle of the invisible hand and its tripartite constituents – enlightened self-
interest, limited government, and a solid currency in a free-market economy – is
still widespread, both in academia where it is taught in most, if not all, introductory
courses to economics as well as in the common knowledge of the general public.5
Roughly at the same time as Sen, Herbert A. Simon objected to this concept of
perfect rationality arguing that “the decision-making process in problem situations
conforms closely to the models of bounded rationality”.6 According to Simon,
“rationality is bounded when it falls short of omniscience”, that is to say, whenever
there is imperfect information in regard to alternatives, relevant exogenous events
or, more generally, whenever we are unable to properly calculate consequences.7
This criticism is fundamental: it is directed against the assumption of perfectly
rational behaviour. Taking up Simon’s critique, Daniel Kahneman and Amos
Tversky published Prospect Theory (1979), a behavioural economic theory based
on empirical research describing how individuals choose between probabilistic
alternatives that involve risk, where the probabilities of the outcomes are known.
The theory further weakened the assumption of perfect rationality as Kahneman and
Tversky found that people are generally loss averse and hence fail to maximise their
expected utility, as fully rational individuals would do, when risk is involved.8 Since

For an early critique of rational choice theory see for example Sen 1967.
Walsh 2000, p. 5.
Cf. Smith 1776/2008.
Sen 2000, p. 272.
See e.g.
(last accessed on 10 April 2015).
Simon 1979, p. 507 (emphasis added).
See Simon 1979, p. 502.
See Kahneman and Tversky 1979, pp. 263–292.
6 Nudging Is Judging: The Inevitability of Value Judgments 71

then, many empirical studies have independently found a myriad of other so-called
systematic cognitive heuristics, which may give rise to biased behaviour.9
In the wake of this development away from rational choice economics to behav-
ioural economics, a new branch of law and economics – called behavioural law and
economics – emerged. Ultimately, in so far as the aim of behavioural law and eco-
nomics is to “influence the choices of affected parties in a way that will make choos-
ers better off”, it is paternalistic.10 This sort of paternalism has been called ‘soft’ or
‘libertarian’ because circumventing heuristics and biases is achieved by means of a
‘nudge’, which is defined as “an aspect of choice architecture that alters people’s
behaviour in a predictable way without forbidding any options or significantly
changing their economic incentives”.11 Much has been argued in favour of nudging
as well as in its disfavour.12 Two issues lie at the very core of the critique of
• The critique that nudging theorists derive an ‘ought’ from an ‘is’, which is in
violation of Hume’s Law.
• The critique that for nudging theorists, deviating from perfect rationality justifies
the implementation of a nudge, the aim of which it is to re-install perfect ratio-
nality. Other possibly relevant parameters, such as autonomy (from Ancient
Greek αὐτο-νομία: self-governing), are neglected.13
The main criticism can be boiled down to accusing nudging of judging. It tells
people what is ‘good’ and what is ‘bad’ for them. In other words, it tells people that
perfect rationality is desirable and ought to be encouraged while bounded rational-
ity is undesirable and ought to be discouraged. Since autonomy entails bounded

For an overview of the most important heuristics and biases, see Mathis and Steffen 2015,
pp. 31–48. For further details on heuristics and biases, see Schweizer 2005; Rachlinski 2000,
pp. 95–115; Sunstein 2002, pp. 61–107; Jolls and Sunstein 2006, pp. 199–241; Thaler and Sunstein
2008; Kahneman and Tversky 1981, pp. 453–458; Tversky and Kahneman 1974, pp. 1124–1131,
1991, p. 1039; Kahneman 1997, p. 105, 2003, p. 697, 2011; Tversky and Fox 1995, pp. 269–283;
Thomas and Millar 2011, pp. 139–149; Meyerowitz and Chaiken 1987, pp. 500–510; Kahneman
et al. 1990, pp. 1325–1348; Kahneman et al. 1991, pp. 193–206; Christensen-Szalanski and
Fobian-Willham 1991, pp. 147–168. For a general discussion of behavioural law and economics,
see Sunstein 2000; Arlen 1998, p. 1765; Issacharoff 1998, p. 1729; Jolls et al. 1998, pp. 1471–
1550; Kelman 1998, p. 1577; Korobkin and Ulen 2000, pp. 1051–1144; Parisi and Smith 2004;
Jolls 2009; Englerth 2007, pp. 60–130. For an overview of libertarian paternalism and the prob-
lems associated with it, see van Aaken 2015 (focussing on the legal limits of paternalistic nudges);
Mitchell 2005, pp. 1–42; Sunstein and Thaler 2003, pp. 1159–1202.
Sunstein and Thaler 2003, p. 1162. Note that Sunstein and Thaler do not elaborate on how the
nudger is reliably able to find out how the nudgee judges his or her choice. It is implied, however,
that the nudgee’s revealed preferences (outset) systematically deviate from his or her rational pref-
erences (goal). The rational preferences are supposed to be value-neutral and universally accepted,
derived from the allegedly equally value-neutral and universally accepted rational choice theory. In
other words, rational preferences serve as an axiom and are closed to debate.
Thaler and Sunstein 2008, p. 6 (emphasis added).
For arguments in favour of nudging, see for example Sunstein and Thaler 2003, pp. 1159–1202.
For arguments against nudging, see for example van Aaken 2015 and Mitchell 2005, pp. 1–42.
See van Aaken 2015.
72 A.D. Steffen

rationality, nudging can never truly propagate autonomy. Furthermore, nudging

presupposes a value-neutral, universal and complete theory of rationality and of
‘good’ (perfect rationality) and ‘evil’ (bounded rationality).14 Proponents of nudg-
ing seem to imply that, since libertarian paternalism focusses on systematic heuris-
tics and biases only, it is indeed value-neutral and universal. In this respect, they use
the term ‘rational’ in its positive sense (they make the empirical observation that
people systematically deviate from perfect rationality). At the same time, however,
they make a value-judgment when they differentiate between ‘good’ (perfect) ratio-
nality and ‘evil’ (bounded) rationality. In this respect, they use the term ‘rational’ in
its normative sense.
If we break this down to the level of preferences, nudging theorists seem to argue
that people’s revealed (boundedly rational) preferences are systematic deviations
from their (perfectly) rational preferences, the latter of which being, on the one
hand, value-neutral and universal aims of both individuals and society and, on the
other hand, inherently ‘good’.15 This seems illogical and indeed it is. In fact, the
nudging theorists’ claim is untenable due to the entanglement of fact and value.
Rational choice theory is neither value-neutral, nor universal, nor complete. As a
result, it is a conscious non-value-neutral choice to prefer perfect rationality to
bounded rationality or, in other words, to prefer (perfect) rationality to (boundedly
rational) autonomy based on the idea that rational choice is faultless whereas auton-
omous choice is faulty. Hence, the argument goes, everybody can be made better off
(in terms of both economic efficiency and personal preference) when rationality
rather than autonomy is maximised. I shall argue that this argument rests on shaky
ground. Not only will it be hard, many times impossible, to unequivocally deter-
mine the (most) perfect rational choice (and doing so may lead to high cost), but
autonomy in itself might be inherently important and valuable to people, even if it
leads to them making rationally ‘wrong’ choices.16 In sum, people might be better
off choosing wrongly but autonomously than choosing rightly but heteronomously.
The name 'liberal paternalism' seems to suggest that it is a form or liberal-
ism. However, in liberalism it has long been acknowledged that individuals know
best what is best for them. We should not forget that autonomy is one of the fore-
most liberties. Jeopardising autonomy may lead to unintended effects. To name just
a few, preventing mistakes may lead to stultification and foregone missed learning
effects, as von Humboldt already pointed out in the eighteenth century.17
Furthermore, the multitude of lifestyles arising from autonomy via originality and

Note how this is closely connected with the ‘naturalistic fallacy’, which claims that it is falla-
cious to explain e.g. the concept of ‘goodness’ reductively in terms of natural properties, such as
‘being pleasant’ or ‘being desirable’. For further information on the naturalistic fallacy, see George
Edward Moore, Principia Ethica, 1903, §10.
See for example the complete transcript of the BBC HARD talk interview with Richard Thaler,
held in October 2012, accessible via
thaler-behavioural-economist.html (last accessed on 10 April 2015).
See van Aaken 2015.
See von Humboldt 1792/1989.
6 Nudging Is Judging: The Inevitability of Value Judgments 73

creativity may be important for the wellbeing of both individuals and society (free-
dom of choice of lifestyle may lead, via innovative creativity, to new forms of liv-
ing, which may be better adapted to new living conditions on an individual level,
which may promote innovation on the level of society).18
In the following, I will elaborate – from a philosophical point of view – why
behavioural law and economics is not the long-yearned-for value-neutral doctrine
of salvation it purports to be. More specifically, after an introduction to the fact/
value dichotomy (6.2), I will analyse Putnam’s critique of rational choice theory
and its repercussions for behavioural (law and) economics (6.3). This is followed by
concluding remarks on nudging and judging (6.4). I will close with an outlook (6.5).

6.2 The Fact/Value Dichotomy

Although following almost directly from W. V. O. Quine’s critique of the analytic/

synthetic dichotomy in his eminent essay Two Dogmas of Empiricism (1951), and
although agreed upon by most, if not all, pragmatists, the view that fact and value
are inextricably entangled is one that remains contested by the majority of the less
analytically-minded, be it by researchers in philosophy, other subjects, or by the
general public. The most prominent account of the collapse of the fact/value dichot-
omy has hitherto been given by Hilary Putnam.19 According to Putnam, the fact/
value dichotomy and the analytic/synthetic dichotomy made up the very basis of
classical empiricism and logical positivism. To abandon these dichotomies, there-
fore, means treading new ground. Or, to put it in Star Trek terminology: to boldly
go where no man has gone before. As Putnam puts it:
[T]he fact/value dichotomy (“is” versus “ought”) and the analytic-synthetic [sic] dichot-
omy (“matters of fact” versus “relations of ideas”), was foundational for classical empiri-
cism as well as for its twentieth-century daughter, logical positivism. Thus to come to think
without these dogmas is to enter upon a genuine “post-modernism” – to enter a whole new
field of intellectual possibilities in every important area of culture.20

I will now give a brief overview of how Putnam arrived at his reasoning. First, I
will introduce the reader to the basic concept of the analytic/synthetic and fact/value
dichotomy, with which the validity of (classical) empiricism stands and falls. A
special emphasis is laid on Hume’s Law. Also, logical positivism as its heir will be
discussed. Second, Quine’s collapse of the analytic/synthetic dichotomy will be
analysed. Indeed, it is a journey with far-reaching consequences we will embark on,
but one that is worthwhile partaking of. After all, if we truly believe in progress in
the way the natural sciences do, we cannot choose but abandon our old idées reçues
grounded in metaphysics. We do indeed have to boldly go where only few have
dared to go before.

See Mill 1859/1869.
See Putnam 2002/2004.
Putnam 2002/2004, p. 9.
74 A.D. Steffen

6.2.1 From Empiricism and Hume’s Law to Logical Positivism

As researchers and especially as scientists, it is unlikely that there is a more devas-

tating reproach than when someone insinuates that one’s research includes, or even
worse is based on, undeclared value judgments. It is deeply engrained in both our
ordinary as well as scientific mind and language that what is meant by that reproach
is that one’s research is subjective rather than objective. It is ‘unscientific’ and there-
fore, at least from an academic point of view, nonsense.21
At the very heart of this reasoning lies Immanuel Kant’s distinction of analytic
truths – true by virtue of their meaning/definition (e.g. oncologists are doctors) –
and non-analytic truths, which Kant termed synthetic truths – true only via infer-
ence (e.g. oncologists are rich). In other words, whenever the subject-concept
(oncologist) contains the predicate-concept (to be a doctor), a proposition is ana-
lytic. Whenever the subject-concept (oncologist) does not include the predicate-
concept (to be rich), it is synthetic.22 Even before Kant, David Hume stated that an
‘ought’ cannot be inferred from an ‘is’ (today known as Hume’s Law).23 In other
words, when we are confronted with positive facts (‘is’), it is invalid to derive from
them normative values (‘oughts’). Anticipating Kant’s analytic/synthetic distinc-
tion, “Hume assumed a metaphysical dichotomy between ‘matters of fact’ and
‘relations of ideas’”.24 The problem with this is that Hume’s Law stands and falls
with his metaphysics of ‘ideas’, which he seems to have adapted from John Locke’s
concept of ‘perceptions’. In An Essay Concerning Human Understanding (1689),
Locke writes that perceptions are “whatsoever the mind perceives in itself, or is the
immediate object of perception, thought or understanding”.25 Hume divided these
perceptions into ‘impressions’ and ‘ideas’, the latter being “faint copies” of impres-
sions “less forcible and lively”.26 Thus, ‘ideas’ only resemble the actual thing, which
is apprehended via ‘perception’ and which induces a ‘sentiment’, or an emotion.27
According to Hume, other ‘facts’ do not exist. Using the example ‘crime’, Hume
Where is that matter of fact that we call crime; point it out; determine the time of its exis-
tence; describe its essence or nature; explain the sense or faculty to which it discovers itself.
It resides in the mind of the person who is ungrateful. […] [T]he crime of ingratitude is not
any particular fact; but arises from a complication of circumstances which being presented

See Putnam 2002/2004, p. 7.
See Kant 1781/1998, A6-7/B10-11. Immanuel Kant further distinguishes between a priori propo-
sitions (propositions whose justification do not rely on experience, such as “2+2=4”) and a poste-
riori propositions (propositions whose justification rely on experience, such as “chairs exist”).
Cf. Hume 1738/1978, pp. 469–470.
Putnam 2002/2004, p. 14.
Locke 1689/2008, II.viii.8.
Hume 1748/2008, II.3–4.
Hume 1748/2008, II.1.
6 Nudging Is Judging: The Inevitability of Value Judgments 75

to the spectator excites the sentiment of blame, by the particular structure and fabric of his

We may thus conclude that by Hume’s standards, judgments of vice and virtue
are not facts, but sentiments: no one can “point out” something that “resides in the
mind”. Per analogiam, any utterances containing normative concepts – such as
‘good’, ‘bad’, ‘right’, ‘wrong’, etc. – are emotions and not facts.
Indeed, this is exactly what the logical positivists did: they built on Hume’s (and
Kant’s) concept that ethical judgments are not facts, and hence are non-scientific
nonsense. As Rudolph Carnap puts it in The Unity of Science (1934):
All statements belonging to Metaphysics, regulative Ethics, and (metaphysical)
Epistemology have this defect [namely, that they are “pseudoproblems” barred from ratio-
nal discourse – AS], are in fact unverifiable and, therefore, unscientific. In the Vienna
Circle we are accustomed to call such statements nonsense. […] [Such] [c]onceptions can
be associated with any arbitrarily compounded series of words.29

Putnam criticises Carnap’s conclusion, on the one hand, because it precludes

(non-metaphysical) epistemological value judgments used in science on a daily
basis (such as ‘coherence’ and ‘plausibility’) and, on the other hand, because it turns
ethics into an irrational, unscientific subject on grounds of inflating a mere (useful)
distinction between ethical judgments and other sorts of judgments into a full-
grown metaphysical dichotomy (metaphysical because facts are claimed to exist
while ethics is claimed to be an “arbitrarily compounded series of words”).30
According to Putnam, to make the step from a mere distinction to a dichotomy is the
decisive element of logical positivism. This step, however, is invalid. At best,
Putnam says, “the fact/value dichotomy is […] not a distinction [in the sense of an
either/or dualism – AS] but a thesis, namely the thesis that ‘ethics’ is not about
‘matters of fact’”.31 The decisive element in the positivists reasoning was that they
pretended to know what a ‘fact’ was:
According to the positivists, in order to be knowledge, ethical ‘sentences’ would have to be
either analytic, which they manifestly are not, or else ‘factual’. And their confidence that
they could not be factual, just like Hume’s confidence that ‘the crime of ingratitude is not
any particular fact,’ derived from their confidence that they knew exactly what a fact was.32

However, logical positivists failed to clearly define the concept of ‘fact’.

Relativity theory and quantum physics radically changed what might be considered
‘fact’ and what not. Certainly it is no longer tenable to adhere to the definition of
fact as sense impression when facing curved space-time.33 Although Carnap revised
his theory in Foundations of Logic and Mathematics (1938) – so that the system of
scientific statements as a whole had factual content – he still held that the language

Hume 1777/1975, pp. 287–288 (original emphasis).
Carnap 1934, p. 22.
Putnam 2002/2004, p. 19, p. 31.
Putnam 2002/2004, p. 19 (original emphasis).
Putnam 2002/2004, p. 21 (original emphasis).
See Putnam 2002/2004, p. 22.
76 A.D. Steffen

of science, in order to be cognitively meaningful, ought to resemble the language of

physics (i.e. ‘factual’ language must at least be reducible to observation terms).34
Although ultimately untenable, this narrow definition of ‘scientific language’ rely-
ing on a narrow definition of ‘fact’, according to Putnam, was a great mistake that
had far-reaching implications up to the present. As Putnam concisely summarises:
The logical positivist fact/value dichotomy was defended on the basis of a narrowly scien-
tistic picture of what a ‘fact’ might be, just as the Humean ancestor of that distinction was
defended upon the basis of a narrow empiricist psychology of ‘ideas’ and ‘impressions’.
The realization that so much of our language is a living counterexample to both (classical
empiricist and logical positivist) pictures of the realm of ‘fact’ ought to shake the confi-
dence of anyone who supposes that there is a notion of fact that contrasts neatly and abso-
lutely with the notion of value. […] A deeper problem is that, from Hume on, […] many
[…] failed to appreciate the ways in which factual description and valuation can and must
be entangled.35

After a brief overview of how Quine let the analytic/synthetic dichotomy col-
lapse (namely, by showing that scientific statements conflate ‘conventions’ and
‘facts’), we will examine Putnam’s argumentation more closely and in doing so see
how “factual description and valuation” are indeed often inextricably “entangled”.36

6.2.2 The Collapse of the Analytic/Synthetic Dichotomy

Following W. V. O. Quine

According to Putnam, “it is widely recognized since Quine’s 1951 attack on this
overblown form of the analytic-synthetic dichotomy that it has collapsed”.37 In
fact, in 1951 Quine showed in his eminent essay Two Dogmas of Empiricism that
there are two essential, but ill-founded, dogmata in modern empiricism: one is the
belief in a clear-cut cleavage between analytic truths and synthetic truths, the other
is the belief that meaningful statements are reducible to logical constructs referring
to immediate experience.38 Quine argues that abandoning these dogmata results in
a blurring between the boundary of ‘soft’ metaphysical – in the sense of non-phys-
ical, thus allegedly speculative – sciences (such as liberal arts and moral sciences),
and ‘hard’ fact-based natural science.39 This leads, as Quine argues, to a shift away
from empiricism and towards pragmatism.40 Quine stresses that his analysis is not
limited to ordinary language, but that it extends to scientific language and even
holds true for precise constructed and formal languages (such as logic) with explicit

See Putnam 2002/2004, pp. 23–25.
Putnam 2002/2004, pp. 26–27 (original emphasis).
Putnam 2002/2004, pp. 26–27.
Putnam 2002/2004, p. 8.
Quine 1951, p. 20.
Quine 1951, p. 20.
Quine 1951, p. 20.
6 Nudging Is Judging: The Inevitability of Value Judgments 77

rules for interpretation since these rules, in order to reveal analytic statements,
must presuppose the existence, or at least the notion, of analyticity.41 As Quine suc-
cinctly puts it:
[T]ruth in general depends on both language and extralinguistic fact. […] Hence the temp-
tation to suppose in general that the truth of a statement is somehow analyzable into a lin-
guistic component and a factual component. Given this supposition, it next seems reasonable
that in some statements the factual component should be null; and these are the analytic
statements. But, for all it’s a priori reasonableness, a boundary between analytic and syn-
thetic statements simply has not been drawn. That there is such a distinction to be drawn at
all is an unempirical dogma of empiricists, a metaphysical article of faith.42

Quine’s reasoning has a profound impact on the logical positivists’ verification

theory of meaning, which contends that only statements about the world that are
empirically verifiable (or, in the case of analytic statements, logically necessary) are
cognitively meaningful. Consequently, liberal arts and moral sciences – in fact, all
of science concerned with evaluative judgements – are cognitively meaningless,
carry no objective truth value and hence are but an “arbitrarily compounded series
of words”.43 In short, they lose their raison d’être as science.
As far as the dogma of reductionism is concerned, Quine argues that it is “inti-
mately connected with the other dogma”.44 According to Quine, the focus on indi-
vidual statements is too narrow. It is impossible that an individual statement can,
“taken in isolation from its fellows, admit of confirmation or information at all”.45
Statements about the factual world, Quine says, are never individually confirmed
via sense experience but always as a cluster, a “corporate body” as he calls it.46
Quine goes even further and argues that “[t]he unit of empirical significance is the
whole of science.”47 If this reasoning is sound, then it makes no sense to speak of
analytic statements at all. To summarize in Quine’s own words:
The totality of our so-called knowledge or beliefs, from the most casual matters of geogra-
phy and history to the profoundest laws of atomic physics or even pure mathematics and
logic, is a man-made fabric which impinges on experience only along the edges. […] A
conflict with experience at the periphery occasions readjustments in the interior of the field.
[…] Re-evaluation of some statements entails re-evaluation of others, because of their logi-
cal interconnections […]. But the total field is so undetermined [sic] by its boundary condi-
tions, experience, that there is much latitude of choice as to what statements to re-evaluate
in the light of any single contrary experience. […] [I]t becomes folly to seek a boundary
between synthetic statements, which hold contingently on experience, and analytic state-
ments, which hold what may. Any statement can be held true come what may, if we make

Quine 1951, pp. 33–34.
Quine 1951, p. 34.
Carnap 1934, p. 22.
Quine 1951, p. 38.
Quine 1951, p. 38.
Quine 1951, p. 38.
Quine 1951, p. 39.
78 A.D. Steffen

drastic enough adjustments elsewhere in the system. […] Conversely, by the same token, no
statement [including basic logic laws – AS] is immune to revision.48

Carnap and the logical positivists realized early on that different categories for
ontological issues and scientific hypotheses can only be upheld by postulating a
clear and “absolute distinction between the analytic and the synthetic”.49 In other
words, they make a category mistake (they treat the analytic and the synthetic as if
they were two distinct categories, when in effect they belong to the same category,
namely myths about reality – the two concepts differ only in degree, not in kind). In
a similar vein, Jürgen Habermas, Karl-Otto Apel and Robert Alexy realized that
different categories for objective normative truths and subjective normative claims
can only be upheld by postulating a clear and absolute distinction between facts and
values. The latter has to undergo the scrutinizing trial of discourse in order to
become the former.50 It is a vain if heroic attempt to save the concept of universal,
absolute truth by clinging on to the dogma of a clear-cut cleavage between the fac-
tual and the counter-factual. Habermas, Apel and Alexy’s method prescribes that a
set of rules, so-called Diskursregeln, be introduced in order to realize an ideal
speech situation.51 Thus, Habermas, Apel and Alexy stumble into the same pitfall as
Carnap and the logical positivists have before them: both fail to see that the rules
they introduce already presuppose (and hence cannot prove) the existence, or at
least the notion, of – in the case of Carnap and the positivists – analyticity and – in
the case of Habermas and the discourse analysts – facticity. In addition, rules are
never void of values. In fact, they seem to be the very essence of values. It is of little
help to seek to objectify these first-order rules by introducing yet another set of
second-order rules which all first-order rules have to comply with. This would only
lead to an infinite regress, but not to new insight.
To sum up, the totality of science is underdetermined by experience. The rest is
fiction. However, this is not to say that these fictitious myths do not have their place
in real life. Quite on the contrary, they make up most of human reality and are
indeed very useful. But they are neither true ipso facto, nor are they in any way
reducible to sense experience, nor are they objectifiable through discourse. In sci-
ence, it is only natural that a myth lacking in some aspect is superseded by its supe-
rior cousin. All science is an infinite series of evolving myths, constantly adjusted
and readjusted to actual and very real current human needs and wants. Hence, if we
want to be upright scientists, be it in the natural sciences or in the liberal arts, we
must not cling to the overcome myth that a clear-cut distinction can be drawn
between the analytic and the synthetic; between facts and values; between right and
wrong; between living being and dead un-being (think Schrödinger’s cat). This is
not to say that there is no real difference between these states, but that this difference
is always a matter of degree (a gradual difference) and never a matter of kind

Quine 1951, pp. 39–40.
Quine 1951, p. 43.
See Habermas 1992/1998, 1992/1996, 1999. Also see Brune 2010.
For an overview see Ott 2001, pp. 165–166.
6 Nudging Is Judging: The Inevitability of Value Judgments 79

(a dichotomy). The latter is a clear category mistake. Furthermore, the difference,

being gradual, tends to be fuzzy rather than discrete. This means that there is always
uncertainty as to where exactly, say, a given sentence is located in the ‘fact-value
spectrum’. As in the case of Schrödinger’s cat, it is often the case that we have no
other choice but to assume that both states are equally true: the cat is already dead
even while it is still alive and kicking. Put differently, the sentence “Kant was ratio-
nal” may equally be read positively/descriptively, namely Kant put his reason to
use, or normatively/prescriptively, meaning that it was good that Kant put his reason
to use.

6.3 A Critique of Rational Choice and Behavioural

(Law and) Economics

In 6.2, we saw that the analytic/synthetic dichotomy cannot be upheld and that, by
analogy, the fact/value dichotomy is equally doomed. In this chapter, I shall scruti-
nise rational choice theory, on which behavioural (law and) economics is based,
against the backdrop of our findings.
In the next section, we will see that rational choice theory is incomplete.52 It is
incomplete because of Gödel’s first incompleteness theorem which, broadly speak-
ing, states that there are statements in the language of a consistent formal system F
which can neither be proved nor disproved in F (they may still be true, however).53
Insofar as rational choice and behavioural (law and) economics affirm the fact/value
dichotomy, it logically follows that a statement in these systems must either be a
fact or a value, tertium non datur. However, statements that are at the same time
facts and values exist. These can be neither proved nor disproved in rational choice
or behavioural (law and) economics. The problem with complete and transitive
preferences, a central theorem in rational choice, will serve as an example. Thus,
these theories are incomplete. It follows that they are neither objective nor univer-
sal. Hence, the decision to maximise ‘rationality’ rather than any other parameter is
a conscious, non-value-neutral choice. As such, its pros and cons need to be care-
fully weighed against other conscious, non-value-neutral choices. This, however, is
usually neglected in the literature of behavioural (law and) economics.
In a second step, I will briefly show that maximising rationality is highly likely
to lead to a lower utility level than maximising autonomy. Further empirical research
needs to be carried out in this field.

An axiomatic theory T is said to be complete if, and only if, it is consistent and if for every sen-
tence σ in the language of T it holds true that either T ˫ σ (the sentence σ is provable in T) or T ˫ ¬
σ (the sentence not-σ is provable in T, i.e. the sentence σ is refutable in T).
For more information on the two Gödel incompleteness theorems, please refer to the respective
entry in the Stanford Encyclopedia of Philosophy, available at:
goedel-incompleteness/ (last accessed: 23.10.2015).
80 A.D. Steffen

Finally, we will see that ‘opting out’ is in fact a bogus claim. It is not the re-
instalment of autonomy it purports to be, but since the nudge leads to a transvalua-
tion of people’s preferences, deciding against the ‘socially desirable’ option by
opting out is bound to lead to an inner conflict, also known as ‘cognitive disso-
nance’ in social psychology. Psychologists agree, and it has been widely empiri-
cally confirmed, that people strive to reduce the negative emotions associated with
cognitive dissonance by avoiding dissonance whenever possible.54 As a result, only
very few people will ever opt out of a nudge. Again, further empirical research
needs to be undertaken in this subject.

6.3.1 Rational Choice and Behavioural (Law and) Economics

Are Incomplete

To follow in the wake of Vivian Walsh and Hilary Putnam, I maintain that rational
choice theory is incomplete.55 The basis for my reasoning is that logical positiv-
ism – particularly verificationism – is too narrow in claiming that statements must
be verifiable either on logical or on empirical grounds to carry meaning. Whereas
rational choice is based on logical positivism insofar as it endorses the viewpoint
that ‘rational’ is logically verifiable (analytic, in other words), behavioural econom-
ics is based on logical positivism insofar as it deems ‘rational’ to be empirically
verifiable (synthetic, in other words). Thus, rational choice and behavioural eco-
nomics affirm logical positivism as well as the analytic/synthetic and the fact/value
dichotomy implicit therein. Rational choice theory as well as behavioural econom-
ics claim that the sort of ‘rational’ they mean is strictly positive/descriptive rather
than normative/prescriptive. This, however, can only be maintained as long as we
believe in clear-cut dichotomy between positive/descriptive facts and normative/
prescriptive values. As we have seen, this seems implausible. It is much more plau-
sible to argue that the relationship between facts and values is one of a fuzzy, grad-
ual difference and that most of the time facts and values are entangled.
The incompleteness of rational choice and in its wake behavioural economics
becomes particularly clear when focussing on the rationality of preferences. Both
theories rely on the existence of completely rational agents. A completely rational
agent is characterised and defined by a complete and transitive preference ordering.
A preference ordering is complete and transitive, if and only if, the agent knows
whether she prefers a certain bundle of commodities to another or is indifferent
amongst them (completeness) and if her preference ordering is consistent, i.e. does
not contradict itself (transitivity).56 In other words, if it is true for all x, y, and z that

See for example Aronson et al. 2008, pp. 163 et seqq.
Cf. Putnam 2002/2004, pp. 79–95.
Cf. for example Mas-Collel 1982, pp. 72–82.
6 Nudging Is Judging: The Inevitability of Value Judgments 81

(where R here refers to the weak preference relation and may be read as is weakly
preferred to, ∨ means or, ∧ means and, and ⇒ means it follows):
1. xRy ∨ yRx (completeness)
2. xRy ∧ yRz ⇒ xRz (transitivity)
This, however, presupposes that the two commodity bundles are in fact compa-
rable, that is to say, that they belong to the same category. If the categories differ,
then the commodity bundles are incomparable. Notwithstanding, the decision
maker may still consider both. Let us look at a concrete example. Suppose Dr
Faustus is considering two ways of living which he holds to be both equally worthy
of pursuing and incommensurable: the religious life (z) and the sensual life. If he
chooses the sensual life, he would prefer to have a romantic relationship with the
intelligent Gretchen (x) rather than her less intelligent little sister, Betchen (y).
Thus, he prefers alternative x to alternative y (xRy). Since he has not made up his
mind whether he prefers either of the two alternatives (x or y) over the religious life
(z), however, he is yet indifferent between x and z as well as between y and z (more
accurately, he does neither weakly prefer x to z or z to x, nor y to z or z to y). In
other words, for all x, y, and z it holds true that:

¬xRz  ¬zRx and ¬yRz  ¬zRy (where ¬ means not )

Now, since the negation of the weak preference relation (R) must also be transitive,
it holds furthermore true that:

xRz  zRy  xRy

From a logical point of view, the agent cannot at the same time weakly prefer x to
y (xRy) and not weakly prefer x to y (¬xRy) as this is inconsistent with transitivity.
Hence, the traditional view deems such behaviour to be irrational. But in reality it
may be rational to hold both xRy and ¬xRy at the same time, provided that the agent
has not yet decided which one to prefer. Following the traditional view, however, a
regulating authority might conclude that, since ¬xRz and since z seems preferable
from the point of view of the authority, the agent will be given z. In a second step,
the Mephistophelian authority might see that ¬zRy and since y is readily available
while z is not, will choose to give y to the agent. To come back to our example, a
regulating authority may move the (rationally acting) agent Dr Faustus – e.g. with a
nudge – from a romantic relationship with Gretchen to a religious life (since he is
indifferent between x and z) and then from a religious life to a romantic relationship
with Betchen (since he is also indifferent between z and y). Thus, Dr Faustus will
eventually end up with the less intelligent Betchen instead of the intelligent
Gretchen, although he clearly and rationally prefers the latter. The regulating
authority will not increase, but decrease welfare, even if we assume transaction
costs to be zero. What to do? There is an easy way out of this conundrum: simply
leave the doctor alone and let him reach a decision on his own. Sapere aude!
82 A.D. Steffen

6.3.2 Autonomy Instead of Rationality?

The point that has been overlooked in the previous discussion is autonomy.57 It
might be a mistake to believe that, because the agent has not yet formed a weak
preference – and hence is deemed by the regulator to be indifferent – between x and
z (z and y), the agent cannot have any meaningful reasons to object being given z
rather than x (y rather than z).58 The agent might want to reflect and to choose
according to her own free will rather than being given a predefined solution, even if
this entails a ‘wrong’ choice in the sense of prescriptive rationality. In other words,
the autonomous end result is incommensurable with the nudged end result, even if
the result is the same. Furthermore, neglecting the importance of free choice leads
to a commodification: the inherent value of autonomy is neglected and in its stead
autonomy and heteronomy become interchangeable goods. By the same token, the
choosable alternatives become commodities too, if the regulating authority is only
concerned about maximizing rationality. In its extreme form, such reasoning would
prefer a dictatorship propagating rationality to a democracy propagating autonomy.
The twentieth century has shown how this road leads: to serfdom. Indeed, as Putnam
points out, “[t]hinking of everything as a ‘commodity’ will necessarily blind one to
the most elementary facts about moral life.”59 A bit less provocatively, we might
state that in certain circumstances, some people may have a (weak) preference for
autonomous decision-making independent of the outcome. This is all it takes to
prove that paternalism, no matter how soft, may decrease rather than increase wel-
fare as measured by total individual well-being.
To sum up, if this reasoning is sound, then replacing autonomy with a nudge will
always result in a decrease in the utility of some people. For the nudge to be effi-
cient, two conditions have to be met: (1) the losses incurred by the people who value
autonomy have to be outweighed by the gains of the people who do not value auton-
omy. Even if condition (1) is met, it might still be the case that the transaction costs
accompanying the implementation of a nudge outweigh the gains. Thus, condition
(2) requires the resulting profits to be at least as large as transaction costs. In most
cases, this hurdle will probably not be cleared. To be sure, only a combination of
qualitative and quantitative studies will be able to yield a conclusive answer as to
whether and, if so, how much people value autonomy and whether the resulting
profit after transaction costs will be positive.
But perhaps ‘autonomy instead of rationality’ is the wrong question? After all,
most nudging theorists agree that people should always have the option to ‘opt out’
of an imposed default (a pre-set, nudged solution to a given problem). Does this not
re-install autonomy once again and hand back the full power to make decisions to
the individual? I shall argue in the next subsection that, due to psychological rea-
sons, it does not. In fact, opting out seems to be a bogus claim.

Cf. Putnam 2002/2004, pp. 82–86.
Putnam 2002/2004, p. 82.
Putnam 2002/2004, p. 83.
6 Nudging Is Judging: The Inevitability of Value Judgments 83

6.3.3 ‘Nudging’ as Transvaluation: ‘Opting Out’ a Bogus


We have seen that people may have a (weak) preference for autonomous decision-
making independent of the outcome. Followers of Schopenhauer might raise the
following objection: “Is it not irrational to let the agent ponder on her preferred
choice if what determines her choice eventually is nothing but the pre-existing
strongest (external) motive rather than her (internal) free will?” This strategy makes
two claims: (1) that the rational preference ordering is predetermined, and (2) that it
makes no difference whether external variables or the (equally external) decision of
the legislator tip the scales in favour of an alternative (in Schopenhauer’s sense,
neither of them is autonomous, both are heteronomous). While the first proposition
presupposes a deterministic world, the second proposition assumes that there is no
rational basis to form a preference for autonomous decision-making over heterono-
mous decision-making since both are external. Autonomy does not exist because
the free will is not free. In other words, Schopenhauer transvalues autonomy into
Both of these topics are far too complex to explore in any depth in this essay.
Suffice it to say that the first argument only works in a deterministic world which,
from today’s scientific point of view (in light of modern physics) seems highly
unlikely to be universally true. The second argument, however, is misleading insofar
as it builds upon a perverted notion of ‘autonomy’: autonomy does not entail that
everything lying outside of my self is precluded from influencing me. To interpret
autonomy in the strict sense of absolute freedom is a false interpretation of the origi-
nal concept of autonomy. The Ancient Greek αὐτο-νομία literally means self-
governing. Thus, I am not completely free, but I rule over myself. This does by no
means exclude external motives from influencing my (autonomous) decision. In
fact, it is quite clear that in its original Greek form, autonomy precludes absolute
freedom and encourages one to thoroughly analyse and carefully weigh different
external motives against each other, as good rulers does. Schopenhauer’s interpreta-
tion of autonomy is unsound. Even worse, it may justify a paternalistic intervention
in a two-step transvaluation: first, it ‘unwills’ the agent, undermines her set of val-
ues and superimposes a specific set of values (the set of values of the Schopenhauerian
speaker, the regulating authority or the legislator) in its stead; second, it judges
agents on grounds of the superimposed set of values rather than on the set of values
of their own. In other words, by negating the agent’s second order desires (to have
the will one wants to have), this strategy also negates the agent’s first order desires
(to do what one wants to do).60 By means of this simple strategy, any paternalistic
intervention may be justified.
Surely though, no one today would be so shrewd as to apply this strategy. After
all, we are living in an enlightened world. Notwithstanding, the strategy is similar
to what Sunstein and Thaler are doing when they declare that one’s own will may

Cf. Bernstein 1991, pp. 203–204. Also see Frankfurt 1988, p. 22.
84 A.D. Steffen

be irrational (boundedly rational, to be precise) and thus ought to be discouraged.

To be fair, Sunstein and Thaler introduce the option to opt out into the equation. But
opting out clauses do not in fact reintroduce autonomous decision-making, since,
when an opting out clause becomes effective, the transvaluation will already have
taken place. Superficially it seems like the agent’s second and first order desires are
reinstated. In fact, however, the agent commits a transgression against the new
order as soon as she opts out. This transgression is likely to lead to cognitive dis-
sonance, that is to say, a situation involving conflicting attitudes, beliefs or behav-
iours. This conflict produces a feeling of discomfort. In an attempt to reduce this
negative feeling and to restore balance, the agent seeks to alter one of the conflicting
attitudes, beliefs or behaviours in such a way that the conflict disappears.61 In other
words, by staying true to herself, the agent opts for ‘irrationality’ in the new order.
Her decision is not approved by society and might meet with heavy criticism, even
from friends. In a worst case scenario, she might be discriminated against and ostra-
cized. She might anticipate this and even seriously consider to not opt out. Consider
a smoker who is well-informed about the risks of smoking but still decides to smoke.
In today’s society, her decision is bound to be criticized. The resulting tension is a
form of cognitive dissonance. In an attempt to avoid this tension, she might refrain
from smoking. This, however, makes her worse off as she does not maximize her
individual, boundedly rational benefit, but the perfectly rational benefit of the ideal
society as interpreted by the nudger. Apart from (in most cases admittedly rather
low) transaction costs, this may be a reason why people refrain from opting out.
Hence, it is plausible to assume that people only very rarely opt out, which indeed
appears to be the case.62

6.4 Conclusion: Nudging Is Judging

The repercussions for rational choice and behavioural (law and) economics in gen-
eral, and nudging in particular, are straightforward: rational preferences, while
empirically observable, are only a marginal phenomenon. It is not an axiom (from
Ancient Greek ἀξίωμα: that which is thought to be self-evident). This has been
empirically tested by behavioural economics. However, to believe that this short-
coming may be remedied by paternalism, and this is the final goal of behavioural
law and economics, is an elementary mistake. Empirics can give us valuable insights
into our behavioural patterns. However, ‘rationality’ – treated by rational choice
and behavioural (law and) economics as a non-normative, positive concept – is in
fact a fuzzy term, that is positive and normative at the same time. The positive and
the normative are like two sides of the same coin. The concept is positive only inso-
far as it is treated as a neutral supposition (as in the question: do people have

See e.g. Aronson et al. 2008, pp. 163 et seqq.
Cf. Carney 2007. Retrieved on 31.08.2015 via:
6 Nudging Is Judging: The Inevitability of Value Judgments 85

perfectly rational preferences?). Behavioural economics has shown that this sup-
position is observable some of the time, but not all of the time. The concept becomes
normative as soon as it loses its neutrality as a supposition and is treated as a goal
for individuals, companies or whole societies. In this manner, rational choice and
behavioural law and economics propagate and superimpose a specific value-system,
which is claimed to be non-normative and value-neutral because it only leads back
to the point of origin (completely rational preferences). This is a mistake. ‘Perfect
rationality’ becomes the touchstone against which all behaviour is judged with the
implicit notion that whatever is ‘perfectly rational’ is ‘good’ (‘boundedly rational’
is ‘bad’) and hence must (not) be in the interest of the individual, the company,
society. In doing so, other parameters – such as autonomy – are not only neglected,
but (on grounds of their alleged normativity) negated. Autonomy and rationality
may both be a fact and a value at the same time. They may have a qualitatively dif-
ferent, equally important, utility to the agent. Indeed, it seems plausible that people
are only able to build up a sound internal value system via autonomous experience.
The interests, values and reasons so acquired may very aptly be called their ‘ratio-
nality’. After all, since they have reasons for doing so, they follow their interests
‘rationally’ and hold their values equally ‘rationally’, even if these interests are
deemed to be ‘irrational’ by the touchstone of perfect rationality. To say that the
concept ‘rational’ means ‘perfectly rational’ is equally nonsensical as to interpret
the concept ‘free will’ as ‘perfectly free’. Paternalistic measures may increase ‘per-
fectly rational welfare’. However, this is the wrong unit of measure. It is not equal
to the ‘boundedly rational welfare’ experienced by individuals. Additionally, indi-
viduals will hardly ever choose to ‘opt out’ in order to prevent cognitive dissonance.
Moreover, missed learning effects might necessitate implementing further paternal-
istic measures in the future.
It would thus appear that, even if the regulatory agency were perfectly informed
or at least better informed than its citizens in certain matters (a supposition that is in
itself highly contested), it should still refrain from engaging in paternalistic mea-
sures, as soft as they may be, since they seem to generally decrease rather than
increase welfare. From a legal point of view, government should limit its activity to
a comprehensible legislation that is mostly limited to negative obligations of its citi-
zens since positive obligations – such as the duty to act rationally – may quickly
conflict overarching negative obligations.63 Governments, however, may have a
positive obligation to focus on encouraging the free development of their citizens’
‘autonomous rationality’ (via providing information and room for experience)
rather than simply forcing them to adopt a specific concept of rationality. As has
been shown, it is nonsense to assume the existence of a neutral, universal and ulti-
mately justified rationality.
A further implication of the fact/value dichotomy is that nudging theorists can-
not delegate the delicate task of identifying ethical goals to philosophers or politi-
cians, as they themselves engage in both positive and normative science all the time.
Today, many researchers present behavioural (law and) economics as a strict

See for example: Bastiat 1848/1995, chapter 2.
86 A.D. Steffen

positive, empirical science. In doing so, an unscientific double standard is main-

tained. This could and should be remedied by making the normative explicit and by
incorporating it. Being part of a social science, propagators of nudging should also
take heed not to fall prey to apriorism (i.e. the doctrine that knowledge rests upon
unquestionable principles that are self-evident) or authoritarianism (i.e. the require-
ment to obey postulated rules or laws without allowing for individual freedom).
Both, apriorism and authoritarianism, are unreasonable. Researchers need to realize
that all science, both positive and normative, is fallible but may still yield funda-
mental knowledge in a step-by-step evolutionary way. The idea of a complete (i.e.
objective and universal) behavioural (law and) economics is absurd. Reality is char-
acterized by pluralistic values and these values are inseparably entangled with facts
(and vice versa). Ultimately, what is imperative in behavioural (law and) economics
as well as in other (positive-normative) sciences, is to make the values behind the
facts (and the facts behind the values) transparent by explicitly accepting and incor-
porating rather than shunning the normative.

6.5 Outlook: Amalgamation of Kant and Aristotle,

or ‘Pragmatic Ethics’

To sum up, judgments of reasonableness and rationality presuppose normative

judgments, epistemic or otherwise.64 In the day of modern science, looking for uni-
versal truths is chasing the overcome ghost of Plato. An approach both more suit-
able to the challenges of the twenty-first century and more fruitful too would be
pragmatism, also in regard to ethical questions that arise in behavioural law and
economics and other such sciences. An ethics that is at the same time fallible and
practical. Pragmatic ethics is the belief that ethics, as well as other normative sci-
ences, develop in the same way as positive sciences do: by inquiry into the truth of
hypotheses. That a given hypothesis cannot be rejected does not prove the hypoth-
esis. But until it is superseded by a more adequate hypothesis, scientists act upon the
hypothesis as if it were true. It is thus important to leave enough room to test
hypotheses rather than accept them as universally true axioms.
Such a pragmatic ethics could be an amalgamation of the ethics of Kant and
Aristotle in the pragmatist spirit of John Dewey.65 It is often said that the ethics of
Kant and Aristotle stand in such stark contrast to each another that the two will
never meet. In fact, however, the stark-naked skeleton – stripped of its metaphys-
ics – of abstract Kantian ethics (as an abstract method) could profitably be fleshed
out by the meaty chunks of concrete Aristotelian ethics on Eudaimonia (from
Ancient Greek εὐ-δαιμονία: literally ‘good spirit’, usually translated as ‘the good

Putnam 2002/2004, p. 145.
Cf. Putnam 2005, pp. 31–32.
6 Nudging Is Judging: The Inevitability of Value Judgments 87

life’ in the sense of ‘a virtuous way of living’) (content).66 In pragmatic ethics, we

can still speak of (conditional) truth and (conditional) objectivity, but not of ‘final
truths’ and ‘perfect objectivity’.
If behavioural law and economics accepted its normative content, it could gain a
lot from incorporating rather than shunning its normative content including ethical
issues. A ‘pragmatic turn’, for example, would open the gates to concepts hitherto
frowned upon for being unscientific, such as autonomy. This would allow to impar-
tially weigh the pros and cons of rationality and autonomy against each other. Only
a pragmatic ethics can do justice to the pluralistic values in life and strive to restore
individual autonomy while at the same time being part of scientific discourse.

Acknowledgement I would like to sincerely thank Prof. Dr. iur. Klaus Mathis, MA in Economics,
and Katrien Van Den Bergh, MA, for their valued input and careful proofreading.


Arlen, Jennifer. 1998. Comment: The future of behavioral economic analysis of law. Vanderbilt
Law Review 51: 1765.
Aronson, Elliot, Timothy D. Wilson, and Robin M. Akert. 2008. Sozialpsychologie, 6th ed.
München: Pearson Studium.
Bastiat, Frédéric. 1848/1995. The law. In Selected essays in political economy. Trans. Seymour
Cain. Chapter 2. Irvington-on-Hudson: The Foundation for Economic Education.
Bernstein, Jay. 1991. Autonomy and solitude. In Nietzsche and modern German thought, ed. Keith
Ansell Pearson, 203–204. London/New York: Routledge.
Brune, Jens Peter. 2010. Moral und Recht. Zur Diskurstheorie des Rechts und der Demokratie von
Jürgen Habermas. Freiburg im Breisgau: Karl Alber.
Carnap, Rudolf. 1934. The unity of science. London: Kegan Paul, Trench, Hubner.
Carnap, Rudolf. 1938. The foundations of logic and mathematics. In International Encyclopedia
of United Science vol. 1, Part 1, 139–214. Chicago: Chicago University Press.
Carney, Scott. 2007. The case for mandatory organ donation. Wired, 8 May.
Christensen-Szalanski, Jay J., and Cynthia Fobian-Willham. 1991. The hindsight bias: A meta-
analysis. Organizational Behavior and Human Decision Processes 48: 147–168.
Englerth, Markus. 2007. Behavioral law and economics – eine kritische Einführung. In Recht und
Verhalten, ed. Christoph Engel, Markus Englerth, Jörn Lüdemann, and Indra Spiecker genannt
Döhmann, 60–130. Tübingen: Mohr Siebeck.
Frankfurt, Harry Gordon. 1988. The importance of what we care about, 22. Cambridge: Cambridge
University Press.
Habermas, Jürgen. 1992/1996. Between facts and norms. Trans. William Rehg. Cambridge: Polity
Habermas, Jürgen. 1992/1998. Faktizität und Geltung. Frankfurt am Main: Suhrkamp.
Habermas, Jürgen. 1999. Wahrheit und Rechtfertigung. Philosophische Aufsätze. Frankfurt am
Main: Suhrkamp.
Hume, David. 1738/1978. A treatise of human nature, eds. L.A. Selby Bigge, and P.H. Nidditch.
Oxford: Oxford University Press.
Hume, David. 1748/2008. An Enquiry concerning human understanding. Oxford World’s Classics.
Oxford: Oxford University Press.

Cf. Putnam 2002/2004, p. 134.
88 A.D. Steffen

Hume, David. 1777/1975. Enquiries concerning human understanding and concerning the prin-
ciples of morals, 3rd ed, eds. L.A. Selby Bigge, and P.H. Nidditch. Oxford: Oxford University
Issacharoff, Samuel. 1998. Can there be a behavioral law and economics? Vanderbilt Law Review
51: 1729.
Jolls, Christine. 2009. Behavioral law and economics. In Law and economics – A reader, ed. Alain
Marciano. New York: Routledge/Chapman & Hall.
Jolls, Christine, and Cass R. Sunstein. 2006. Debiasing through law. Journal of Legal Studies 35:
Jolls, Christine, Cass R. Sunstein, and Richard Thaler. 1998. A behavioral approach to law and
economics. Stanford Law Review 50: 1471–1550.
Kahneman, Daniel. 1997. New challenges to the rationality assumption. Legal Theory 3: 105.
Kahneman, Daniel. 2003. A perspective on judgment and choice. Mapping bounded rationality.
American Psychologist 58: 697.
Kahneman, Daniel. 2011. Thinking, fast and slow. New York: Farrar, Straus and Giroux.
Kahneman, Daniel, and Amos Tversky. 1979. Prospect theory: An analysis of decision under risk.
Econometrica XVLII: 263–292.
Kahneman, Daniel, and Amos Tversky. 1981. The framing of decisions and the psychology of
choices. Science 211(4481): 453–458.
Kahneman, Daniel, Jack Knetsch, and Richard Thaler. 1990. Experimental tests of the endowment
effect and the coase theorem. Journal of Political Economy 98: 1325–1348.
Kahneman, Daniel, Jack Knetsch, and Richard Thaler. 1991. Anomalies: The endowment effect,
loss aversion, and status quo bias. The Journals of Economic Perspectives 5(1): 193–206.
Kant, Immanuel. 1781/1998. The critique of pure reason. Trans. P. Guyer, and A.W. Wood.
Cambridge: Cambridge University Press.
Kelman, Mark. 1998. Behavioral economics as part of a rhetorical duet: A response to Jolls,
Sunstein, and Thaler. Stanford Law Review 50: 1577.
Korobkin, Russell, and Thomas Ulen. 2000. Law and behavioral science: Removing the rationality
assumption from law and economics. California Law Review 88: 1051–1144.
Latin, Howard. 1994. “Good” warnings, bad products, and cognitive limitations. UCLA Law
Review 41: 1193–1295.
Locke, John. 1689/2008. An essay concerning human understanding. Oxford World’s Classics.
Oxford: Oxford University Press.
Mas-Collel, Andreu. 1982. Revealed preference after Samuelson. In Samuelson and neoclassical
economics, ed. George R. Feiwel, 72–82. Heidelberg/New York/Dordrecht/London: Springer.
Mathis, Klaus, and Ariel Steffen. 2015. From rational choice to behavioural economics: Theoretical
foundations, empirical findings and legal implications. In European perspectives on behav-
ioural law and economics, economic analysis of law in European legal scholarship 2, ed. Klaus
Mathis, 31–48. Cham/Heidelberg/New York/Dordrecht/London: Springer.
Meyerowitz, Beth E., and Shelly Chaiken. 1987. The effect of message framing on breast self-
examination: Attitudes, intentions, and behavior. Journal of Personality and Social Psychology
52: 500–510.
Mill, John Stuart. 1859/1869. On liberty, 4th ed. London: Longman, Roberts and Green Co.
Mitchell, Gregory. 2005. Libertarian paternalism is an oxymoron. Northwestern University Law
Review 99(3): 1–42.
Moore, George Edward. 1903. Principia Ethica. Cambridge: Cambridge University Press.
Ott, Konrad. 2001. Moralbegründungen zur Einführung. Hamburg: Junius.
Parisi, Francesco, and Vernon Smith. 2004. The law and economics of irrational behavior.
Stanford: Stanford University Press.
Putnam, Hilary. 2002/2004. The collapse of the fact/value dichotomy and other essays, including
the rosenthal lectures. 3rd printing. Cambridge, MA/London: Harvard University Press.
Putnam, Hilary. 2005. Ethics without ontology. Cambridge, MA/London: Harvard University
6 Nudging Is Judging: The Inevitability of Value Judgments 89

Quine, Willard Van Orman. 1951. Two dogmas of empiricism. Philosophical Review 60: 20–43.
Rachlinski, Jeffrey J. 2000. A positive psychological theory of judging in hindsight. In Behavioral
law and economics, ed. Cass R. Sunstein, 95–115. Cambridge: Cambridge University Press.
Schweizer, Mark. 2005. Kognitive Täuschungen vor Gericht. Eine empirische Studie. Zürich:
Ohne Verlag.
Sen, Amartya. 1967. The nature and classes of perspective judgments. Philosophical Quarterly
17(66): 46–62.
Sen, Amartya. 2000. Development as freedom. New York: Anchor Books.
Simon, Herbert A. 1979. Rational decision making in business organizations. American Economic
Review 69: 493–513.
Smith, Adam. 1776/2008. An inquiry into the nature and causes of the wealth of nations, ed.
Kathryn Sunderland. Oxford: Oxford World’s Classics.
Sunstein, Cass R. (ed.). 2000. Behavioral law and economics. Cambridge: Cambridge University
Sunstein, Cass R. 2002. Probability neglect: Emotions, worst cases, and law. Yale Law Journal
112: 61–107.
Sunstein, Cass R. 2014. Why nudge? New Haven/London: Yale University Press.
Sunstein, Cass R., and Richard H. Thaler. 2003. Libertarian paternalism is not an oxymoron. The
University of Chicago Law Review 70(4): 1159–1202.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge: Improving decisions about health, wealth
and happiness. New Haven/London: Yale University Press.
Thomas, Ayanna K., and Peter R. Millar. 2011. Reducing the framing effect in older and younger
adults by encouraging analytic processing. The Journals of Gerontology Series B: Psychological
Sciences and Social Sciences 67B(2): 139–149.
Tversky, Amos, and Craig R. Fox. 1995. Weighing risk and uncertainty. Psychological Review
102(2): 269–283.
Tversky, Amos, and Daniel Kahneman. 1974. Judgment under uncertainty: Heuristics and biases.
Science. New Series. 185(4157): 1124–1131.
Tversky, Amos, and Daniel Kahneman. 1991. Loss aversion in riskless choice: A reference-
dependent model. Quarterly Journal of Economics 107: 1039.
van Aaken, Anne. 2015. Judge the nudge: In search of the legal limits of paternalistic nudging in
the EU. In Nudging and the law. What can EU law learn from behavioural sciences?, eds.
Alberto Alemanno, and Anne Lise Sibony. Forthcoming. Oxford: Hart Publishing.
von Humboldt, Wilhelm. 1792/1989. Ideen zu einem Versuch, die Grenzen der Wirksamkeit des
Staates zu bestimmen. Stuttgart: Reclam Philipp jun. GmbH & Co. KG.
Walsh, Vivian. 2000. Smith after Sen. Review of Political Economy 12(1): 5–25.
Part II
Chapter 7
Nudging and the Principle of Proportionality
Obliged to Nudge?

Mark Schweizer

Abstract While the ethics and politics of nudging have received a great deal
of attention, the legality of non-coercive interventions aimed at changing
human behaviour has not been the object of much study. In this contribution, I
examine firstly which limits, if any, the principle of proportionality
(‘Verhältnismäßigkeitsgrundsatz’) as applied by the German Federal Constitutional
Court imposes on governmental ‘choice architects’. While nudges as such generally
do not interfere with fundamental rights, including the very broad ‘right of personal
development’ of Art. 2(1) Basic Law for the Federal Republic of Germany, their
implementation will often interfere with the fundamental rights of citizens other
than the decision makers and therefore trigger constitutional scrutiny. Since pater-
nalistic motives are – in principle – not accepted as legitimate ends that may justify
an interference with fundamental rights, nudges solely intended to protect the deci-
sion maker from self-harm may not pass the proportionality test. Secondly, I exam-
ine whether the necessity prong of the proportionality principle may force the
legislator to choose a non-coercive nudge over a more traditional coercive measure.
Given the empirical evidence of the limited effectiveness of nudges in changing
behaviour compared to traditional regulatory means, I come to the conclusion that
the proportionality principle does not compel the use of nudges.

M. Schweizer (*)
Law School, University of St. Gallen, Dufourstrasse 50, 9000 St. Gallen, Switzerland

© Springer International Publishing Switzerland 2016 93

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_7
94 M. Schweizer

7.1  Introduction

The ethics and politics of ‘nudging’ citizens towards a choice that – in the govern-
ment’s opinion – is in the citizens’ own best interests have been hotly debated.1 The
legality of nudging has received much less attention.2 The question whether govern-
ments might be obliged to nudge has been largely ignored so far. Prima facie, the
principle of proportionality – a German legal export widely adopted in (continental)
European legal doctrine3 – might provide a rationale that compels nudging. Reduced
to its core, the principle of proportionality requires that a government action which
abridges a fundamental right is suitable for achieving a legitimate public interest
and is necessary in the sense that there is no less restrictive alternative – i.e. a mea-
sure that infringes the fundamental right(s) less – that achieves the same end. A
nudge, as a non-coercive measure to influence choice, is generally a milder measure
than traditional coercive measures (i.e., prohibitions or commandments). If the
nudge achieves the same end, the principle of proportionality dictates that it has to
be chosen over the coercive measure.
In this contribution, I will examine whether the above plausible conjecture holds
up to closer scrutiny. The article is structured as follows: first, nudging is defined
and an example is introduced that will be used throughout the text to illustrate the
concept. Secondly, the principle of proportionality is explained in some depth. I will
primarily rely on the German doctrine, as it is more finely tuned than that of many
other jurisdictions,4 the modern principle of proportionality being a German con-
cept that plays a huge role in German constitutional law.5 Then, each prong of the
proportionality test is applied to nudging to see whether it prohibits or promotes
nudging. Unsurprisingly, it turns out that things are not as simple as set out above.
Tentative conclusions end this contribution.

 An exhaustive overview of the literature is impossible, some starting points are Camerer et al.
2003; Thaler and Sunstein 2003; Rachlinski 2003; Glaeser 2006; Hill 2008; Sugden 2008; Bovens
2009; Hausman and Welch 2010; White 2010; Selinger and Whyte 2011; Blumenthal-Barby and
Burroughs 2012; Goodwin 2012; Sunstein 2014; 2015 and the contributions by Bruno S. Frey/Jana
Gallus, Peter G. Kirchschläger and Mark D. White in this volume; from a German perspective van
Aaken 2006; Schnellenbach 2011; Kirchgässner 2013; Neumann 2013; Fischer and Lotz 2014.
 But see Alemanno and Spina 2014; van Aaken 2015a, b.
 Saurer 2012, p. 8; Wahl 2013, p. 823
 Classen 2012, p. 659; Wahl 2013, p. 836.
 Pieroth 2014, para. 289; Sodan and Ziekow 2014, p. 227. For an overview over the historic devel-
opment see Koch 2003, pp. 4 et sq.
7  Nudging and the Principle of Proportionality 95

7.2  ‘Choice Architecture’ and ‘Nudging’

One of the core insights of the psychological study of judgment and decision ­making
is that judgments and choices are context-sensitive. How a choice is framed influ-
ences the choice. An outcome framed as a loss is treated differently than the same
outcome framed as a gain.6 The bigger a plate, the more one eats.7 Whether a choice
has immediate consequences or the decision maker commits to an action in the (far)
future makes a difference – I want a chocolate bar right now, but next week, I will
eat an apple a day.8 Whether an option is a default or requires an active choice (‘opt-
in’) makes a huge difference.9 Few people choose to be organ donors, however,
when organ donation is the default, few people opt out.10 The list of ‘biases’ seems
One takeaway is that humans are hopelessly incompetent to make choices that
actually further their own best interests, and that the government should – if not
making those choices for them – at least direct them towards the ‘right’ choice. This
justification of nudging has received a lot of criticism.12 Another view is that since
choice is unavoidably context-based, there is no such thing as an unbiased choice
that truly reveals the individual’s preference.13 Replacing one biasing context by
another is justified if there is evidence that one option is actually better than the
other, and the new context leads to that option being chosen more often.14 The activ-
ity of organizing the context in which a decision is made is termed ‘choice
architecture’.15 A nudge is defined as “any aspect of the choice architecture that
alters people’s behaviour in a predictable way without forbidding any options or
significantly changing their economic incentives.”16 In other words, the choice set
remains the same, it is only structured differently.
Sunstein and Thaler use the example of a buffet at a school cafeteria to illustrate
the concept of choice architecture.17 If meats and fatty foods are presented promi-
nently, they will be chosen more often than when they are presented less conspicu-
ously. Since there is evidence that eating too much processed meat and partially

 Tversky and Kahneman 1981.
 Wansink and van Ittersum 2006.
 Read and van Leeuwen 1998.
 E.g., Nessmith et al. 2007; but see Bronchetti et al. 2011; Russell 2015.
 Johnson and Goldstein 2003.
 The chapter headings in Pohl 2004 give an overview over the most widely studied fallacies and
biases in thinking, judgment and memory.
 Rachlinski 2003; Englerth 2007, pp. 23 et sq.
 Thaler and Sunstein 2003, p.175.
 Thaler and Sunstein 2008, p. 237.
 Thaler and Sunstein 2008, p. 3.
 Thaler and Sunstein 2008, p. 6.
 Thaler and Sunstein 2008, p. 2.
96 M. Schweizer

hydrogenated oils (which contain trans fats) leads to coronary heart disease,18 it
would be desirable for most people to reduce their intake of those foods. Sunstein
and Thaler suggest that the presentation of the food be structured such that healthy
foods, such as salads and vegetables, are placed more prominently, for example by
highlighting them.19 Other nudges include putting unhealthy food (e.g. cheese) fur-
ther out of reach than healthy food (e.g. broccoli) and providing spoons for healthy
foods items and tongs for unhealthy food items (tongs have less capacity). This is
shown to reduce intake of the unhealthy food items by as much as 8–16 %.20 Even
simply placing menu items at the beginning or end of their category increases their
popularity by about 20 %.21
The example also shows that while a nudge is non-coercive on the decision
maker, its mandated introduction may well rest on coercion. I.e., if the cafeteria is
run by a private company and the government legislates that food items in school
cafeterias must be presented in a certain way, the compliance of the company is
presumably enforced by coercive means or changing economic incentives (i.e.,
non-compliant companies will lose the contract).22 Behind a nudge for the consumer
lies a coercive measure for service providers. We will come back to this.

7.3  The Principle of Proportionality

Proportionality has been a concept of law since antiquity, primarily in criminal law
(‘let the punishment fit the crime’, i.e., be proportionate).23 Its modern form was
developed in nineteenth century German administrative law as a restriction on the
executive (‘police’) power.24 It is only after the Second World War that it has
become an important part of German constitutional law, restricting the freedom of
the legislator.25 Today, the principle is recognized in many countries as part of con-
stitutional law.26 It is enshrined in the Charter of Fundamental Rights of the
European Union (Art. 52(1)) and plays an important role in the adjudication of the

 Mensink et al. 2003; Micha et al. 2010.
 Thaler and Sunstein 2008, p. 2.
 Dayan and Bar-Hillel 2011.
 Rozin et al. 2011.
 Since the government is largely free to set contract terms (as long as they are not discriminatory),
contracting only with companies that agree to adhere to ‘behavioral food presentation’ may not
trigger any constitutional concerns. It is easy to imagine, however, that the government would try
to mandate these measures for all restaurants, which would clearly entail coercion.
 Koch 2003, pp. 39 sq.
 Koch 2003, pp. 45 sq.
 Schlink 2001, p. 445.
 Koch 2003, pp. 48 sqq; Saurer 2012.
7  Nudging and the Principle of Proportionality 97

European Court of Human Rights.27 American law, on the other hand, has not been
receptive to balancing in the context of fundamental rights adjudication.28
While the principle of proportionality as such is recognized in many jurisdic-
tions, its application varies to a great degree. Namely the level of scrutiny applied
when assessing the proportionality of a measure differs greatly. A discussion of the
application of ‘the’ principle of proportionality to nudging must therefore remain at
a very abstract level, limiting its usefulness. I will therefore concentrate on German
doctrine, which is very finely tuned. I will also limit myself to the principle of pro-
portionality in the context of fundamental rights. The principle of proportionality
has other applications, but it is hard to see how they could relate to nudging.
The principle of proportionality is usually presented as a three prong test for
administrative measures that assesses (a) whether a measure interfering with a right
is suitable to achieve its objective, (b) whether it is necessary for that purpose and
(c) whether it excessively burdens the individual compared with the benefits it aims
to secure.29 In the assessment of legislative acts it becomes a four prong test, the first
prong being the assessment whether the measure serves a legitimate public inter-
est.30 This prong is often neglected because in practice, it has little bite. The German
Constitutional Court understands legitimate public interest so broadly that almost
any interest suffices. It may, however, have some bearing on the constitutionality of
paternalistic nudges, as will be explained below.
The principle of proportionality limits the interference of the state with funda-
mental rights. The German Constitution (‘Grundgesetz’) recognizes explicitly that
laws may interfere with fundamental rights, or in other words, the scope of funda-
mental rights may be restricted by parliamentary act.31 The proportionality principle
materially limits the extent such interference is permissible; it restricts the restric-
tions (‘Schranken-Schranke’).32 The principle of proportionality is therefore trig-
gered by an interference with a fundamental right, which raises the question whether
a ‘nudge’ can ever be considered an abridgment of a fundamental right.

 Arai-Takahashi 2002, pp. 14 sq.
 See District of Columbia v. Heller, 554 U.S. 570 (2008), for a discussion Cohen-Eliya and Porat
2009; historical overview Porat 2014.
 E.g. Michael 2001a, p. 148; Stein and Frank 2010, p. 245.
 E.g. Zippelius and Würtenberger 2008, p. 219; Ipsen 2014, para.184; Sodan and Ziekow 2014,
p. 228.
 See Art. 3(2), 5(2), 8(2), 12(1) Basic Law, and more often. For the understanding of ‘law’ as
parliamentary act see Pieroth 2014, paras. 272 et sq.
 Pieroth 2014, paras. 285 et sq.
98 M. Schweizer

7.4  Does Nudging Interfere with Fundamental Rights?

7.4.1  A Constitutional Right Protecting Personal Autonomy

The German Constitution recognizes a fundamental right to freedom of personal

development and the right to private and family life (Art. 2(1) Basic Law for the
Federal Republic of Germany, hereinafter ‘Basic Law’). In the early days of the
German Constitution, there was considerable debate whether the freedom of per-
sonal development only protected important lifestyle choices or any choice at all.33
The German Constitutional Court soon sided with the proponents of a wide inter-
pretation of Art. 2(1) holding that the freedom of personal development was to be
understood as a general freedom of choice. The wording of the original draft that
“everybody can do as he pleases” had only been replaced because it was not consid-
ered solemn enough for a constitution, but no change in scope had been intended.34
Subsequently, the Constitutional Court held for example that horse riding in the
forest and feeding pigeons in a park were activities protected by the right to freedom
of personal development.35
The view that art. 2(1) protects the freedom of choice as such can be considered
reigning today,36 despite recurring criticism.37 The protection afforded by the right
to personal development extends to ‘everything that is in the interest of a person’s
autonomy’.38 It is undisputed that any state act interfering with personal choice is
subject to constitutional scrutiny.39 However, since nudging never prohibits a
choice, freedom of choice seems unaffected. But autonomy is more than freedom of
choice – while freedom concerns the ability to act, autonomy concerns the indepen-
dence and authenticity of the desires that move one to act in the first place.40 It has
been argued that the consciously employed organization of the choice context to
influence the decision maker interferes with an individual’s autonomy.41
While there are many definitions of personal autonomy, personal autonomy is
“generally understood to refer to the capacity to be one’s own person, to live one’s
life according to reasons and motives that are taken as one’s own and not the prod-
uct of manipulative or distorting external forces.”42 At the heart of the debate

 Peters 1963, pp. 47 et sq.
 BVerfGE 6, 32, pp. 36 sq. – ‘Elfes decision’.
 BVerfGE 54, p. 143 – ‘pigeon feeding’; BVerfGE 80, 137, pp. 152 et sq. – ‘riding in the
 Ipsen 2014, para. 771; Pieroth 2014, paras. 386 et sq.
 See the references cited in Cremer 2003, p. 81.
 Möller 2014, p. 163.
 Zippelius and Würtenberger 2008, p. 244.
 Christman 21.3.2015, 1.1 Basic Distinctions.
 Hill 2008; Sugden 2008; White 2010; Blumenthal-Barby and Burroughs 2012; Goodwin 2012;
Miller and Gelinas 2013; Wilkinson 2013; van Aaken 2015b.
 Christman 21.3.2015, Introduction; cited approvingly by van Aaken 2015b, p. 9.
7  Nudging and the Principle of Proportionality 99

whether nudging interferes with personal autonomy lies therefore the question
whether nudges – some or all of them – are ‘manipulative and distorting external

7.4.2  Nudging and Personal Autonomy

From the perspective of autonomy, not all nudges are created equal. Nudges that
merely provide information, or make certain information more salient, clearly do
not interfere with autonomy and do not trigger constitutional scrutiny. Autonomy
requires informed consent, so nudges that are designed to inform the decision maker
about the consequences of his decision actually enhance autonomy.43 In a similar
vein, the Federal Constitutional Court has held that the provision of truthful infor-
mation by the government cannot be considered an interference with fundamental
rights, even if the information ‘names and shames’ certain competitors and causes
them considerable losses.44 Efficient markets depend on transparency, and measures
enhancing market transparency are prima facie constitutional.45
While it is widely accepted that the provision of purely factual information does
not interfere with fundamental rights, Lüdemann argues convincingly that educa-
tional campaigns designed to change social norms or instil new social norms can be
an indirect interference with fundamental rights, namely personal freedom in the
sense of Art. 2(1) Basic Law.46 The violation of a social norm can have conse-
quences that are at least as severe as a minor fine. While these consequences – such
as social shaming or even ostracism – are not directly attributable to the govern-
ment, if the government actively tries to establish a social norm, it is indirectly
responsible for those consequences. According to Lüdemann, such educational
campaigns require a legal basis in a law in the formal sense and for them to be pro-
portional it must be shown that other measures are not effective in changing the
­citizens’ behaviour in the desired way.47
The more difficult issues arise not with informational nudges designed to engage
the audience on a conscious, cognitive level, but with nudges that work on a sub-
conscious or emotional level. Critics argue that such nudges are manipulative and
irreconcilable with personal autonomy.48 Van Aaken distinguishes nudges targeting
the formation of preferences and ‘nudges’ aimed at correcting cognitive errors in
order to help people pursue their own preferences rationally. The former require

 Sunstein 2015, p. 30; van Aaken 2015a, p. 13.
 BVerfGE 105, 252, p. 268 – ‘glycol warning’.
 BVerfGE 105, 252, p. 267 – ‘glycol warning’.
 Lüdemann 2004, pp. 123 et sq.
 Lüdemann 2004, pp. 127 et sq.
 Blumenthal-Barby 2013, pp. 190 et sqq.; Yeung 2012; Wilkinson 2013.
100 M. Schweizer

greater scrutiny than the latter, because they interfere with personal autonomy.49
Particularly problematic are nudges that engage the automatic, unconscious ‘System
I’ of thinking and from which the individual, for lack of awareness, cannot escape
by reflection.50 Such nudges, which van Aaken terms ‘invisible’, are held to inter-
fere with personal autonomy.51
The argument that nudges aimed at changing preferences deserve special scru-
tiny seems to rest on the premise that there are stable preferences that are indepen-
dent of the decision context, and the psychological literature teaches us that this is
not the case.52 If all preferences depend on context, it is hard to see why one con-
text – for example framing a decision as choosing between losses instead of gains –
is more manipulative than another one.53 There must be some context of the decision,
and we might as well choose a context that is more likely to lead to better choices
for more people.54 Even appeals to emotion are hard to be considered manipulative
if one truly accepts the context-dependence of any choice.55 Graphic displays of the
health consequences of smoking on cigarette packaging may well appeal not only to
the rationality of smokers, but also to their emotions. But beautifully shot displays
of healthy, handsome, happy smokers conquering the desert, jungle or the opposite
sex are no less appealing to emotion. It is simply impossible to say what the ‘true’,
context-free preference for smoking is, and there is no indication that the current
context is the one that reveals the ‘true’ preferences of the decision maker.
The aforesaid could lead to the conclusion that non-coercive ‘choice architec-
ture’ never interferes with personal autonomy, but that would not be justified. There
are instances of nudges that are truly manipulative and cannot be reconciled with
respect for personal autonomy. Blumenthal-Barby uses the example of someone
who wishes to convince his overweight roommate to lose weight by replacing all
mirrors in the house with distorting mirrors that make the person look fatter.56 Even
if the goal is ‘good’ it is hard to dispute that the measure taken interferes with the
autonomy of the target, not only because the target is unware, but also because the
nudge is based on providing false information.57 Pointing out that the choice archi-
tects do not pursue selfish motives is hardly sufficient to alleviate the fears over such

 van Aaken 2015b, p. 10; Yeung 2012, p. 138, similarly distinguishes between ‘non-exploitative’
nudges that seek to ensure a rational, informed choice, and ‘expoitative’ nudges that “are designed
to alter behaviour by taking advantage of cognitive irrationalities”.
 van Aaken 2015a, p. 10; 2015b, p. 15.
 van Aaken 2015a, p. 18; 2015b, p. 14.
 For an overview see Lichtenstein and Slovic 2006.
 Thaler and Sunstein 2003, p. 175.
 Thaler and Sunstein 2008, p. 237.
 More careful Sunstein 2015, p. 38.
 Blumenthal-Barby 2013, p. 190.
 Yeung 2012, p. 134.
 Kelly 2013, p. 220.
7  Nudging and the Principle of Proportionality 101

Some ‘invisible’ nudges therefore potentially infringe on personal autonomy,

and if one subscribes to the notion that Art. 2(1) Basic Law in essence protects
autonomy, interfere with a fundamental right. But which nudges are truly ‘invisi-
ble’? Van Aaken considers a nudge invisible if the decision maker is not aware of
the effect of the nudge on his or her choice, because the cues are processed by the
automatic ‘System I’ of thinking.59 I argue that this is not sufficient to consider a
nudge as interfering with personal autonomy. To illustrate with an example, in a
widely cited study a picture of eyes was put above the coffee machine in an office
environment in which the payment for the coffee relied on an honour system, i.e.,
users of the coffee machine were expected to pay into a coin box next to the machine,
but the machine would dispense coffee irrespective of payment. The result of the
field experiment was that in weeks where a poster of human eyes was displayed,
voluntary contributions were almost three times as high as in weeks when a poster
of flowers was displayed.60
This nudge is invisible in the sense that the coffee drinkers were not aware that
the poster altered their behaviour. But can it be considered interfering with auton-
omy? To say so, one would have to argue that the ‘true self’ is revealed in the pres-
ence of a picture of flowers, or a naked wall, while in the presence of the poster of
the eyes, the person is not true to himself or herself. If one accepts that no choice is
context-free and that there is therefore no ‘unbiased’ environment that reveals the
true preference of the individual, this argument is hard to sustain. Accepting the
context-dependence of any choice means that most changes to the choice architec-
ture cannot be considered an interference with personal autonomy. It is not suffi-
cient that the decision maker is unaware of the (intended) effect of the manipulation;
he or she has to be unaware of the manipulative cues – such as in the case of sub-
liminal advertising or the subtly distorting mirrors – that one can consider the nudge
an interference with personal autonomy. The nudges suggested by Sunstein and
Thaler do not fall into this category. Such nudges do not interfere with the right ‘to
do as one pleases’ even if one accepts that the state must respect the individual’s
autonomy. Changing the default for organ donation from opt-in to opt-out does not
affect an individual’s autonomy61; it merely assigns the burden differently.62 Pictures
of people dying of cancer on cigarette packaging are not an interference with auton-
omy since there is no ‘emotion free’ preference for smoking. Replacing spoons with
tongs – certainly manipulative in the sense that most people won’t notice the choice
architecture – does not affect people’s preference for certain foods. It increases,
marginally, the effort required to fulfil those preferences. But having to reach twice
for the same amount of cubes cannot be considered an interference with personal
autonomy even under the most stringent standard. And providing a (large) spoon is
just as ‘manipulative’ as providing tongs, as it forces people that wish to select only

 van Aaken 2015a, p. 11; 2015b, pp. 13 et sqq.
 Bateson et al. 2006.
 As long as it is reasonably easy to opt-out. For a view of default options as effective mandates see
Bubb and Pildes 2014, pp. 1616 et sq.
 For a different view see van Aaken 2015a, p. 10.
102 M. Schweizer

one or two cheese cubes to be extra careful. The only truly context free choice
would be to provide no serving utensils and ask students to bring their own.
To summarize, the provision of purely factual information will enhance rather
than diminish personal autonomy. Nudges aimed merely at increasing the aware-
ness and information of citizens do not lead to constitutional concerns. Reminding
students to enrol in higher education by sending them text messages,63 for example,
is no interference with personal autonomy. Educational campaigns aimed at chang-
ing social norms, on the other hand, can indirectly interfere with fundamental
right(s). Changes to the decision context intended to influence choice will not gen-
erally interfere with personal autonomy as long as they are not imperceptible in the
sense that not only their effect is unnoticed, but the manipulated cue as such is
unnoticeable. Most of the commonly discussed forms of nudging therefore leave
the decision maker’s autonomy unaffected and do not trigger constitutional
However, as the example with the tobacco health warning demonstrates, many
nudges will coerce actors other than the decision maker, in this case tobacco com-
panies. Unsurprisingly, the German Supreme Court held that mandated health
warnings interfere with the economic rights of Art. 12(1) Basic Law (but came to
the conclusion that the interference was proportional).64 Many nudges are coercive
for actors other than the decision maker. Informing home owners that they use more
energy than their neighbours may induce them to use less energy without using clas-
sic (financial) incentives,65 but forcing energy providers to supply the information
interferes with their economic freedom. A 14 day cooling off period that allows
rescinding any contract entered into online may be a nudge from the consumers’
perspective,66 but it is certainly coercive on the vendors that have to accept it.
Many nudges will therefore interfere with fundamental rights not of the decision
maker, but other citizens. They will therefore trigger constitutional scrutiny, which
leads to the first prong of the proportionality test: do they serve a legitimate end?

7.5  Applying the Principle of Proportionality to Nudging

7.5.1  Legitimate Public Interest and Nudging

The first prong of the proportionality test requires the identification of a legitimate
public interest (‘Gemeinwohlziel’) that is served by the proposed measure. The
legitimate public interest may itself be derived from the constitution, but does not

 Castleman et al. 2014.
 BVerfGE 95, p. 173 – ‘warnings on cigarette packaging’. The plaintiffs also argued that their
freedom of speech was affected, the Constitutional Court denied this, id. at 181.
 Allcott 2011.
 According to van Boom and Ogus 2010, p. 2. I am not even sure I would call this a nudge at all.
7  Nudging and the Principle of Proportionality 103

necessarily have to be.67 It can be any economic, cultural or social purpose. The
protection of minors, minorities, socially disadvantaged persons, public health, the
environment or the advancement of families or gender equality are often cited
examples of legitimate public interests.68 Goals that are unconstitutional, such as
discriminatory purposes, are obviously not legitimate.69
Measures are considered paternalistic if they interfere with the liberty or auton-
omy of a person without that person’s consent and do so just because the measure
will improve the person’s welfare.70 According to the prevailing German doctrine,
an interest is not a ‘public’ interest if the goal of the measure is paternalistic in this
sense.71 The measure must rather be aimed at preventing social harm, or in modern
economic parlance, a negative externality. Paternalistic motives cannot sustain an
interference with a fundamental right except when the person protected from self-­
inflicted harm is a minor or mentally handicapped adult.72 In its early case law, the
Federal Constitutional Court has explicitly endorsed this view. It held that the state
does not have the task of ‘educating’ its citizens and therefore no right to interfere
with their autonomy merely to ‘better’ them.73 A ‘missionary’ purpose is not a legit-
imate purpose for interfering with fundamental rights.
This does not bode well for paternalistic nudges, which is a rather large sub-­
category of nudges.74 The fact that most nudges will only trigger constitutional scru-
tiny because they interfere with the liberty or autonomy of actors other than the
person being nudged does not make them non-paternalistic. Measures that interfere
with the autonomy of one person to protect another person from self-inflicted harm
are ‘indirectly paternalistic’.75 Their ultimate motive lies in the protection of the
other person from self-harm. If paternalistic motives are insufficient to justify the
interference with fundamental rights of the person to be protected from self-harm,
they are (all the more) insufficient to justify interference with the fundamental rights
of other persons.76 A full prohibition of the sales of tobacco products based on pater-
nalistic motives could be challenged not only by smokers, but also by the suppliers

 Hufen 2014, p. 119; Ipsen 2014, para. 187; unless the fundamental right intefered with is an
‘unconditional’ right (ohne Gesetzesvorbehalt), in which case it may only be intefered with based
on competing constitutional right(s), Michael 2001b, p. 655. For a different view see Fischer 1997.
 e.g., Hufen 2014, p. 119.
 Instead of many Sodan and Ziekow 2014, p. 229.
 Dworkin 2013, p. 29.
 Hillgruber 1992, p. 119; Fischer 1997, p. 273; van Aaken 2006, p. 136. Möller 2005 accepts
paternalistic motives as legitimate if the measure is aimed at preserving integrity.
 The protection from self-harm of people considered incapable of fully using their mind is con-
sidered autonomy-preserving, van Aaken 2015b, p. 3. ‘Soft paternalistic’ measures are widely con-
sidered legitimate and will not be considered further here.
 BVerGE 22, 180, pp. 219 et sq. – ‘youth welfare’.
 Not all nudges are paternalistic, of course. Non-paternalistic nudges will generally pass the legit-
imate aim prong, van Aaken 2015a, p. 20.
 Feinberg 1986, pp. 9 et sq.
 Enderlein 1996, p. 136.
104 M. Schweizer

of tobacco products, on the basis that paternalistic motives cannot justify the inter-
ference with a fundamental right.
The case law of the Federal Constitutional Court shows, however, that it does not
stringently apply the condition that the measure must be aimed at preventing a
social harm.77 In later cases, it held that ‘generally’, an interference with a funda-
mental right must be aimed at the protection of others,78 leaving open the possibility
that exceptionally, the protection from self-inflicted harm may be a legitimate pur-
pose.79 It upheld helmet laws and seatbelt laws,80 and laws curtailing sports bet-
ting.81 In the former case, one may argue that it is the public that bears the health
care costs in a country that knows compulsory health insurance, and in the latter
case, it is the public that through mandatory social security supports the gambling
addict that is unable to feed him or herself. These arguments are, however, hardly
convincing.82 Not primarily because they would justify almost any paternalistic
intervention, but because mandatory health insurance and social security are them-
selves paternalistic interferences with personal freedom. There is no opt-out of
health insurance or social security in Germany. Everybody who is able to must
contribute to them. The benefit of basic health care is forced upon the individual for
his or her own good. It seems hard to justify further paternalistic interventions based
on a first paternalistic intervention that spreads the consequences of self-inflicted
harm onto other citizens. In any case, harm to third parties by way of social security
could be easily avoided by the less restrictive alternative of merely excluding cer-
tain harms from coverage, e.g., compulsory health insurance would not pay for
injuries sustained because the injured failed to wear a helmet while riding a
In its helmet and seat belt law decisions, the Federal Supreme Court has right-
fully not only relied on the indirect social consequences of not wearing a helmet or
a seat belt, but sought direct social harm. In case of the motorcyclist not wearing a
helmet it pointed out that he may be unable to help others after having caused an
accident, while the helmet-wearing rider may remain conscious and able to help.83
The driver not wearing a seat belt may be thrown from the car and collide with
bystanders.84 These reasons seem rather strained, and they show to which argumen-
tative lengths the Federal Supreme Court goes to avoid accepting a purely paternal-
istic motive as legitimate public interest. A fair summary of the Federal Supreme
Court’s case law with regards to measures aimed at protecting the life and health of
citizens from self-harm is that it accepts paternalistic motives as a legitimate i­ nterest

 van Aaken 2015a, p. 20.
 BVerfGE 90, 145, p. 172 – ‘cannabis’.
 In BVerfGe 60, 123, p. 132 – ‘transgender I’, it even held without further elaboration that protec-
tion from ‘serious’ self-inflicted harm is a legitimate end.
 BVerfGE 59, p. 275 – ‘motorcycle helmet’; BVerfG, NJW 1987, p. 180 – ‘seat belt law’.
 BVerfGE 115, p. 276 – ‘sports betting’.
 But see Ueltzhöffer 2005, p. 17.
 BVerfGE 59, p. 275 – ‘motorcycle helmet’.
 BVerfGE, NJW 1987, p. 180 – ‘seat belt law’.
7  Nudging and the Principle of Proportionality 105

if the averted harm is very severe and irreversible, namely severe bodily harm or
death. Doctrinally, this has been justified by the argument that personal autonomy
includes an objective dimension of integrity, i.e., the state has a duty to protect a
person’s integrity from irreparable harm.85
Most nudges will not, however, fall into this category. The harm from not saving
enough for retirement is hardly on the same level as smashing one’s brain through
a windshield. Even health nudges – a very popular sub-set of nudges – are for the
most part aimed at preventing long-term health issues resulting from, for example,
obesity. While coronary heart diseases and diabetes are serious and potentially
deadly afflictions, they do not fall into the category of serious bodily harms that the
Federal Supreme Court has so far held to justify paternalistic interventions. No mat-
ter what ethical justifications there are for such nudges, it is hard to square paternal-
istic nudges that trigger constitutional scrutiny with the requirement that any
interference with personal autonomy must serve a public good, i.e., prevent social
Another area where the Federal Constitutional Court has grappled with the legit-
imate means test is consumer protection. Consumer protection measures are mostly
aimed not at preventing bodily harm, but rather financial harm to consumers.
Preventing someone from entering into an agreement that is disadvantageous for
him or her seems like a classic case of paternalism. There is a loss of freedom when
consumers are prevented from, for example, buying consumer goods from commer-
cial vendors without a 2 year period to claim defects.87 A well informed consumer
may rationally prefer a cheaper good without warranty to the more expensive good
with a warranty.88
The Federal Constitutional Court has relied on two lines of arguments to uphold
potentially paternalistic consumer protection regulation in the face of constitutional
scrutiny. Firstly, it is argued that there often is a structural imbalance between con-
tracting parties which leaves one party essentially at the mercy of the other. Under
circumstances where one party can in reality dictate the terms of the agreement
there is no free choice, the weaker party – typically the consumer or employee – is
forced to accept the offer as it is. The court then considers it justified to restrict
freedom of contract to prevent exploitation.89 Arguably, this is not protection from
self-inflicted harm, but protection from harm by third parties.
Secondly, the court argues that the inexperience of the decision maker with the
specific transaction and its associated risks or social pressure may justify cogent

 Möller 2005.
 van Aaken 2015a, p. 21.
 Art. 437(3) in connection with Art. 474 Bürgerliches Gesetzbuch.
 Warranty is used in the sense of ‘Gewährleistung’, i.e., the vendor must only warrant that the
good was not defective when the risk passed. If the defect occcurs within 6 months after the pass-
ing of the risk, it is presumed that it existed when the risk passed, Art. 476 Bürgerliches Gesetzbuch.
 BVerfGE 81, 242, p. 255 – ‘sales agents’; BVerfGE 89, 214, p. 232 – ‘debt guarantee’.
106 M. Schweizer

private law.90 These motives are hard to interpret as anything but paternalistic.91
One cannot shake the impression that while in principle and in theory, paternalistic
motives are not accepted as legitimate ends for interfering with a fundamental right,
in practice the Federal Constitutional Court accepts paternalistic motives in the con-
text of consumer protection and employee protection.92
The conclusion so far is therefore that nudges that trigger constitutional scrutiny
must in principle serve other than purely paternalistic goals to pass the legitimate
means test of the proportionality principle, unless they are aimed at preventing
irreparable harm to a person’s integrity. The case law of the Federal Constitutional
Court namely in the area of consumer protection law indicates that the court does
not apply the prohibition of paternalistic motives strictly, however.

7.5.2  Suitability and Nudging

A means interfering with a fundamental right in the interest of a legitimate public

good is suitable when it is, at least in principle, possible that it furthers the stated
public goal.93 It is not required that it is the optimal means to further the public
interest,94 or that it serves all the stated public interests if more than one is pleaded.95
It is also not required that the state shows that the means demonstrably furthers the
public interest, the ‘abstract potential’ that it helps achieving it is enough.96
Unsurprisingly, hardly any legislative measure ever fails this prong of the propor-
tionality principle.97
Measures that change the context of a choice to nudge the decision maker
towards a choice preferred by the policy maker will pass this prong even if their
effectiveness is potentially more limited than early advocates of nudging cared to
admit (see next section). The evidence that nudging works to some degree is over-
whelming, and that is all that is required under the suitability test. The fact that not
all decision makers make the desired choice is irrelevant as long as more decision
makers make the desired choice after the intervention than before the intervention.
Nudging is therefore, like any not wholly irrational legislative measure, suitable in
the sense of the principle of proportionality.

 BVerfGE 89, 214, p. 235 – ‘debt guarantee’; BVerfGE NJW 1999, pp. 2749 et sqq.; BVerfGE 60,
329, p. 345 – ‘pension adjustment’.
 Enderlein 1996, p. 143.
 See also van Aaken 2015a, p. 21, who considers the Constitutional Court’s jurisprudence ‘eva-
sive’ and ‘less rigorous’ than the doctrine.
 BVerfGE 115, 267, p. 308 – ‘old debts’.
 BVerfGE 30, 292, p. 316 – ‘oil reserves’.
 Michael 2001b, p. 656.
 BVerfGE 67, 157, p. 175 – ‘surveillance law’. For a critique see Bickenbach 2014, pp. 320 et sq.
and the references cited therein.
 Hufen 2014, p. 119.
7  Nudging and the Principle of Proportionality 107

7.5.3  Nudging as the Least Restrictive Alternative?

A means is required (or necessary) when there is no other means available that is
equally suitable to further the public interest but less infringing on the constitution-
ally protected right(s).98 When there are several equally effective means to achieve
a given objective the state is required to choose the ‘mildest means’, i.e., the inter-
vention least restrictive of personal liberties.99 If there is doubt whether the pro-
posed measure or a less restrictive alternative are equally effective, the less
restrictive measure must be chosen.100 A measure is less restrictive if it interferes
less intensely with a given fundamental right, or interferes with a less weighty fun-
damental right, or affects less people.101 The legislator is not obliged to choose the
less restrictive alternative, even if it is considerably less restrictive, if it serves even
one of the stated public interests marginally less than the proposed policy.102 On the
other hand it is not required that the least restrictive alternative does not interfere
with any fundamental right at all, it is sufficient that it interferes less.103
A nudge will generally be less restrictive than a coercive measure. This even
holds for nudges that are coercive on actors other than the ultimate decision makers,
because the coercive alternative will often also coerce those actors. Take the exam-
ple of health warnings on cigarette packaging. These warnings – clearly intended to
influence smokers to quit – are burdensome on tobacco companies. A traditional
financial incentive with a similar effect is increasing the tax on tobacco products.
But if this measure is as effective as health warnings (in reducing smoking), it is no
less coercive on the tobacco companies, and arguably more coercive on smokers.104
The same goes for forcing the operators of cafeterias to provide tongs rather than
spoons for unhealthy food items. The coercive alternative of banning certain foods
altogether, or increasing their price by taxation, is more coercive on the consumer,
and equally coercive or more coercive on the operator. Changes to the choice archi-
tecture will therefore generally pass the necessity prong of the proportionality
Van Aaken argues that from a set of potential non-coercive interventions, the
government should choose transparent nudges that rely on information and raising
awareness, rather than nudges that work on a subconscious level.105 As desirable as

 BVerfGE 67, 157, p. 176 – ‘surveillance law’.
 BVerfGE 53, 135, p. 145 – ‘chocolate easter bunny’; Sodan and Ziekow 2014, p. 230.
 Schlink 2001, p. 458, argues that the burden of proof that the measure is required lies with the
state, therefore, if there remain doubts, the measure is not justified.
 Michael 2001b, pp. 656 et sqq.
 Michael 2001b, p. 657 and Michael and Morlok 2014, p. 308.
 Michael and Morlok 2014, para. 620.
 Health warnings on cigarette packaging may be covered by special sleeves, thereby preventing
exposure to the warning. There is no legal way to avoid paying tobacco tax, however. The provision
of sleeves by tobacco companies may be illegal, but nobody seems to argue that the end consumer
is prohibited from using such sleeves, Wilson et al. 2006.
 van Aaken 2015b, pp. 31 et sqq.
108 M. Schweizer

this may be from an ethical perspective seeking to maximize autonomy, legally, it

only follows from the proportionality principle if the transparent, information-based
measure is as effective as the subconscious nudge. This is at least doubtful.106 It
would have to be empirically shown that, for example, providing the patrons of a
cafeteria with information on the risks and benefits of certain food items is as effec-
tive as placing those food items at the back of the line, or that letting the users of a
coffee machine in an office know that without their contribution, no more coffee
can be bought, increases donations as much as hanging a picture of eyes next to the
The more interesting question is whether the state may actually be forced to
choose a non-coercive measure over a coercive measure under this prong. This is
the case if the non-coercive measure is as effective as the proposed coercive mea-
sure. While early proponents were convinced that nudging is an effective way of
changing behaviour, often as effective as coercive approaches,107 empirical evi-
dence casts doubt on this.108 After the UK government famously installed a
‘Behavioural Insights Team’ – a small team of civil servants and academics based in
the Cabinet Office tasked with advising the British government in non-coercive
interventions to change public behaviour – the House of Lords Select Committee on
Science evaluated the effectiveness of non-coercive forms of behaviour change in a
2011 report.109 The Committee heard a large number of witnesses, many of them
academics both engaged in primary research as well as those tasked with changing
behaviour at government agencies and private companies.110 The report came to the
conclusion that nudges – a subset of what the report describes as “non-regulatory
interventions”111 – “would be unlikely to have a significant effect [on behaviour] if
used in isolation.”112 The Select Committee on Science “urge[d] ministers to ensure
that policy makers are made aware of the evidence that non-regulatory measures are
often not likely to be effective if used in isolation and that evidence regarding the
whole range of policy interventions should be considered before they commit to
using non-regulatory measures alone.”113
Lauren E. Willis specifically examined the effectiveness of default rules to
induce behaviour change. She notes that while there is evidence that automatic

 Englerth 2007, p. 239.
 Thaler and Sunstein 2008, p. 8.
 Marteau et al. 2011; Rayner and Lang 2011; Salazar 2012; Costa-Font et al. 2014; Bubb and
Pildes 2014, pp. 5 et sq.; Russell 2015, pp. 22 et sq.; Ito et al. 2015.
 House of Lords Science and Technology Select Committee 19.07.2011, see id. Box 7 at para.
5.3 for the role of the Behavioural Insights Team. Today, the Behavioural Insights Team is a private
enterprise – partially owned by the UK government – “whose mission is to help organisations in
the UK and overseas to apply behavioural insights in support of social purpose goals”, see www. (last visited 8 February 2016).
 See list of witnesses in House of Lords Science and Technology Select Committee 19.07.2011,
Appendix 2.
 House of Lords Science and Technology Select Committee 19.07.2011, Table 1 at para. 2.3.
 House of Lords Science and Technology Select Committee 19.07.2011, paras. 5.8, 8.14.
 House of Lords Science and Technology Select Committee 19.07.2011, paras. 5.15, 8.16.
7  Nudging and the Principle of Proportionality 109

(default) enrolment in a retirement savings plan dramatically increases the partici-

pation rate, this is not the case for many other defaults intended to protect consum-
ers from self-harm. She attributes the difference primarily to the incentives of the
private actors. When defaults are deployed in an environment where one side of the
bargain – usually the better informed side – has an interest to change the default, the
default turns out to be not very sticky at all. Worse yet, those consumers who opt
out of the default option are often the ones that would benefit the most from it,
because conversely, the contractual partner has the strongest incentive to make
them switch.114 Willis concludes that “where a party with the means to influence
consumer choice opposes them, government nudges are likely to fail.”115 To be
effective in the face of opposition, nudges “must be supported by a web of regula-
tion and government marketing so strong that the cost of cutting through that web
exceeds the benefits to firms of doing so”,116 which makes them effectively as bur-
densome as traditional coercive regulation. In the same vein, Bronchetti et al. find
that default options are not sticky in the face of pre-determined spending plans by
low-income tax filers.117 Similar to Willis, Bubb and Pildes argue that more than
nudges are needed to achieve the stated policy goals of increasing retirement sav-
ings, protecting consumers from predatory lending practices and improving fuel
economy.118 They accuse proponents of nudging of deliberately avoiding drawing
the full conclusions from the body of (psychological) research they are building on,
which would lead them to endorse traditional regulatory interventions that limit
freedom of choice, to offer a political third way – ‘libertarian paternalism’ – that
generates political consensus.119
Both empirical evidence as well as theoretical considerations – namely taking
into account the strategic behaviour of market participants – therefore puts the
effectiveness of freedom of choice preserving interventions in furthering public
policy goals in doubt. Since under the necessity test of the principle of proportional-
ity a less restrictive alternative only has to be preferred if it is equally effective as
the proposed measure in all regards, the legislator will rarely be compelled to choose
nudging over more traditional regulatory means. Perhaps unsurprisingly,120 it all
comes down to the last and most controversial prong of the principle of proportion-
ality, balancing proper.

 Willis 2013, p. 1186, p. 1228.
 Willis 2013, p. 1161.
 Willis 2013, p. 1229.
 Bronchetti et al. 2011.
 Bubb and Pildes 2014, pp. 1647 et sqq. and more often.
 Bubb and Pildes 2014, pp. 1596 et sqq.
 Classen 2012, p. 652; Möller 2014, p. 156.
110 M. Schweizer

7.5.4  Balancing and Nudging

The last prong of the proportionality test requires the interference of the proposed
intervention with the fundamental right(s), and its other undesired effects, to be
weighed against the importance of the public interest the means seeks to achieve
(‘Verhältnismässigkeit im engeren Sinn’, ‘Angemessenheit’ or ‘Zumutbarkeit’).121
Balancing requires that the “weight of the interference must not be disproportionate
to the weight of its justifying reasons”.122 The legislator must put the abridged indi-
vidual right and the public interest served by the intervention into perspective. The
balancing test may result in the finding that a suitable and required means is uncon-
stitutional because the interference with the fundamental right(s) weighs more
heavily than the interests the measure seeks to protect.123 The less severe the inter-
ference with fundamental rights is, the more easily it can be accepted in view of the
protection of important goals by the measure.124
If this all sounds a bit vague to you, you are not alone. The Federal Supreme
Court acknowledges the fuzziness of the balancing test insofar as only dispropor-
tionate measures are unconstitutional.125 In other words, the balancing test does not
require that the measure is proportionate; it only requires that it is not disproportion-
ate. This is meant to preserve the legislator’s freedom to choose among several
suitable and required measures (‘Einschätzungsprärogative’).126 It is nonetheless
evident that the balancing test is a value judgment much more than the suitability
and necessity prongs of the proportionality test, which require an empirical assess-
ment of the proposed measure’s impact.127 For critics of the balancing test, the
courts overreach into the proper domain of politics when they engage in unstruc-
tured balancing.128 The weighing of interests to reach a compromise is a typical
activity of legislators and should be reserved for democratically elected
Considerable doctrinal effort has therefore been spent on structuring the balanc-
ing test and making it more predictable. The best known of these efforts is probably
Robert Alexy’s weight formula.129 In a partially simplified form, the weigh formula
takes into account the interests (or principles) to be balanced:

 Hufen 2014, p. 121; Sodan and Ziekow 2014, p. 230.
 BVerfGE 118, 168, p. 195 – ‘basic account data’: “Dieses Gebot verlangt, dass die Schwere des
Eingriffs bei einer Gesamtabwägung nicht außer Verhältnis zu dem Gewicht der ihn rechtfertigen-
den Gründe stehen darf”.
 BVerfGE 115, 320, pp. 345 et sq. – ‘dragnet search II’.
 BVerfGE 118, 168, p. 195 – ‘basic account data’.
 BVerfGE 100, 226, p. 242 – ‘heritage site preservation’.
 Zippelius and Würtenberger 2008, pp. 124 et sqq.; Ipsen 2014, para. 195; Michael and Morlok
2014, para. 626.
 Schlink 2001, p. 458.
 Schlink 2001, pp. 461 et sqq.; Pieroth 2014, para. 303; and the references cited in Saurer 2012,
p. 31.
 Alexy 2003, pp. 783 et sqq.; for an English exposition see Jestaedt 2012, pp. 155 et sqq.
7  Nudging and the Principle of Proportionality 111

a = interest pursued by the interference with a fundamental right (aim of the

b = interest burdened by the interference
The further elements of the weight formula are W = weight (or value) of the
interest (‘the importance of satisfying the interest’), P = probability of interference
and I = intensity (or gravity) of interference. The weight formula then reads

Wa Pa I a Ga
Ga ,b =
Wb Pb I b Gb

A measure is disproportionate if Ga,b <1. Put into words, the weight formula requires
taking into account the weight of the fundamental right, the intensity and probabil-
ity of its infringement by the proposed measure on the one hand, and the weight of
the public interest to be served, the probability and degree of its achievement by the
proposed measure on the other side, and finds the measure disproportionate if the
latter is larger than the former. Besides the difficulty of expressing such abstract
notions as the ‘weight of an interest’ as real numbers – which has to be possible
unless one considers the interests incommensurable130 – the problem with the
weighing formula is that it explains a lot, but predicts little.131 Few would disagree
that the weight formula identifies correctly the parameters required for proper bal-
ancing, but it gives no guidance how the proper weight is assigned to various inter-
ests, or principles, in specific cases.132
Alexy and his supporters posit that it is at the least possible to put the affected
interests into three categories of ‘little’, ‘intermediate’ and ‘most’ worth protect-
ing.133 Applied to fundamental rights, the freedom of personal development (Art.
2(1) Basic Law) is least protected, while more specific rights such as the right to life
(Art. 2(2)), freedom of opinion (Art. 5(2)) or the privacy of correspondence and
telecommunication (Art. 10(1)) fall into the second category. The highest category
is reserved for unconditional134 rights, such as freedom of religion (Art. 4(1) Basic
Law). In the same way the public interest(s) the measure seeks to further can be
categorized, with the protection of constitutional goods having the most weight. In
this category fall not only the protection of the fundamental rights of others, but also
the protection of constitutionally mandated public goods, such as the protection of
the environment (Art. 20a Basic Law). More controversially, it is claimed that other

 Alexy 2003, p. 781, argues that the German constitution, by allowing the inteference with fun-
damental rights based on parliamentary act, assumes commensurability of rights and public inter-
ests. For a more pessimistic view see Endicott 2014.
 Klement 2008, p. 761.
 Jestaedt 2012, pp. 164 et sqq.
 Alexy 2003, p. 773; Wienbracke 2013, p. 153; Michael and Morlok 2014, para. 624.
 ‘Vorbehaltlose Grundrechte’ in German. In this context this means that the German Constitution
does not expressely state that these rights may be abridged by a law, as it does for many fundamen-
tal rights.
112 M. Schweizer

public interests can be categorized according to their weight, too.135 Finally, the
epistemic probability of the measure having the desired effect or the feared interfer-
ence with the fundamental right can be divided into the three categories of certainty,
plausible and ‘not evidently false’.136
Given the vagueness of the balancing test it is difficult to assess on an abstract
level how nudging interventions will fare. Nudges will often only interfere with the
economic freedom (‘Berufsfreiheit’, Art. 12 Basic Law) in the sense of the
‘Berufsausübungsfreiheit’, i.e. the freedom of how to specifically conduct a gainful
activity, of those having to implement the nudge, rather than completely prohibiting
certain gainful activities.137 This economic freedom is afforded little weight unless
the exercise of a profession is completely forbidden,138 which would indicate that
most interventions in choice architecture will pass the balancing test. Since the bal-
ancing test also requires taking into account the weight of the public interest pur-
sued by the measure, a more definite answer is impossible without knowing which
public interest the nudge seeks to further. The fact that the state chose a non-­coercive
and therefore most likely less effective measure to further a public may by itself
indicate that the public interest pursued is not very important.139 Nonetheless it is
hard to see how a nudge that serves a legitimate public interest and is both suitable
and required would fail the balancing test unless the sub-set of people that would
exhibit the undesired behaviour absent the nudge is so small that even the relatively
minor inconvenience imposed by certain nudges on the overwhelming majority of
citizens outweighs the benefits to those whose behaviour is changed by the
The balancing test, both as proposed by Alexy as well as practiced by the Federal
Constitutional Court, does not take into account the availability of less restrictive
alternatives. The relevant question is only whether the interference of the proposed
measure with a fundamental right is proportionate in view of the harm that likely
occurs in the absence of the measure. Therefore, the balancing test cannot force the
legislator to choose a non-coercive intervention over a proposed coercive measure.

7.5.5  Consistency and Nudging

In recent times, the Federal Constitutional Court has explicitly referred to the lack
of consistency (‘Folgerichtigkeit’) of a proposed measure when judging its propor-
tionality.140 When implementing measures protecting a public interest, the legislator
must not create exceptions without a rational basis, or in other words, the proposed

 Alexy 2003, pp. 780 et sqq.; Michael and Morlok 2014, para. 624 (p. 310 at the top).
 Alexy 2003, p. 789.
 The scope of protection of the ‘Berufsfreiheit’ is far reaching, see Ipsen 2014, paras. 634 et sq.
 Ipsen 2014, paras. 656 et sq.
 Payandeh 2011, p. 607 (for inconsistent legislation).
 BVerfGE 115, 267, p. 310 – ‘sports betting’; BVerfGe 121, 317, pp. 362 et ssq – ‘smoking in
7  Nudging and the Principle of Proportionality 113

measure must be consistent with its stated aim. The ‘consistency requirement’ had
been developed in the court’s tax jurisprudence, where it is based on the principle of
equality (Art. 3 Basic Law) and requires taxation according to the subject’s eco-
nomic capacity.141 Its application in the framework of the proportionality test for the
interference with fundamental rights has been greeted with a healthy dose of scepti-
cism if not outright hostility. It was pointed out that requiring consistency in the
interference with fundamental rights may lead to the absurd consequence that a
measure that interferes more substantially with a fundamental right, or the funda-
mental right(s) of more persons, may be proportional, while a less restrictive mea-
sure may be considered disproportionate.142 These concerns are mostly based on the
specific reasoning of the court in its decision regarding the constitutionality of
smoking bans in bars, restaurants and the like.143 The specific legislation permitted
separate smoking rooms if the inn also offered a non-smoking room. Small corner
bars were unable to benefit from the exemption – since they only had one room,
which room had to be smoke-free. The court considered that the proposed legisla-
tion did not pass the balancing test because it had a disproportionate impact on
small bars, which would lose customers to larger establishments where smoking
remained possible.144
At first blush, the requirement of consistency may put choice architectural mea-
sures at risk. By their nature non-coercive, they are inconsistent in the sense that
they allow interested subjects avoiding the intended behavioural change. If the gov-
ernment really believed that certain harms should be averted, should it not make
sure that they are actually averted, rather than leaving compliance up to the
A closer look at the consistency requirement puts these doubts to rest, however.
In practice, it serves two purposes. On the one hand, the consistency requirement is
simply an application of the principle of equal treatment specifically to competi-
tors.145 While regulations that differ based on the facts of the case rather than the
person are generally compatible with the equal treatment requirement of Art. 3
Basic Law, if a facts-based distinction disparately impacts certain competitors, their
economic freedom is infringed.146 Nudges as such will not lead to unequal treatment
of competitors; they equally affect everybody in a given market. They may if they
are combined with other regulations – such that only companies above a certain size
must implement them – but the disparate impact then results not from the nudge as
such, but rather from the circumstances of its implementation.
On the other hand, inconsistent legislation is a signal that the state has hidden
motives other than the protection of a legitimate public interest.147 When regulating

 Bumke 2010, pp. 87 et sq.
 BVerfGe 121, 317, p. 385 – ‘smoking in bars’ (Mansing dissenting).
 Payandeh 2011, p. 606.
 BVerfGE 121, 317, pp. 355 et sq. – ‘smoking in bars’.
 Michael 2008.
 Michael 2008, p. 878.
 Petersen 2013.
114 M. Schweizer

sports betting, for example, the government may restrict economic freedom to pro-
tect gambling addicts from financial ruin. It may not, however, restrict economic
freedom to protect its fiscal interests by exempting state owned entities from regula-
tions that apply to private outfits.148 The regulation therefore has to be consistently
aimed at protecting gambling addicts to be constitutional.149 The fact that the gov-
ernment chooses a nudge over more traditional coercive regulation may indicate
that the true motive is paternalistic. Nudges are tempting because they seem to
interfere little with fundamental rights. If the government was fully convinced that
a legitimate public interest was at stake, coercive measures could easily be justified
if the only fundamental rights at stake are personal and economic freedom. If the
court comes to the conclusion that the end of a choice architectural intervention is
purely paternalistic, however, the proposed regulation should fail the first prong of
the proportionality principle, rather than be considered inconsistent.
In sum, changes in the decision context are not inconsistent under the ‘consis-
tency requirement’ developed in the latest case law of the Federal Constitutional
Court. The fact that not all individuals exposed to the changed decision context will
change their behaviour in the expected way is a virtue, not a vice, of nudging.

7.6  Conclusion

Subjecting governmental non-coercive changes to a decision context designed to

influence human behaviour in a predictable way to the proportionality test raises a
number of interesting questions. Firstly, it has to be assessed whether a true ‘nudge’,
which does not limit the choice set, can be considered an interference with a funda-
mental right under the German Constitution. In this contribution I argue that most
nudges per se do not interfere with personal autonomy, as there is no truly ‘unbi-
ased’ context that would reveal the ‘true’ preference of the decision maker. All
choices are influenced by context in some way, and there is no right to a context-­
free choice. However, the implementation of nudges will often involve coercive
measures against persons other than the decision maker, and therefore trigger con-
stitutional scrutiny as they interfere with the personal freedom of those forced to
implement the nudge.
Secondly, nudges that are solely aimed at preventing the decision maker from
harming him- or herself do not serve a legitimate end that would justify the interfer-
ence with a fundamental right unless the harm prevented is serious bodily harm or
death. It does appear, however, that the Federal Constitutional Court allows
­paternalistic interventions in the domain of consumer and employee protection if
not in theory, then at least in practice.
Thirdly, most nudges – like any not wholly irrational intervention – will pass the
suitability prong of the proportionality principle. Fourthly, nudges will usually pass

 BVerfGE 115, 276, pp. 303 et sqq. – ‘sports betting’.
 BVerfGE 115, 276, pp. 309 et sqq. – ‘sports betting’.
7  Nudging and the Principle of Proportionality 115

the necessity prong as well, as there will generally not be any less restrictive alterna-
tives. It is namely doubtful that purely cognitive interventions aimed at providing
information to the decision maker are as efficient as measures that alter the decision
context. On the other hand, the legislator will rarely if ever be forced to choose a
nudge over a coercive measure under the necessity test, because nudges are hardly
ever as effective in changing behaviour as traditional prohibitions or
Fifthly, it is difficult to assess in the abstract whether nudges will pass the bal-
ancing test. Since most nudges will only interfere with personal freedom or eco-
nomic freedom, which are given little weight in the balancing against competing
(public) interests, nudges should generally fare well under the balancing test. On the
other hand, the fact that the government chooses nudging rather than a coercive
measure may indicate that it does not consider the interest to be furthered by the
measure to be very important.
Finally, nudges will generally be consistent in the sense of the ‘consistency
requirement’ recently employed by the Federal Constitutional Court also in the
assessment of the proportionality of interventions with fundamental rights. While
nudges will by their nature not affect every decision maker, this does not mean that
they are inconsistent in the sense that they violate the equal treatment clause of the
German Constitution because all the decision makers are exposed to the interven-
tion in the same way.


Alemanno, Alberto, and Alessandro Spina. 2014. Nudging legally: On the checks and balances of
behavioral regulation. International Journal of Constitutional Law 12(2): 429–456.
Alexy, Robert. 2003. Die Gewichtsformel. In Gedächtnisschrift für Jürgen Sonnenschein. 22.
Januar 1938 bis 6. Dezember 2000, ed. Jürgen Sonnenschein, Joachim Jickeli, Peter Kreutz,
and Dieter Reuter, 771–792. Berlin: De Gruyter Recht.
Allcott, Hunt. 2011. Social norms and energy conservation. Journal of Public Economics 95(9–
10): 1082–1095.
Arai-Takahashi, Yutaka. 2002. The margin of appreciation doctrine and the principle of propor-
tionality in the jurisprudence of the ECHR. Antwerp: Intersentia.
Bateson, Melissa, Daniel Nettle, and Gilbert Roberts. 2006. Cues of being watched enhance coop-
eration in a real-world setting. Biology Letters 2(3): 412–414.
Bickenbach, Christian. 2014. Die Einschätzungsprärogative des Gesetzgebers. Analyse einer
Argumentationsfigur in der (Grundrechts-)Rechtsprechung des Bundesverfassungsgerichts.
Tübingen: Mohr Siebeck.
Blumenthal-Barby, J.S. 2013. Choice architecture: A mechanism for improving decisions while
preserving liberty? In Paternalism. Theory and practice, ed. Christian Coons and Michael
Weber, 178–196. Cambridge, MA: Cambridge University Press.
Blumenthal-Barby, J.S., and Hadley Burroughs. 2012. Seeking better health care outcomes: The
ethics of using the ‘Nudge’. The American Journal of Bioethics 12(2): 1–10.
Bovens, Luc. 2009. The ethics of nudge. In Preference change. Approaches from philosophy, eco-
nomics and psychology, ed. Till Grüne-Yanoff and Sven Ove Hansson, 207–219. Dordrecht:
116 M. Schweizer

Bronchetti, Erin Todd, Thomas S. Dee, David B. Huffman, and Ellen Magenheim. 2011. When a
nudge isn’t enough. Defaults and saving among low-income tax filers, NBER Working Paper
Series (16887). National Bureau of Economic Research, Cambridge (Mass.), USA.
Bubb, Ryan, and Richard H. Pildes. 2014. How behavioral economics trims its sails and why.
Harvard Law Review 127: 1595–1678.
Bumke, Christian. 2010. Die Pflicht zur konsistenten Gesetzgebung. Der Staat 49: 77–105.
Camerer, Colin, Samuel Issacharoff, George Loewenstein, Ted O’Donoghue, and Matthew Rabin.
2003. Regulation for conservatives: Behavioral economics and the case for ‘Asymmetric
Paternalism.’ University of Pennsylvania Law Review, 151(3): 1211–1254.
Castleman, Benjamin L., and Lindsay C. Page. 2014. Summer nudging: Can personalized text
messages and peer mentor outreach increase college going among low-income high school
graduates? Journal of Economic Behavior & Organization, 115:144–160.
Christman, John. 2015 March 21. Autonomy in moral and political philosophy. http://plato.stan-
Classen, Claus Dieter. 2012. Das Prinzip der Verhältnismäßigkeit im Spiegel europäischer
Rechtsentwicklungen. In Der grundrechtsgeprägte Verfassungsstaat. Festschrift für Klaus
Stern zum 80. Geburtstag, ed. Michael Sachs and Helmut Siekmann, 651–667. Berlin: Duncker
& Humblot.
Cohen-Eliya, Moshe, and Iddo Porat. 2009. The hidden foreign law debate in Heller: The propor-
tionality approach in American Constitutional Law. San Diego Law Review 46: 367–415.
Costa-Font, Joan, David R. Just, Barbara Fasolo, and Nattavudh Powdthavee. 2014. Nudging our-
selves? The limits of incentivizing ‘Good Behavior’. Applied Economic Perspectives and
Policy 36(1): 1–5.
Cremer, Wolfram. 2003. Freiheitsgrundrechte. Funktionen und Strukturen. Tübingen: Mohr
Dayan, Eran, and Maya Bar-Hillel. 2011. Nudge to nobesity II: Menu positions influence food
orders. Judgment and Decision Making 6(4): 333–342.
Dworkin, Gerald. 2013. Defining paternalism. In Paternalism. Theory and practice, ed. Christian
Coons and Michael Weber, 25–38. Cambridge, MA: Cambridge University Press.
Enderlein, Wolfgang. 1996. Rechtspaternalismus und Vertragsrecht. München: C.H. Beck.
Endicott, Timothy. 2014. Proportionality and incommensurability. In Proportionality and the rule
of law. Rights, justification, reasoning, ed. Grant Huscroft, Bradley W. Miller, and Grégoire
C.N. Webber, 311–342. Cambridge, MA: Cambridge University Press.
Englerth, Markus. 2007. Vom Wert des Rauchens und der Rückkehr der Idioten – Paternalismus als
Antwort auf beschränkte Rationalität? In Recht und Verhalten. Beiträge zu behavorial law and
economics, ed. Christoph Engel, Markus Englerth, Jörn Lüdemann, and Indra Spiecker, 231–
258. Tübingen: Mohr Siebeck.
Feinberg, Joel. 1986. Harm to self. The moral limits of the criminal law. New York: Oxford
University Press.
Fischer, Kai. 1997. Die Zulässigkeit aufgedrängten staatlichen Schutzes vor Selbstschädigung.
Frankfurt am Main: P. Lang.
Fischer, Mira, and Sebastian Lotz. 2014. Ist sanfter Paternalismus ethisch vertretbar? Eine differ-
enzierende Betrachtung aus Sicht der Freiheit. Sozialer Fortschritt 63(3): 52–58.
Glaeser, Edward L. 2006. Paternalism and psychology. Regulation 29(2): 32–38.
Goodwin, Tom. 2012. Why we should reject ‘Nudge’. Politics 32(2): 85–92.
Hausman, Daniel M., and Brynn Welch. 2010. Debate: To nudge or not to nudge*. Journal of
Political Philosophy 18(1): 123–136.
Hill, Claire. 2008. Anti-anti-anti paternalism. New York University Journal of Law and Liberty 2:
Hillgruber, Christian. 1992. Der Schutz des Menschen vor sich selbst. München: F. Vahlen.
House of Lords Science and Technology Select Committee. 2011July 19. Behaviour Change.
Report. HL Paper 179. London.
7  Nudging and the Principle of Proportionality 117

Hufen, Friedhelm. 2014. Staatsrecht. München: Beck.

Ipsen, Jörn. 2014. Staatsrecht. Neuwied: Luchterhand.
Ito, Koichiro, Takanori Ida, and Makoto Tanaka. 2015. The persistence of moral suasion and eco-
nomic incentives. Field experimental evidence from energy demand, NBER Working Paper
Series (No. 20910).
Jestaedt, Matthias. 2012. The doctrine of balancing – strengths & weaknesses. In Institutionalized
reason. The jurisprudence of Robert Alexy, ed. Matthias Klatt, 152–172. Oxford/New York:
Oxford University Press.
Johnson, Eric J., and Daniel G. Goldstein. 2003. Do defaults save lives? Science 302:
Kelly, Jamie. 2013. Libertarian paternalism, utilitarism, and justice. In Paternalism. Theory and
practice, ed. Christian Coons and Michael Weber, 216–230. Cambridge, MA: Cambridge
University Press.
Kirchgässner, Gerhard. 2013. Sanfter Paternalismus, meritorische Güter, und der normative
Individualismus. In Grenzen der Konsumentensouveränität, ed. Martin Held, Gisela Kubon-­
Gilke, and Richard Sturn, 41–62. Marburg: Metropolis.
Klement, Jan Henrik. 2008. Vom Nutzen einer Theorie, die alles erklärt. Robert Alexys
Prinzipientheorie aus der Sicht der Grundrechtsdogmatik. JuristenZeitung: 63(15–16):
Koch, Oliver. 2003. Grundsatz der Verhältnismässigkeit in der Rechtsprechung des Gerichtshofs
der europäischen Gemeinschaften. Berlin: Duncker & Humblot.
Lichtenstein, Sarah, and Paul Slovic. 2006. The construction of preference. An overview. In The
construction of preference, ed. Sarah Lichtenstein and Paul Slovic, 1–40. Cambridge/New
York: Cambridge University Press.
Lüdemann, Jörn. 2004. Edukatorisches Staatshandeln. Steuerungstheorie und Verfassungsrecht
am Beispiel der staatlichen Förderung von Abfallmoral. Baden-Baden: Nomos.
Marteau, T.M., D. Ogilvie, M. Roland, M. Suhrcke, and M.P. Kelly. 2011. Judging nudging: Can
nudging improve population health? BMJ (British Medical Journal) 342: d228.
Mensink, Ronald P., Peter L. Zock, Arnold D.M. Kester, and Martijn B. Katan. 2003. Effects of
dietary fatty acids and carbohydrates on the ratio of serum total to HDL cholesterol and on
serum lipids and apolipoproteins: A meta-analysis of 60 controlled trials. The American
Journal of Clinical Nutrition 77(5): 1146–1155.
Micha, Renata, Sarah K. Wallace, and Dariush Mozaffarian. 2010. Red and processed meat con-
sumption and risk of incident coronary heart disease, stroke, and diabetes mellitus a systematic
review and meta-analysis. Circulation 121(21): 2271–2283.
Michael, Lothar. 2001a. Die drei Argumentationsstrukturen des Grundsatzes der
Verhältnismäßigkeit. Zur Dogmatik des Über- und Untermaßverbotes und der Gleichheitssätze.
JuS 41(2): 148–155.
Michael, Lothar. 2001b. Grundfälle zur Verhältnismäßigkeit. JuS 41(7): 654–659.
Michael, Lothar. 2008. Folgerichtigkeit als Wettbewerbsgleichheit. Juristenzeitung 63: 875–882.
Michael, Lothar, and Martin Morlok. 2014. Grundrechte. Baden-Baden: Nomos.
Miller, Franklin G., and Luke Gelinas. 2013. Nudging, autonomy, and valid consent: Context mat-
ters. The American Journal of Bioethics 13(6): 12–13.
Möller, Kai. 2005. Paternalismus und Persönlichkeitsrecht. Berlin: Duncker & Humblot.
Möller, Kai. 2014. Proportionality and rights inflation. In Proportionality and the rule of law.
Rights, justification, reasoning, ed. Grant Huscroft, Bradley W. Miller, and Grégoire
C.N. Webber, 155–172. Cambridge, MA: Cambridge University Press.
Nessmith, William E., Stephen P. Utkus, and Jean A. Young. 2007. Measuring the effectiveness of
automatic enrollment. Vanguard Center for Retirement Research 31: 1–19.
Neumann, Robert. 2013. Libertärer Paternalismus. Theorie und Empirie staatlicher
Entscheidungsarchitektur. Tübingen: Mohr Siebeck.
Payandeh, Mehrdad. 2011. Das Gebot der Folgerichtigkeit: Rationalitätsgewinn oder Irrweg der
Grundrechtsdogmatik? Archiv des öffentlichen Rechts 136(4): 578–615.
118 M. Schweizer

Peters, Hans. 1963. Das Recht auf freie Entfaltung der Persönlichkeit in der höchstrichterlichen
Rechtsprechung. Wiesbaden: VS Verlag für Sozialwissenschaften.
Petersen, Niels. 2013. Gesetzgeberische Inkonsistenz als Beweiszeichen. Archiv des öffentlichen
Rechts 138: 108–134.
Pieroth, Bodo. 2014. Grundrechte. Heidelberg/München: Müller.
Pohl, Rüdiger (ed.). 2004. Cognitive illusions. A handbook on fallacies and biases in thinking,
judgement and memory. Hove/New York: Psychology Press.
Porat, Iddo. 2014. Mapping the American debate over balancing. In Proportionality and the rule of
law. Rights, justification, reasoning, ed. Grant Huscroft, Bradley W. Miller, and Grégoire
C.N. Webber, 397–416. Cambridge, MA: Cambridge University Press.
Rachlinski, Jeffrey J. 2003. The uncertain psychological case for paternalism. Northwestern
University Law Review 97: 1165–1225.
Rayner, Geof, and Tim Lang. 2011. Is nudge an effective public health strategy to tackle obesity?
No. BMJ (British Medical Journal) 342(5): d2177.
Read, Daniel, and Barbara van Leeuwen. 1998. Predicting hunger: The effects of appetite and
delay on choice. Organizational Behavior and Human Decision Processes 76(2): 189–205.
Rozin, Paul, Sydney Scott, Megan Dingley, Joanna K. Urbanek, Hong Jiang, and Mark Kaltenbach.
2011. Nudge to nobesity I: Minor changes in accessibility decrease food intake. Judgment and
Decision Making 6(4): 323–332.
Russell, Jacob Hale. 2015. The separation of intelligence and control. The retirement savings crisis
and the limits of soft paternalism. William & Mary Business Law Review 6: 35–87.
Salazar, Alberto. 2012. Libertarian paternalism and the danger of nudging consumers. King’s Law
Journal 23(1): 51–67.
Saurer, Johannes. 2012. Die Globalisierung des Verhältnismässigkeitsgrundsatzes. Staat 51: 3–33.
Schlink, Bernhard. 2001. Der Grundsatz der Verhältnismäßigkeit. In Festschrift 50 Jahre
Bundesverfassungsgericht. Zweiter Band, ed. Peter Badura and Horst Dreier, 445–465.
Tübingen: Mohr Siebeck.
Schnellenbach, Jan. 2011. Wohlwollendes Anschubsen: Was ist mit liberalem Paternalismus zu
erreichen und was sind seine Nebenwirkungen? Perspektiven der Wirtschaftspolitik 12(4):
Selinger, Evan, and Kyle Whyte. 2011. Is there a right way to nudge? The practice and ethics of
choice architecture. Sociology Compass 5(10): 923–935.
Sodan, Helge, and Ziekow. 2014. Grundkurs Öffentliches Recht. Staats- und Verwaltungsrecht.
München: Beck.
Stein, Ekkehart, and Götz Frank. 2010. Staatsrecht. Tübingen: Mohr Siebeck.
Sugden, Robert. 2008. Why incoherent preferences do not justify paternalism. Constitutional
Political Economy 19(3): 226–248.
Sunstein, Cass R. 2014. Why nudge?: The politics of libertarian paternalism. New Haven: Yale
University Press.
Sunstein, Cass R. 2015. Nudging and choice architecture: Ethical considerations, Yale Journal on
Regulation Forthcoming.
Thaler, Richard H., and Cass R. Sunstein. 2003. Libertarian paternalism. American Economic
Review 93(2): 175–179.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge. Improving decisions about health, wealth,
and happiness. New Haven: Yale University Press.
Tversky, Amos, and Daniel Kahneman. 1981. The framing of decisions and the psychology of
choice. Science 211(4481): 453–458.
Ueltzhöffer, Christian. 2005. Die staatliche Einflussnahme auf den Tabakkonsum von Kindern und
Jugendlichen in Deutschland. Wien: Springer.
van Aaken, Anne. 2006. Begrenzte Rationalität und Paternalismusgefahr. In Paternalismus und
Recht. In memoriam Angela Augustin (1968–2004), ed. Michael Anderheiden, Peter Bürkli,
Hans Michael Heinig, Stephan Kirste, and Kurt Seelmann, 109–144. Tübingen: Mohr Siebeck.
7  Nudging and the Principle of Proportionality 119

van Aaken, Anne. 2015a. Constitutional limits to nudging. A proportionality assessment. University
of St. Gallen Law & Economics Working Paper (2015–03): 1–27.
van Aaken, Anne. 2015b. Judge the nudge. In search of the legal limits of paternalistic nudging in
the EU. University of St. Gallen Law & Economics Working Paper (2015–01): 1–35.
van Boom, Willem H., and Anthony Ogus. 2010. Introducing, defining and balancing autonomy v.
paternalism. Erasmus Law Review 3: 1–5.
Wahl, Rainer. 2013. Der Grundsatz der Verhältnismäßigkeit. Ausgangslage und
Gegenwartsproblematik. In Verfassungsstaatlichkeit im Wandel. Festschrift für Thomas
Würtenberger zum 70. Geburtstag, ed. Dirk Heckmann, Ralf P. Schenke, and Gernot Sydow,
823–854. Berlin: Duncker et Humblot.
Wansink, Brian, and Koert van Ittersum. 2006. The visual illusions of food: Why plates, bowls, and
spoons can bias consumption volume. The FASEB Journal 20(4): A618.
White, M.D. 2010. Behavioral law and economics: The assault on consent, will, and dignity. In
Essays on philosophy, politics & economics. Integration & common research projects, ed.
Christi Favor, Gerald F. Gaus, and Julian Lamont, 201–224. Stanford: Stanford Economics and
Wienbracke, Mike. 2013. Der Verhältnismäßigkeitsgrundsatz. Zeitschrift für das Juristische
Studium 2: 148–155.
Wilkinson, T.M. 2013. Nudging and manipulation. Political Studies 61(2): 341–355.
Willis, Lauren E. 2013. When nudges fail. Slippery defaults. University of Chicago Law Review
80: 1155–1229.
Wilson, Nick, George Thomson, Philippa Howden-Chapman, and Louise Signal. 2006. Regulations
should ban the sale of cigarette pack covers of health warnings. The New Zealand Medical
Journal 119(1243): U2251.
Yeung, Karen. 2012. Nudge as fudge. The Modern Law Review 75(1): 122–148.
Zippelius, Reinhold, and Thomas Würtenberger. 2008. Deutsches Staatsrecht. Ein Studienbuch.
München: C.H. Beck.
Chapter 8
Nudging in Swiss Contract Law?
An Analysis of Non-mandatory Default Rules from a
Legal, Economic and Behavioural Perspective

Klaus Mathis and Philipp Anton Burri

Abstract This essay discusses various interpretations of the function of non-

mandatory contract law and the ensuing requirements with regard to formulating its
content adequately. After some preliminary remarks on non-mandatory contract law
(Sect. 8.1) initially two theories are introduced: Friedrich Carl von Savigny’s theory
based on the autonomy of contracting parties and the theory of Oskar Bülow and
Rudolf Stammler focusing on heteronomous evaluations (Sect. 8.2.1). There fol-
lows a presentation of a classical economic perspective on contract law, which con-
ceives of non-mandatory norms as means for reducing transaction costs and which
upholds efficiency as the central standard (Sect. 8.2.2). Next, attention is turned on
the behavioural economic analysis of non-mandatory contract law, and the funda-
mental possibility of using non-mandatory law as a nudging instrument to steer
behaviour is introduced, making reference to two concrete examples of its applica-
tion in labour law and in surety contract law. In conclusion, the most important
findings are summarised (Sect. 8.3).

8.1 Introduction

Present-day private law systems are based on the principle of private autonomy.
According to this principle private legal subjects are fundamentally free to arrange
their legal relationships themselves, according to their own will, within the bounds
of the legal system.1 One of the means available to them for precisely this kind of
private autonomous arrangement is the instrument of the contract. Under the
principle of freedom of contract, it is fundamentally left up to them to determine

Gauch et al. 2008a, p. 59.
K. Mathis (*) • P.A. Burri
Faculty of Law, University of Lucerne, Frohburgstrasse 3,
6002 Lucerne, Switzerland

© Springer International Publishing Switzerland 2016 121

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_8
122 K. Mathis and P.A. Burri

whether and with whom they wish to conclude, terminate or amend a contract.2 In
particular, it is also left to the parties to define the content of the contract within the
bounds of the law: provided that they do not thereby violate any compelling regula-
tion, agree any impossible or illegal content, or contravene public morals,3 they are
free to regulate performance, consideration, and all other conditions of the contract
as they wish.4
For a private agreement to be recognised as a contract, the legal order demands
that the parties must have agreed a “meaningful whole”.5 This means that at least
the essential points, the essentialia negotii, of the (natural6 or normative7) consensus
must be covered. Any lack of agreement about (non-essential) ancillary points, the
accidentalia negotii, on the other hand, does not prevent the contract from coming
into being (for Switzerland cf. Art. 2 para. 1 of the Swiss Code of Obligations).8
Although such ancillary points can likewise be regulated freely – within the
aforementioned bounds – there is, theoretically, an infinite number of ancillary
points and it is unreasonable to expect the parties, and indeed not even possible for
practical reasons, to anticipate and consensually regulate all eventualities that might
come into play during the lifetime of their contract.9 This means that all contracts –
even after an interpretation of the contractual provisions – contain loopholes to a
greater or lesser extent.10
This initial situation makes it clear that a system of contract law based on private
autonomy is dependent on a mechanism for closing such loopholes, if any disputes
arising over non-contractually regulated points are not to go unanswered by the
court.11 The closing of these loopholes is referred to as completion of the contract.12
As possible instruments for such contract completion, the following cascade of
completion mechanisms has become established: non-mandatory statutory law,
customary law and contract completion by the court.13 In the case of the last method,
the court has to be guided by the hypothetical will of the parties, and apply those
rules which the parties themselves would have agreed if the question concerned had

Schwenzer 2012, p. 171.
Decision of the Swiss Federal Court BGE 96 II (18), 20, E.1.
Cf. on the entire issue, Schwenzer 2012, 180–181, 249–257; Gauch et al. 2008a, pp. 122–152.
Schönenberger and Jäggi, marg. no. 84 to Art. 1 of the Swiss Code of Obligations.
Gauch et al. 2008a, p. 312.
Gauch et al. 2008a, p. 317.
Gauch et al. 2008a, pp. 330 et seqq..
Belser et al. 2000, p. 174.
Korobkin 1998, p. 609; Schäfer and Ott 2012, p. 455.
On the theoretically conceivable options for reacting to incomplete contracts in the event of dis-
putes, see Cziupka 2010, p. 1 and for a brief comparative presentation on the treatment of loop-
hole-closing instruments in civil law and common law systems, see ibid. pp. 59–63.
Gauch et al. 2008a, pp. 283 et seqq.; Schwenzer 2012, p. 269.
Gauch et al. 2008a, p. 287.
8 Nudging in Swiss Contract Law? 123

led to a contractual provision.14 If the balance of interests in the case to be ruled

upon fits a generalizable constellation, the court must act modo legislatoris to estab-
lish a general norm and apply this to the concrete case.15 In exceptional cases, non-
mandatory statutory law makes reference to the discretion of the court or to custom
and practice.16
The present study focuses on the analysis of non-mandatory statutory law and,
for that reason, refrains from presenting the other methods of contract completion.

8.2 Theories on Non-mandatory Contract Law

Non-mandatory contract law17 comprises those norms of private law which only
come into application if the parties to a contract have not agreed their own provision
overriding the statutory regulation.18 It satisfies the principle of private autonomy
that a large number of the norms in private law exhibit this permissive character,
whereby they cede precedence to the regime arrived at consensually between the
contracting parties.19
Thus, a major part of private law and especially contract law seems to consist of
a kind of substitute regime20 which is only conceived, from the start, for the eventu-
ality that the parties fail to regulate a certain legal question themselves, and is only
intended to take effect if their contract turns out to contain a loophole in this
respect.21 Seen in this light, the state is providing a mere service to private individu-
als by holding a safety net of norms in readiness, its sole purpose being – in cases
where the common will of the parties to the contract has not been stated com-
pletely – to complete this and to resolve any disputed issues appropriately.
This is not, however, the only way of looking at non-mandatory law, which can
also be understood as a kind of state proposed rule22 from which the parties can,
and in some circumstances must, take guidance.23 Considered from this angle,
non-mandatory norms sketch out the legislatorial idea of an adequate regulation of
the given aspect of life, and thereby give expression to normatively charged

Gauch et al. 2008a, p. 285.
Gauch et al. 2008a, p. 286.
Gauch et al. 2008a, p. 284.
Also: dispositive law, ius dispositivum, voluntary law (under civil law).
For Swiss contract law: Huguenin 2012, p. 6; Gauch et al. 2008a, p. 283–285; Schwenzer 2012,
p. 252.
Cf. Bucher 1977, p. 249.
Thus, Schäfer and Ott 2012, p. 455.
Schmid and Stöckli 2010, p. 11.
Cf. Cziupka 2010, p. 207.
On the “guiding principle” function of non-mandatory law, with further citations, Cziupka 2010,
pp. 126 et seqq.
124 K. Mathis and P.A. Burri

behavioural expectations.24 From this standpoint, a divergent private agreement

appears to be an exception from the “normal order”25 regarding the matter con-
cerned and, as such, may impose an obligation upon the parties to justify their
divergent provision.26
What has been said until now indicates that the phenomenon of non-mandatory
law in our private law system is of central importance, on the one hand, and that on
the other hand, the function of this type of law is neither definitively explained and
nor does it appear to be trivial. This essay takes up these problems and inquires into
the function of non-mandatory law and, in conjunction, into the adequate content of
non-mandatory norms, which on the one hand emerge as a necessary supplement
for the incomplete contracts, and on the other hand may serve the parties as a legis-
latorial example of contractual regulation of their legal relationships.
To this end, initially the two major juridical theories27 on non-mandatory law are
presented: on the one hand, the theory of Friedrich Carl von Savigny, focusing on
autonomy and free will; on the other hand, an approach that goes back to Oskar
Bülow and Rudolf Stammler, which brings to the forefront the objective regulatory
function as well as imperatives of justice that are codified not only in mandatory
norms, but are also reflected in non-mandatory norms.28
Further, two different theories arising from an economic analysis of non-
mandatory contract law will be presented. The first to be introduced is a classical
economics perspective, which understands non-mandatory contract law essen-
tially as a vehicle for facilitating legal transactions and as a mere instrument for
lowering transaction costs. Ultimately, non-mandatory law can also be understood
as a potential nudging instrument.29 After introducing some fundamental insights
from behavioural economics, the specific mode of operation of non-mandatory
contract law will be described and the resulting nudging opportunities presented
with reference to specific applied examples in the domains of labour law and
surety contract law.

Cziupka 2010, p. 5.
Raiser 1935, p. 293.
On this, cf. citing Enneccerus et al. 1959, § 49; Cziupka 2010, p. 91.
There are numerous variations and modes of both theories. For the purposes of the present essay,
it suffices to give a schematic presentation illustrating the fundamental field of tension between
autonomous and heteronomous norming in non-mandatory contract law.
For a comprehensive presentation of these theories, see Cziupka 2010, pp. 66 et seqq.
Cziupka 2010, pp. 486 et seqq.
8 Nudging in Swiss Contract Law? 125

8.2.1 Juridical Theories on Non-mandatory Contract Law

Savigny and the Hypothetical Will of the Parties

Friedrich Carl von Savigny’s entire legal thinking is characterised throughout by

autonomy and the freedom of the private legal subjects.30 Consequently he con-
ceives of non-mandatory law as a mere interpretation of the common will of the
parties, which has only been expressed incompletely in the contract.31 He thereby
attests non-mandatory law to have a pure convenience function.32 Its task consists
solely in completing the incompletely articulated definitions of the will of the con-
tracting parties.33 To that extent, the role assigned to non-mandatory law is merely
Logically, then, the will of the contracting parties has to be taken as the only
reasonable standard for the content of non-mandatory law.35 Since this will, in the
case in question, has definitely not been expressed by the parties explicitly, nor can
it be deduced implicitly from their behaviour, reference must be made to a
presumed,36 a hypothetical37 will of the parties in order to close any eventual loop-
hole in the contract. According to Bernhard Windscheid’s wording, the content of
this hypothetical will is nothing but what the parties themselves would have agreed
if they had reached an agreement on the question concerned, and thus, the actual
will of the parties.38 This hypothetical will of the parties is reflected in the typical39
will of the parties that can be recognised in legal dealings in the contractual arrange-
ments covering uniformly recurring business situations.40 Hence, the origin of non-
mandatory law is referred to as sedimentary in nature, in contrast to the eruptive
creation of law by the legislator.41
It is evident without further discussion that in such a conception, the parties are
free without any justification to reach an agreement diverging from non-mandatory
law, considering that according to this view, the very fact of an autonomous agree-
ment is deemed to have an intrinsic “value in itself”42 which legitimises a deviation

On this, see Cziupka 2010, p. 70.
Von Savigny 1840, p. 58.
Cziupka 2010, p. 67.
See von Savigny 1840, p. 59.
Cziupka 2010, p. 72.
Cziupka 2010, p. 72.
Cziupka 2010, p. 72.
Cziupka 2010, p. 72.
Windscheid 1887, pp. 263–264.
von Savigny 1840, p. 58.
Cziupka 2010, p. 73.
Laband 1888, p. 164.
Cziupka 2010, p. 81.
126 K. Mathis and P.A. Burri

from non-mandatory law.43 A contractual agreement that also covers ancillary points
appears desirable indeed, given that non-mandatory law defines itself as a mere
instrument for facilitating legal dealings and seeks to claim only subsidiary validity
from the outset.44

Bülow, Stammler and the Binding Force of the Legal Order

The following theoretical approach, which essentially goes back to Oskar Bülow
and Rudolf Stammler, must be characterised almost as an antithesis to Savigny’s
theory. Whereas Savigny made the autonomy of private persons the starting point
of all considerations, in Bülow and Stammler the “binding force of the legal order”
is absolutely primary to the conception of non-mandatory law and contract law.45
The institute of the contract is not understood here as a consequence arising from
any pre-legal autonomy of individuals, but as an instrument made available by state
law.46 Non-mandatory law in this context reflects the order laid down by the legisla-
tor as an adequate provision, whilst also empowering private individuals if need be
to choose a provision better adapted to the given circumstances of life as a proxy for
the legal order.47 In this respect, non-mandatory law does not complete the incom-
plete contract; on the contrary, the contract is understood as a medium for the con-
cretisation of objective law.48 Thus, it becomes impressively clear that non-mandatory
law, too, has a mandating dimension.49
From this perspective, the orientation point for the content of non-mandatory
norms is obviously no longer the hypothetical will of the parties; far rather, it is
“constantly the material and ethical needs of the totality” that the legal order must
serve.50 The pure convenience function of non-mandatory law in Savigny’s concep-
tion is marginalised here, and substituted over broad areas by a value-setting func-
tion.51 Bülow and Stammler address precisely this area of tension when they write
that non-mandatory law is “not exclusively subservient to the declaration of private
will”52 but a “power independently encountering” the will of the parties.53

Cziupka 2010, p. 67.
Cziupka 2010, p. 78.
With further citations, Cziupka 2010, p. 91.
See also Huber 1933, p. 40: “There is no natural freedom to conclude a contract, but rather, there
is a state-regulated institution of the free contract”.
See extensive discussion of this in Cziupka 2010, pp. 108–109, pp. 108-109; Stammler 1886,
p. 24.
In this vein, Cziupka 2010, pp. 108–109.
With further citations, Cziupka 2010, p. 205.
Stammler 1886, p. 25.
Cziupka 2010, p. 67.
Stammler 1886, p. 24.
Bülow 1881, p. 108.
8 Nudging in Swiss Contract Law? 127

Under this regime, any deviation by the contracting parties from non-mandatory
model provisions is no longer self-evident by any means. Instead, such an approach
calls for objective reasons to be brought forth, i.e. in the form of an “unusual con-
stellation of facts and interests not covered by the law”.54 Thus, instead of leaving
the parties unbounded latitude to make their arrangements, a “varied spectrum of
possible forms of arrangement”55 is made available to them, from which they may
select one of the possible state-recognised provisions.56
This is the underlying concept of the guiding principle theory (Leitbildtheorie)
that is recognised in Germany today, which draws on non-mandatory law as a stan-
dard in the context of controlling the contents of general terms and conditions of

8.2.2 Economic Theories on Non-mandatory Contract Law

Having set out the two foundational juridical theories on non-mandatory contract
law, two economic views on non-mandatory contract law will be introduced in the
following section.58 This involves a change of perspective. Whereas the two juridi-
cal theories are developed directly from a priori values – autonomy for Savigny, and
the “material and ethical needs” of the legal community in the case of Bülow and
Stammler – economic analysis of law begins by examining human behaviour under
various conceivable institutional framework conditions. The focus, then, in this first
stage, is on how people’s expected behaviour changes (positive analysis) when cer-
tain (non-mandatory) legal norms are specified.59 Not until the second stage are the
expected consequences of different legal arrangement options compared with each
other and assessed on the basis of normative criteria (normative analysis).60
Essentially, any criteria may be chosen for the assessment of these consequences.
Economic analysis itself, however, makes use of the instruments of welfare eco-
nomics offer its own criteria for the assessment of social conditions. Accordingly,
the key parameter used for this assessment is the welfare of society as a whole,

Enneccerus et al. 1959, §49 III.
Cziupka 2010, p. 111.
Cziupka 2010, p. 104.
Cziupka 2010, pp. 126 et seqq.
At this point, a further theory discussed in the economic analysis of law discourse can be men-
tioned: this theory describes the possibility of utilising non-mandatory contract norms as penalty
default rules. The pivotal aspect of this is that non-mandatory norms that are disadvantageous for
both parties compel them to find their own contractual provision. In the negotiation process the
parties would then disclose information that they would otherwise have held back, had non-man-
datory law already proposed a provision advantageous to them. Foundational, Ayres and Gertner
1989; critical in relation to the practical relevance of penalty default rules, Posner 2006; also criti-
cal of the whole matter, Kähler 2012, p. 149 et seqq.
Mathis 2009, pp. 7 et seqq.
Mathis 2009, pp. 31 et seqq.
128 K. Mathis and P.A. Burri

which is composed of the individual welfare of each of that society’s members. A

condition associated with higher welfare is preferred over one with lower welfare.
If welfare rises as a result of a measure, the measure concerned is deemed to be
efficient.61 In this regard, two constellations can be differentiated: if a measure or
transaction results in greater welfare for at least one person without making another
person worse off, the measure concerned is Pareto efficient.62 If, in contrast, a mea-
sure that makes at least one person worse off improves the lot of at least one other
person, and if the better-off person’s gain weighs more heavily than the worse-off
person’s loss, this transaction is referred to as Kaldor-Hicks efficient.63
Economic analysis thus directs its attention to how individuals make choices and
to what extent these choices can be steered in a particular direction – normally in the
direction of efficiency – by means of legal frameworks. Individuals do not make
their decisions in a vacuum, but always act in a concrete decision-making context
that is constituted through social and legal norms and rules. The legislator can help
to shape this decision-making context by passing laws. This setting has come to be
discussed as a particular ‘choice architecture’64 within which individuals make their
choices. There are two ways in which influence can be exerted on individual choices
in this process. On the one hand, legal norming can be used to alter the incentive
structure by modifying the cost-benefit relationship of alternative choices. Measures
of this type appeal to people’s rational side; thus they are geared towards the behav-
iour of the classical homo oeconomicus. Levying taxes on fuel is one example of the
use of this method to increase the costs of consuming energy from fossil sources.
Ceteris paribus this leads to a reduction in consumption of the taxed energy sources.
On the other hand, people also react to a multitude of other factors, such as how a
particular decision-making situation is presented to them; they take their guidance
from ‘normal’ behaviour or tend to comply with specified norms per se. Such mea-
sures tend to appeal more to people’s irrational side.

Reducing Transaction Costs by Imitating the Market

The Coase Theorem as the Theoretical Departure Point

To determine the consequences of different norms, economic analysis of law tradi-

tionally reaches for the behavioural model of neoclassical economics theory, namely
the heuristic figure of homo oeconomicus.65 This model is constructed on three fun-

Efficiency or efficient can mean – in a relative sense – that state of affairs A is preferable to state
of affairs B or – in an absolute sense – that a certain state of affairs is the best possible and thus
represents an optimal state. Mathis 2009, p. 32.
For an extensive presentation and critique, see Mathis 2009, pp. 35 et seqq.
For an extensive presentation and critique see Mathis 2009, pp. 42 et seqq.
Thaler and Sunstein 2008, p. 81 et seqq.
Mathis 2009, p. 14; comprehensively on homo oeconomicus, Kirchgässner 2008; on its applica-
tion in legal science in particular, Lüdemann 2007.
8 Nudging in Swiss Contract Law? 129

damental assumptions: human beings (1) think and act rationally, (2) align their
choices exclusively with their own self-interest and (3) are equipped with unlimited
willpower to put their choices into practice.66 Starting from this model of human
behaviour, predictions can be derived about how people will adapt their behaviour
on the basis of changes – especially legal changes – to the framework conditions.
This model of human behaviour can now be made fruitful for the analysis of non-
mandatory contract law. In order to demonstrate this, first a brief introduction is
necessary of an important foundational theorem of economic analysis of law: the
Coase theorem. In his famous essay “The Problem of Social Cost”67 written in 1960,
Ronald H. Coase describes how, in a world in which there are no transaction costs
of any kind and in which all property rights are clearly allocated, economic resources
are constantly directed to where they be will put to most economically efficient
use.68 If a person possesses a right that is valued more highly by another person, this
second person will offer the other a high enough price to secure the first person’s
agreement to transfer the right. Both persons are better off after the transaction than
before; thus, the transaction is Pareto efficient. If there is nobody who is willing to
buy the right from the first person at the price asked, the explanation is that the first
person is the one who values the right most highly.

Reducing Transaction Costs as the Principal Task of Non-mandatory

Contract Law

Reality – and that is precisely the point that Coase intended to make69 – obviously
does not adhere to the idealising assumption of zero transaction costs. In actual fact,
costs are incurred for every transaction to a greater or lesser extent: searching for a
business partner, obtaining information about the product to be bought or the rights
in it, negotiating the contract, checking compliance with the contract, and enforcing
the contract.70 If the costs of a transaction in total are so high that the gain in utility
arising from it equals or even exceeds them, then rational and utility-maximising
individuals will refrain from the transaction. Any potential gain in utility is thus
inhibited by unduly high transaction costs, ruling out a theoretically possible Pareto
Against this backdrop it is now the task of the law to provide institutional frame-
work conditions which reduce the actual transaction costs as far as possible so as to
enable Pareto-efficient transactions or, in the most extreme case, even to render a
transaction superfluous by itself allocating the rights directly to the most efficient
place of use.71 From this viewpoint, the economic role of non-mandatory contract

Mathis 2009, pp. 7 et seqq.
Coase 1960.
For thorough treatment of the Coase theorem, see Mathis 2009, pp. 53 et seqq.
Coase 1960, p. 6.
For an enumerated list see Weigel 2003, p. 46 citing Alchian 1979.
Cf. on this, with further citations, Mathis 2009, pp. 61 et seqq.
130 K. Mathis and P.A. Burri

law now becomes apparent: it consists in reducing the costs of negotiating contracts
and efficiently allocating rights and obligations.72 If contracting parties need not
make contractual provision for every possible contingency that might occur every
time they conclude a contract, but can trust that non-mandatory statutory law will
contain an adequate provision, this saves them a great part of the costs of contract
In terms of content, this means that non-mandatory law should suggest those
rules which the contracting parties might be expected to prefer. Since rationally act-
ing homines oeconomici wish to maximise their utility, in any given case they will
favour the rule with the most likely prospect of efficient outcomes.73 So where the
transaction costs of negotiating an efficient agreement are too high, the utility-
maximising market solution should be imitated by non-mandatory law.74
In effect, this theory exhibits an unmistakable similarity with Savigny’s will-
based theory. Indeed, effectively these two positions are largely convergent.75
Nevertheless, it makes a difference whether a theory is based on the principle of
individual autonomy or of efficiency, since the latter places the focus on collective
welfare. Lorenz Kähler makes this explicit for contract law by means of the follow-
ing example76: It can be argued that a clause (provided by non-mandatory law) in a
transport contract whereby bus passengers undertake to donate blood to each other
in the event of an accident increases their common utility. In spite of this, a theory
based on autonomy would probably register reservations about such a rule. No the-
ory respecting the autonomy of individuals would condone the (albeit non-
mandatory) withdrawal of certain highly personal rights, even if on average a gain
in utility could be expected from such a provision.
The two theories also differ in the ways in which they determine the appropriate
content of non-mandatory contract law. In Savigny’s theory, fundamental impor-
tance is ascribed to the typical provisions chosen by the individuals in similar situ-
ations in contractual reality. Economic theory suggests that in reality, however,
prohibitively high transaction costs prevent the parties themselves from finding a
solution that is efficient. It therefore derives the content of non-mandatory norms
from theoretical considerations of efficiency. It proposes the norm that the parties
would have agreed themselves in a world without transaction costs, and not the
norm that is typically chosen in reality, where transaction costs are a given.77

Schäfer and Ott 2012, pp. 455–456; Kähler 2012, p. 140.
Cziupka 2010, pp. 296 et seqq.
With further citations, Unberath and Cziupka 2009, p. 49; and Cziupka 2010, p. 296.
Unberath and Cziupka 2009, p. 41.
Kähler 2012, pp. 148–149.
Unberath and Cziupka 2009, p. 74.
8 Nudging in Swiss Contract Law? 131

Question of Departure from Non-mandatory Law in the Classical Conception

of Economic Analysis of Law

In Savigny’s theory, a departure of the parties from non-mandatory law is possible

without any problem; as an expression of private autonomy in legal arrangements,
in fact, it is actually welcomed. Under the conception of Stammler and Bülow, in
contrast, such an approach appears suspect; in certain circumstances, objective rea-
sons may be demanded for having defied the order put in place by the legislator, if
the protection of the legal system is to be invoked.
From an economic analysis of law perspective, the question is framed differ-
ently: in which circumstances if any will the parties want to make their own agree-
ment, departing from the statutory rules suggested? The answer to this is supplied
by the economic behavioural model. Rational and utility-maximising contracting
parties will always reach their own agreement if the expected utility from the agree-
ment in question is greater for the parties than the transaction costs incurred for the
additional necessary contract negotiations.78 The observed fact that the parties in a
concrete case have not reached any agreement of their own may occur for a great
many reasons: the statutory suggestion already fits very well into the structure of
their legal relationship; they do not expect any notable benefits from a waiver
because they assume that there is only a very slender probability that the contin-
gency in question will occur; they are confronted with especially high information
costs because they are inexperienced in business; or the low transaction volume
may not justify extensive contractual protection.79

Concrete or Abstract Norms in Non-mandatory Contract Law?

As a precondition for substantial costs to be saved and transactions made easier

through the use of non-mandatory law, it must provide rules that are suitable for the
concrete business transaction of the parties. Otherwise the parties in turn would
have to carry on negotiations in order to replace the unsuitable stipulations of the
law with a provision better adapted to their circumstances.
At first glance, therefore, an ideal condition would be attained if, in every indi-
vidual case, non-mandatory law suggested the provision that the concrete parties
themselves would have agreed if they had invested the necessary time and energy
in the negotiation. In view of the great number and heterogeneity of the contracts
concluded in reality, this is naturally an illusion. The jungle of regulations that
would grow up from regulating the detail of every contract in any kind of regular
use would be preposterous from the viewpoint of practicability.80 For one thing,
such an approach would be doomed to failure by the fact that the legislator does not

Korobkin 1998, pp. 613–617; 2000, with reference to the influence of standard form contracts
pp. 123–125.
Kähler 2012, pp. 141–142.
On the practicability of the law in general, see Mathis and Anderhub 2009.
132 K. Mathis and P.A. Burri

possess the necessary information to pass the corresponding statutory norms; for
another, the application of law in a dispute would likewise be very complicated,
since it would first be necessary to determine the category into which the contract
falls – and hence the applicable norm – which, given a theoretically infinite number
of types of contract, would be a practically impossible exercise.81 Furthermore a
tightly-worded statutory proposed rule may even be a hindrance to the parties in
formulating their contracts; simple and rapidly comprehensible norms can more
easily be waived than complicated ones, however balanced in terms of content.82
Another option for arriving at individually adapted provisions would be for the
legislator to define only a few principles, keeping them abstract, which would then
have to be concretised by the courts in individual cases.83 This would save both the
parties and the legislator from having to pass detailed norms with regard to concrete
individual cases. However, this approach also brings problems in its wake. Firstly,
when they conclude the contract, the parties can barely anticipate how a judge will
resolve any potentially arising dispute; in other words, this approach would leave
them with very little legal certainty.84 Secondly, handing down just rulings will be
very arduous for the courts if they have to ascertain and weight all the details of the
concrete case and first concretise the general principles before ultimately being able
to apply them to the matter at issue. Whatever this variant saves in terms of legisla-
tive work, it results in additional judicial effort, which is ultimately reflected in
higher court costs.85
From what has been said, it can be concluded that neither unduly abstract nor
unduly detailed norms appear to have any merit. The question therefore arises as to
whether a middle path86 might deliver an optimum – although not an ideal – result.
This would require the legislator to confine itself to dividing all common contracts
into a manageable number of groups, which are internally as homogenous as pos-
sible but externally distinct from other groups.87 Next, the essential elements of each
type of contract and frequently occurring matters of dispute would be regulated in
such a way as to find an efficient solution for the majority of contracts actually con-
cluded which fall into the given group. But this means that a minority exists which
now has to live with the fact that they are subject to non-mandatory norms which
are suboptimal for their specific contractual relationship. This can be justified by
means of the following arguments: First, such a solution is Kaldor-Hicks efficient if
it can be assumed that the transaction-cost savings for the majority exceed the costs

Unberath and Cziupka 2009, p. 60.
Kähler 2012, p. 142.
Unberath and Cziupka 2009, pp. 60 et seqq. The same applies to the reference to a hypothetical
will of the parties, which relates to what bona fide parties in the contracting parties’ place would
have agreed in good faith, cf. Schwenzer 2012, p. 270. This concept is likewise very abstract and,
in the application of law, requires great concretisation by the courts.
Cziupka 2010, p. 410.
Unberath and Cziupka 2009, p. 62.
On this, see Cziupka, p. 415.
Unberath and Cziupka 2009, p. 48
8 Nudging in Swiss Contract Law? 133

incurred by the minority due to inadequate provisions or waiving of the same.88

Note that here it is evident that the non-mandatory nature of contract law norms
proves to be extremely helpful from an economic perspective, since it leaves the
parties with the option of waiving a norm – to the tune of the additional transaction
costs – that is inefficient for them.89 Second, it serves the purpose of legal certainty.
Particularly after having signed the contract, the parties can obtain information at
relatively low cost about how a court will most likely rule on a possible dispute.90
Third, in cases in which the parties have not agreed any contractual solution despite
blatantly unfit non-mandatory law, the judicial completion of contracts with refer-
ence to general principles can be invoked as a useful corrective instrument.91 Though
inappropriate as a primary means of loophole closure, judicial completion of con-
tracts which have “fallen through the net of the statutory typological system” can
certainly be used to good purpose here.
In Switzerland, the middle path just outlined seems to be found in positive law.
The typical contracts, in the sense mentioned – in part still subdivided down into
subtypes – are regulated in the special part of the Swiss Code of Obligations.92 Even
on the typical innominate contract, a stable line of adjudication has been established
which can be counted as functionally equivalent to non-mandatory contract law. In
the Swiss literature, the view continues to be advocated that non-mandatory con-
tract law is the first and foremost instrument to be referred to for contract comple-
tion.93 Non-mandatory statutory law can only be called upon to close loopholes,
however, if the norm in question is applicable to the case at all. If the norm does not
fit the content of the contract, so that applying it would give rise to a “disharmony”,
then clearly the norm is not tailored to the particular contract and is not therefore a
workable instrument for contract completion.94 In the case of idiosyncratic con-
tracts, the judge’s task is to find a solution that he deems appropriate to the problem
laid out before him. His guideline in this is the hypothetical will of the parties.95 If,
in the concrete case, the judge spots a generalizable set of interests, he is required to
act modo legislatoris by establishing a general rule and applying this to the case in

Unberath and Cziupka 2009, p. 52. Here the additional assumption is made that the legislation
costs and court costs for this variant are lower than for the two extreme variants. This is a plausible
scenario under the assumption of rising marginal costs.
Cziupka 2010, p. 420.
Cziupka 2010, p. 417.
Schäfer and Ott 2012, pp. 457–458; Cziupka 2010, pp. 417–418.
Schmid and Stöckli 2010, pp. 5 et seqq.
Gauch et al. 2008a, p. 284, citing by way of a rationale the judge’s obligation to the statute (Art. 1
para. 1, Swiss Civil Code); for a contrasting opinion, see Schwenzer (2012, pp. 270–271), who
cites the principle of freedom of contract as a basis for prioritising the hypothetical will of the
Gauch et al. 2008a, pp. 284–285.
Gauch et al. 2008a, pp. 285–286.
Gauch et al. 2008a, pp. 285 et seqq.; Art. 1 para. 2 of the Swiss Civil Code.
134 K. Mathis and P.A. Burri

General terms and conditions of business as a (partial) substitute for non-

mandatory contract law

General terms and conditions of business are a frequent occurrence in contract prac-
tice. They can partially fulfil the function of non-mandatory law. Normally they
contain a comprehensive specification of the rights and obligations that are relevant
in the business relationship concerned, which renders further negotiation over
details unnecessary. Thus, they are eminently suitable for lowering transaction
costs.97 However, general terms and conditions of business are frequently biased
towards the welfare of one individual contracting party, and not towards maximis-
ing the welfare gain of both parties in line with economic efficiency criteria. On the
contrary, market power, asymmetric distribution of information and transaction
costs often mean that general terms and conditions of business provide for an inef-
ficient allocation of rights and obligations.98

8.2.3 Behavioural Economic Analysis: Non-mandatory

Contract Law as Nudging?

Until now, economic analysis of law has suggested that the main reason for the
omission of contractual agreements and the ensuing reliance on non-mandatory
statutory law resides in transaction costs that are too high in relation to the utility
derived by parties from their own agreement. In the following, however, it will be
shown that apart from the issue of transaction costs there is also a multitude of other
reasons why parties often refrain from allocating rights and obligations themselves
at variance with non-mandatory law.
All of these essentially stem from the fact that real people do not match up to the
ideal model of homo oeconomicus but diverge from its strictly rational behaviour in
all manner of ways.99 Thus, empirical studies from cognitive psychology research
show that individuals generally underestimate risks (overoptimism), make distorted
assessments of probabilities on the basis of selectively available information (the
availability heuristic), give too little consideration to the long-term consequences of
their actions (hyperbolic discounting), prefer the status quo over other alternatives
(status quo bias), follow a natural inclination to comply with norms (norm compli-
ance), and allow their choices to be influenced by irrelevant factors (anchor effect).

Schäfer and Ott 2012, p. 456 citing Kötz 2003, pp. 209–214.
Schäfer and Ott 2012, p. 456 citing Kötz 2003, pp. 209–214; see also Belser 2000, pp. 184
et seqq.
Schnellenbach 2012, pp. 595–596, with further citations; see also the extensive discussion in
Mathis and Diriwächter 2011 and Mathis and Steffen 2015. On constrained rationality in general,
see also McFadden 1999.
8 Nudging in Swiss Contract Law? 135

Reasons for the De Facto Stability of Non-mandatory Law in Reality

According to the neoclassical model, the allocation of rights by negotiable law is of

relatively minor importance.100 If the non-mandatory norm does not match the inter-
ests of the parties, they simply elect to waive it (given proportionate transaction
costs).101 Empirical studies show that in practice, however, reliance is placed on
non-mandatory law far more frequently than the neoclassical model would pre-
dict.102 In the following, the above-mentioned behavioural economic reasons for this
will be elucidated.


People tend to overestimate their own abilities and power to influence particular
events, and systematically underestimate the probability that they will be affected
by a negative event (hence overoptimism).103 For example, most people’s estimation
of the probability of a car accident is considerably lower than the probability actu-
ally is for the average car driver.104 So this overoptimism ensures that people sys-
tematically overestimate the probability of positive events and underestimate the
probability that negative events will occur. Accordingly, the parties concluding a
contract tend to neglect explicit provisions concerning risks that could materialise
in the course of the term of their contractual relationship, since they assume too low
a probability of any such occurrence.105

Availability Heuristic

Another reason why parties tend to leave out provisions for particular risks is that,
out of strategic considerations, they do not even mention these during contract
negotiations. For if one side introduces a particular risk into the discussion, the
other contracting party might assume that the counterpart considers the risk to be
especially high.106 This impression is amplified by what is known as the availability
heuristic, whereby people overestimate the probability of events that they can read-
ily associate with a real-life situation.107 The more easily a person can imagine a
particular situation, the higher they estimate the probability that the event in

Korobkin 1998, pp. 621–623.
Unberath and Cziupka 2009, p. 76.
Korobkin 1998, pp. 633–646; see also Listokin 2009.
Jolls 2009, pp. 77–78.
See DeJoy 1989.
With further citations, Kähler 2012, p. 169.
Kähler 2012, p. 169.
See Jolls et al. 1998, p. 1537.
136 K. Mathis and P.A. Burri

question will occur.108 By raising the risk as a topic, a contracting party vividly
draws attention to it in the other party’s mind.109 Therefore, to pre-empt the avail-
ability heuristic, a party who is dependent on the other as a business partner will
often avoid mentioning certain risks, even if, viewed objectively, it would seem
worthwhile to agree provisions on them.110

Hyperbolic Discounting

A further effect which, at the time of concluding a contract, discourages parties

from taking trouble over unduly detailed negotiations is the phenomenon of hyper-
bolic discounting,111 whereby individuals assign far too little weight to the conse-
quences of current choices that will only occur in the distant future, resulting in
time-inconsistent preferences.112 What this means is that people tend to neglect both
positive and negative consequences that may arise from a choice made in the here
and now. Because of this short-sightedness, they later come to regret decisions
taken at an earlier point in time. This phenomenon explains why, for example, many
people have to struggle with unhealthy behaviours such as smoking113 or unhealthy
eating, and – of their own volition – save too little for their old age.114 In the short
term these activities bear fruit in the form of results perceived as benefits; smoking
a cigarette and eating sugary foods serve to gratify an immediate need; neglecting
to save adequately for old age has the direct consequence that more money is avail-
able for consumption today. In the long run, however, the said behaviours often
yield undesirable results: obesity, lung and cardiovascular disease, and poverty in
old age. A parallel can be seen in the signing of half-baked contracts: usually, the
risks against which contractual precautions ought to be taken only manifest them-
selves in the distant future, whereas the transaction costs of negotiating the contract
are incurred right away. The upshot is that many people tend to conclude contracts
hastily, without waiving inappropriate non-mandatory norms as rational individuals
would have done.

Tversky and Kahneman 1974, p. 1127.
Kähler 2012, p. 170.
Kähler 2012, p. 170.
Foundational discussion, Laibson 1997.
Thaler and Sunstein 2008, p. 44.
Gruber 2002, p. 54–55.
Thaler and Sunstein 2008, p. 73, pp. 112–115.
8 Nudging in Swiss Contract Law? 137

Status Quo Bias

For many of the choices encountered in the course of life, one of the possible options
is the non-choice. Many people have a strong inclination to “choose” precisely this
option, thereby clinging onto the existing state of affairs (hence, status quo bias).115
A vivid and prominent example of this phenomenon is the opt-out solution for
organ donorship, whereby all human beings are essentially organ donors after their
deaths – unless they make an explicit choice during their lifetimes not to be. In
comparison with the opt-in solution, whereby organ removal is only possible with
express consent, such a provision may motivate considerably more people to donate
their organs even though the transaction costs for an opt-out declaration are very
For the contracting parties, non-mandatory law – since it takes force when they
do not actively opt for an alternative provision – represents the status quo.117 This
means that its content has far greater influence on contractual reality than the neo-
classical model would predict.
The various phenomena listed as causes of the status quo bias are presented

Endowment Effect
One explanation for the status quo bias is provided by a phenomenon called the
endowment effect. This effect describes how people attach a higher value to some-
thing if they already own it.118 For legal norms, this means that those to whom non-
mandatory law allocates a subjective right are only willing to relinquish this right in
a contract negotiation in return for a relatively high payment.119 The following
extract from a study by DiMatteo will clarify this:
When a penalty clause was part of the status quo (standard form scenario), the subjects
demanded a higher price to remove the clause than they were willing to pay to insert one.
In fact, by an overwhelming margin, the most popular response to the request to remove the
standard clause was: ‘will not agree to remove Penalty Clause at any price.’ […] In con-
trast, when the clause was part of a standard form contract, 74.5 % of the subjects were
unwilling to agree to its removal ‘at any price.’ The survey confirmed Professor Korobkin’s
findings that the framing of a contractual term as the norm impacts the preferences of the
contracting parties.120

In this experiment, the subjects were asked how much they would be prepared to
pay in order to insert into the contract a contractual penalty in their favour, and how
much they would accept in return for giving up a pre-existing contractual penalty in

See Samuelson and Zeckhauser 1988.
Thaler and Sunstein 2008, pp. 187–188.
Cf. Korobkin 1998, pp. 630–633.
Cf. Kahnemanet al. 1991.
Cf. Korobkin 1998, p. 656.
DiMatteo 2006, p. 897, citing Korobkin 1998).
138 K. Mathis and P.A. Burri

their favour. The result is surprisingly decisive. Although the two constellations are
identical from an economic point of view, in the case of the pre-existing penalty
clause 74.5 % of participants were not prepared to give up the contractual penalty at
any price, whereas only 32 % were willing to pay the stated maximum price for the
insertion of the clause.121
If one party is only willing to give up a right to which it is entitled under the
status quo in return for a relatively high payment, and meanwhile the other party
attaches relatively little value to a right that it was not allocated by non-mandatory
law, it can be concluded that the allocation of rights by non-mandatory law substan-
tially influences the final outcome of the distribution of rights and obligations.122

Loss Aversion
A further phenomenon closely related to the endowment effect, which can partially
explain the stability of non-mandatory law in contractual reality, is loss aversion,
whereby individuals attach greater weight to losses than to gains.123 If a person loses
100 francs the loss of utility suffered is greater than the gain in utility experienced
from receiving 100 francs. For this reason, the parties tend to be unwilling to give
up – i.e. lose – rights in order to receive – i.e. gain – other rights in return.124

Omission Bias
The so-called omission bias states that people perceive negative events to be worse
if they are caused by a definite action than by inaction.125 Parents perceive it as
worse, for example, when their child is made ill by a vaccination than by a disease
that a vaccination might have prevented.126 This effect is even stronger if the conse-
quences of the action are unknown at the time of making the decision. Since in
many cases the parties cannot clearly anticipate the consequences of a contractual
provision at the time of concluding the contract, in accordance with the omission
bias the parties will tend to bear the risks assigned to them by statute rather than
making private agreements and taking upon themselves new risks created by their
own deliberate action.127

With further references on the endowment effect in conjunction with non-mandatory law, see
Kähler 2012, p. 180.
Unberath and Cziupka 2009, p. 72.
Kahneman and Tversky 1979.
Korobkin 1998, pp. 655–656.
With further citations, Kähler 2012, p. 182.
With further citations, Kähler 2012, p. 182.
On this, see Kähler 2012, p. 182.
8 Nudging in Swiss Contract Law? 139

Norm Compliance

A further reason that contributes to the tendency of contracting parties to cling to

non-mandatory law is the phenomenon of norm compliance. This is manifested in
the fact that the addressees of norms tend to adhere to the rules laid down, irrespec-
tive of whether their contents confer any benefits.128 This can be motivated by the
associated gain in reputation of being an honourable person, a boost in self-esteem,
and a general trust in the legal system.129

Anchor Effect

Furthermore, people tend to use information as a basis for their choices, even infor-
mation that has no objective connection with the choice situation in question (the
anchor effect).130 In a study by Kahneman and Tversky, students were asked to
estimate the proportion of African states in the United Nations Organization. Before
the estimation task, the experimental subjects spun a wheel of fortune which “ran-
domly” stopped at the number 65 or 10. Students in the group whose lucky number
was 65 estimated the proportion of African states in the UN at 45 % on average; for
the group with the lucky number 10, the mean estimated value was 25 %. Thus, the
completely arbitrary number on the wheel of fortune, by serving the experimental
subjects as an anchor, had a substantial influence on their estimation.131
What this means for contract law is that even if the parties make a divergent
provision, it is not unusual for the non-mandatory legal norm to act as their anchor.132
So private agreements take place in the shadow of non-mandatory law, which func-
tions as an important reference scale.133 Or to put it another way: “The default rule
can be very ‘sticky’”.134

Possibility of Steering Behaviour by Means of Nudging

The analysis of these variations leads to a new behavioural model which more ade-
quately describes human behaviour. Apart from explaining why the contracting
parties often omit a contractual provision, more importantly, these insights also
shed light on the possibility that, in principle, it is possible to use the cognitive
characteristics of real individuals as reference points for measures to influence their

Kähler 2012, p. 182.
With further citations, Kähler 2012, p. 182. For a thorough treatment of trust in the legal system,
see Kähler 2010, pp. 441 et seqq.
Tversky and Kahneman 1974, p. 1128.
Tversky and Kahneman 1974, p. 1128 and Mathis and Diriwächter 2011, p. 61.
On the so-called anchor effect, see Tversky and Kahneman 1974, p. 1128.
Möslein 2010, p. 2864.
Sunstein and Thaler 2003, p. 1175.
140 K. Mathis and P.A. Burri

behaviour.135 Since these measures are not normally dependent on exerting direct
compulsion, but rather act as a kind of signpost for those concerned or steer them
subliminally in the direction of a certain choice, such measures are referred to in the
literature as ‘nudges’. A measure counts as a nudge if it causes negligibly low
costs – if any – to the person affected, and hence has a substantial influence on the
behaviour of homo sapiens, but none – or very little – on that of homo
A new interpretation of the function of non-mandatory contract law now emerges:
It seems natural to take into account the errors and omissions of individuals’
decision-making when formulating non-mandatory norms in contract law, not least
because it opens up options for the legislator to use non-mandatory law to exert
substantial influence on private legal arrangements without having to restrict the
private autonomy of the parties by passing mandatory norms.
With regard to the content of non-mandatory contract law, two points have to be
addressed. The first is that the alignment of the content of non-mandatory law to the
contractual provisions typically chosen in contractual reality, as demanded by
Savigny’s theory, is not very convincing. Often, because of cognitive deficits, the
parties will negotiate suboptimal contracts. Furthermore, such contracts are always
dependent on currently applicable non-mandatory law, so that the real hypothetical
will of the parties is not completely expressed in the contracts actually concluded.137
The question which then arises is what should serve as the guideline for formulating
the content of non-mandatory norms.
In principle it is always up to the legislator which objectives it wishes to pursue
with the said measures. For instance, by means of targeted nudging it may promote
the efficiency of the contracts concluded, but may also bring in ideas of justice, for
example, by seeking to protect the typically weaker contracting party. It must be
borne in mind that the cognitive deficits identified by behavioural economics can
themselves be the reason that makes it seem appropriate for the legislator to inter-
vene correctively in the arrangement of contracts by private individuals. For if, as a
result of overoptimism or other cognitive anomalies, individuals enter into contracts
which they would not have concluded in this form as entirely rational individuals,
the question that arises is whether the legislator should protect these people from
their own weaknesses by means of (non-mandatory) norms.
Such an approach is generally discussed under the heading of “libertarian
paternalism”.138 The paternalistic part of this concept consists of influencing the
decisions of individuals for their own welfare, because they would otherwise fail to
honour their objective interests optimally due to the cognitive distortions.139 The

Jolls et al. 1998, p. 1447.
Thaler and Sunstein 2008, p. 8.
Unberath and Cziupka 2009, p. 74.
Foundational, Sunstein and Thaler 2003; critique of their work, Horn 2013, p. 14 and
Schnellenbach 2012, pp. 594 et. seqq.
Sunstein and Thaler 2003, p. 1162.
8 Nudging in Swiss Contract Law? 141

liberal aspect is that whichever measures are enacted, the individual should essen-
tially be free to choose a different option than that provided by the legislator.140
In order to protect individuals from their own irrational decisions, or to make
targeted use of their cognitive weaknesses in order to steer their behaviour in a par-
ticular direction, various strategies are available. What is known as debiasing141 is
the attempt to alleviate the cognitive anomalies by means of appropriate measures.
The goal of this is to help people towards the possibility of making rational choices.
One example is supplied by the warnings on dangerous products. Based on overop-
timism, consumers tend to underestimate the risk arising from certain dangerous
products.142 Complicated packaging inserts which are factually correct but present
the statistical risk to the consumer in a confusing way are not normally a suitable
way to counteract the overoptimism bias. It is much more effective to draw the
consumer’s attention to the danger by means of intuitively understandable symbols.
This method is far more likely to succeed in helping the consumer to recognise the
risk and make a conscious, rational choice.143
Another possibility of protecting individuals from suboptimal decisions is the
prohibition of certain courses of action.144 This does not prepare the ground for a
rational choice but, far rather, prohibits an irrational choice from the outset. An
example would be the prohibition of the consumption of hard drugs. Ultimately it is
also conceivable in principle to make targeted use of people’s cognitive deficits in
order to steer their behaviour in a certain, socially desirable direction. Such an
approach is referred to as rebiasing. It is neither an attempt to facilitate a rational
choice nor to prevent an irrational choice. The aim is rather to make use of irrational
behaviours in order to accomplish societal objectives. For example, the state could
make a rule whereby electricity from renewable energy sources was the default
option. Purchasing electricity generated by nuclear or coal-fired power plants would
still be possible, but would have to be actively chosen by individuals. Based on the
status quo bias it is likely that, under such a rule, there would be greater demand for
electricity from renewable sources.
Making reference to applied practice in two fields of law, the possible practice of
regulation informed by behavioural economics will be set out below. The first
example concerns labour law. In this field of law there are many norms to protect
the typically weaker contracting party, namely the employee. The aim is to protect
the latter from abuse of the employer’s position of superior economic power. The
relevant norms are frequently mandatory in nature. The following sets out to show
that non-mandatory norms can also play their part in improving labour conditions.
Law on contracts of surety serves as the second example. Here the problem has
a different emphasis. The central concern is not the protection of the weaker party,
but rather, the protection of the surety from himself or from his own limited

Sunstein and Thaler 2003, p. 1161.
Grundlegend Jolls and Sunstein 2006, pp. 199–241; see also Mathis and Diriwächter 2011,
pp. 63 et seqq.
Jolls and Sunstein 2006, p. 208.
Jolls and Sunstein 2006, pp. 207 et seqq.
Jolls and Sunstein 2006, p. 207.
142 K. Mathis and P.A. Burri

cognitive capabilities. Since surety contracts are often agreed – commonly among
close family and friends – as accommodation surety contracts, these are usually
uncompensated. The surety does not profit from any payment for guaranteeing the
principle debtor’s obligations. It can thus be said that here, the protection of the
stronger party from himself is the prime consideration.

Nudging in Labour Law

One possible applied field for such nudges is in labour law.145 According to Swiss
labour law, an employee is fundamentally obliged to perform any hours of overtime
required (Art. 321c para. 1, Swiss Code of Obligations). If the employee agrees,
these hours can be compensated by means of additional days of holiday. Otherwise
the additional work is to be remunerated. According to Art. 321c para. 3, Swiss
Code of Obligations, a premium of 25 % is payable on top of the normal rate of pay.
This stipulation is non-mandatory in nature, however. That is to say, it can be
adjusted both up as well as down, by means of a written agreement, a normal
employment contract or a collective employment contract (cf. Art. 321c para. 3,
Swiss Code of Obligations).146
Against the backdrop of the discussed insights from behavioural economics, this
stipulation can be understood as a nudge in favour of the employee. Initially Art.
321c para. 3 of the Swiss Code of Obligations functions as the default rule, and thus
is always applied if the parties have not agreed anything to the contrary – which will
presumably, thanks to the status quo bias, often be the case. But even if the parties
should reach a divergent agreement, the premium of 25 % specified by non-
mandatory law acts as an anchor. The parties allow themselves to be influenced by
this figure and will therefore tend to agree rates of remuneration higher than 100 %.
Now a possibility of increasing average remuneration in contractual reality comes
to light: if the legislator were to raise the premium to 50 % as standard, the expected
consequence would be that employees on average would receive higher hourly
remuneration for overtime worked.
By making use of nudging in a similar way, for example, the average length of
annual leave actually granted could be increased. According to Art. 329a para. 1 of
the Swiss Code of Obligations, the employee has a right to at least 4 weeks annual
leave per business year.147 Supposing that the legislator altered this stipulation to the
effect that the standard duration of annual leave was now 6 weeks, but never less
than 4 weeks, then on the basis of the outlined behavioural economic phenomena a
considerable number of employees would begin to benefit from longer holidays. It
would be harder for employers to grant just 4 weeks annual leave if statutory
norming succeeded in surrounding a 6-week leave allowance with the “aura of

See, for example, Sunstein 2002, who supports assigning certain rights to employees in accor-
dance with non-mandatory law.
On the whole issue, Dunand 2013, p. 93.
Cerrotini 2013, p. 370.
8 Nudging in Swiss Contract Law? 143

normality”148 and employees increasingly demanded that as “their rightful due”. On

the basis of the anchor effect, a 5-week annual leave allowance would be considered
more reasonable than under the currently valid law, so that holiday durations of
only 4 weeks would probably occur less frequently in contractual reality than today.
At this point, however, it is necessary to mention the limits of such an approach.
For if one party which possesses a certain market power has a great incentive to
diverge from non-mandatory law, this counteracts those forces thanks to which the
parties often follow the non-mandatory statutory suggestion.149 In such situations,
then, the stickiness150 of “default rules” no longer holds true to the same extent, so
that in this context Willis talks about “slippery defaults”.151 Furthermore, in practice
collective labour agreements and – especially outside of labour law – general terms
and conditions of business are widespread. If contracts are no longer negotiated in
detail, the behavioural economic effects of non-mandatory statutory law obviously
lose their impact. Instead, collective labour agreements for employees and general
terms and conditions of business for consumers now take the place of non-mandatory
law as the status quo. These, in turn, exert the impact that the findings of behav-
ioural economics have attested for non-mandatory law.152 To what extent non-man-
datory law might also have an influence on the formulation of general terms and
conditions of business and collective labour agreements remains largely unexplored
at the time of writing.

Surety Contract Law as Nudging

Behavioural Economic Motivation of the 1942 Reform of Swiss Law

on Contracts of Surety
Switzerland’s current law on contracts of surety (Art. 492–512, Swiss Code of
Obligations) can be interpreted as a regulatory intervention motivated by behav-
ioural economics. The legislator inserted these particular stipulations into the Code
of Obligations in the year 1942 as a reaction to an actual surety crisis. In the wake
of the Great Depression of the 1930s, a large number of sureties and their families
became destitute because they had entered into surety contracts, which they could
not cover economically.153 In its dispatch from 1939, the Swiss Federal Council
justified its view of the necessity for reform as follows:
It is immediately clear that these are instances of abuses of the surety contract system,
which – because they are not just rare and exceptional cases – constitute pathological
symptoms in the body of the economy, which the legislator may not heedlessly pass by. Its

Unberath and Cziupka 2009, p. 73).
Willis 2013, pp. 1200 et seqq.
Cf. Sunstein and Thaler 2003, p. 1175.
Willis 2013, p. 1174.
Luth 2010, pp. 195 et seqq.
With further citations, Kuhn 2011, p. 695 et seqq.; Swiss Federal Gazette BBL 1939 841, p. 845.
144 K. Mathis and P.A. Burri

task is here to look to justice, to draw the future surety’s attention to the risks he is proceed-
ing to enter into, to protect him from his own inexperience, credulity and kind-heartedness
but also from his own imprudence. The prevention of excessive burdens through surety
contracts is not just in the interest of the surety himself, but also of the creditor and of any
healthy economic and credit policy whatsoever. It may be true that the experiences of recent
years exert an inhibiting effect in any case, so that today people exercise more caution and
restraint over signing a surety contract. But this applies least of all to those would need it
most, namely the incorrigible optimists. Also, it cannot be disregarded that such experi-
ences are quickly forgotten when the economy improves again, and that children have to try
out their parents’ experiences themselves in order to grow wise. Nor may one succumb to
the deceptive hope that it will suffice to draw attention to the risks by means of public infor-
mation work in schools and for the general public. Even today, every Swiss child knows the
aphorism: “Bürgen tut würgen” [“signing surety leads to strangulation”]. The knowledge
that can be passed on in this way is not enough, however, to protect sureties from falling
prey to the powers of persuasion of a debtor in difficulties and, in a careless moment, plac-
ing his signature on a risky surety obligation in the belief that this time it really is just a
formality. Here, only a reform of the law on contracts of surety can help.154

It is immediately evident that, in this case, the Swiss Federal Council did not set
out from the ideal image of homo oeconomicus, but instead pays attention to those
attributes of real people which are being emphasised once again today by behav-
ioural economics. Indeed the Swiss Federal Council’s arguments could almost have
been taken from a current behavioural economics study, considering that wide-
spread, irrational overoptimism is so exactly identified as one of the main causes of
the dire straits suffered by many sureties.
The Swiss Federal Council also anticipated another behavioural economic phe-
nomenon in its arguments, namely hyperbolic discounting. With regard to the surety
contracts issue, the Swiss Federal Council writes the following in its dispatch of
Psychologically, the risk of entering into burdensome surety contract obligations is that
what is promised is not an immediate but a future payment, which possibly will not materi-
alise at all. This lures optimists into underestimating the risk.155

As becomes evident from this extract, entering into surety contract obligations
involves an interplay between overoptimism and hyperbolic discounting. Firstly, the
surety systematically underestimates the risk of actually being called upon to meet
the liability. Secondly, the surety attaches too little weight to the costs, which will
often only be incurred – if at all – in the distant future.

Interpretation of Swiss Law on Contracts of Surety as Nudging

Of greatest interest to the present study is the manner in which the Swiss Federal
Council wishes to counteract the problem of surety contracts entered into carelessly
and give better protection to sureties. Although the majority of the stipulations in

Swiss Federal Gazette BBL (1939 841, p. 846) (emphasis added by authors).
Swiss Federal Gazette BBL (1939 841, p. 869) (emphasis added by authors).
8 Nudging in Swiss Contract Law? 145

surety contract law are prima vista mandatory in nature,156 nevertheless a set of
norms exist, which are aimed at protecting the surety and are non-mandatory in
If the parties have not expressly agreed a joint and several surety contract, for
example, then the surety only becomes liable if the principal debtor has been
declared bankrupt or has made enforcement of the claim considerably more difficult
by relocating his domicile abroad (known as a simple surety contract; Art. 495 para.
1, Swiss Code of Obligations).157 The parties are free, however, to insert a joint and
several surety provisions in their contract, the upshot of which is that the surety can
be sued as soon as the principal debtor has defaulted on debt repayment and the
creditor has issued a reminder to no avail (Art. 496 para. 1, Swiss Code of
Obligations).158 That is a non-mandatory protective norm favouring the surety.
Moreover, the surety’s maximum liability decreases with the passing of time,
according to non-mandatory law. Pursuant to Art. 500 para. 1 of the Swiss Code of
Obligations, the amount of the liability amortises by 3 % p.a. or by 1 % p.a. in the
case of legally secured mortgage debts if the surety is a natural person. For their
part, the parties are free to waive this protective mechanism favouring the surety by
making a divergent agreement.159 If, in contrast, the contracting parties do not
actively opt out of amortisation, this protective mechanism favouring the surety
comes to bear as the default rule.
As a matter of principle the surety can also demand the liquidation of any pledges
deposited by the principal debtor before he can be sued (Art. 495 para. 2 and 496
para. 2, Swiss Code of Obligations). Equally, the surety can waive this right in a
contractual agreement.160 In the absence of an active waiver, the insistence on prior
liquidation of pledges remains open to him as a means of defence.
The non-mandatory nature of the stipulations of surety contract law is very much
more extensive, however. The reason for this is a peculiarity of Swiss credit security
law: in accordance with the express will of the legislator, the agreement of a con-
tract of guarantee in place of a surety contract is readily possible, which means that
surety contract law can be circumvented relatively easily.161 Therefore, the whole of
surety contract law is de facto non-mandatory in nature, since the parties can
exclude the applicability of the stipulations in question in globo by means of their
own agreement.
With the contract of guarantee, the guarantor promises a payment to a benefi-
ciary in the event that a certain event described in the contract of guarantee occurs.
Indeed, the parties are free to define the non-repayment of the debt as the event

Kuhn 2011, p. 687.
Kuhn 2011, p. 705.
Kuhn 2011, p. 707.
Kuhn 2011, p. 698.
Kuhn 2011, pp. 707–708.
Kuhn 2011, p. 679 citing Decision of the Swiss Federal Court BGE 129 III 702 E.2.3 and
Gutzwiller 1984, p. 127. Another alternative, because functionally related, is the joint assumption
of liability, Kuhn 2011, p. 678.
146 K. Mathis and P.A. Burri

invoking the guarantee.162 This so-called surety guarantee bears a strong resem-
blance to a true surety contract. From an economic perspective its use achieves the
same result. The main difference from a surety contract in terms of juridical techni-
calities resides in the lack of accessoriness of the guarantee payment to the funda-
mental relationship, i.e. the relationship between debtor and creditor.163 What this
means in particular is that – in contrast to the contract of surety – a guarantee pay-
ment must be made even if the debtor in the fundamental relationship could have
raised objections against the creditor.164 Thus, the guarantee is clearly the “more
severe” security instrument. Furthermore, surety contract law contains numerous
other mandatory (formal) rules in favour of the surety, none of which are applicable
to the guarantee.165
Thus, Swiss law provides two different legal arrangement options for the same,
economically identical, transaction: on the one hand, the form-bound contract of
surety, which provides numerous protection mechanisms for the surety, and on the
other, the contract of guarantee, which is valid without formal requirements and
entirely in keeping with freedom of contract.
This situation may seem unsatisfactory at first glance, i.e. the possibility and
indeed the permissibility of circumventing even the mandatory stipulations of surety
law without problems. Furthermore this outset situation leads to difficult questions
of differentiation in practice. Where there is doubt, have the parties agreed a con-
tract of surety or a contract of guarantee?166 Legal theory criticises the legal situa-
tion as “bad legislation”167; the Swiss Federal Supreme Court talks about an
“inconsistency of the legal order”.168 To address the problem, in its adjudication it
asserted that any “surety relationship” in the private sphere should be presumed to
be a contract of surety in the technical sense, if its legal nature does not clearly
emerge from the agreement or other circumstances.169 If the parties are experienced
in business, on the other hand, or if it is an international agreement, then the reverse
assumption applies: in this case, it is fundamentally presumed to be a guarantee
unless the parties clearly intended to agree a surety contract.170

On the concept and outward forms of the contract of guarantee, Gauch et al. 2008b,
p. 334–335.
Gauch et al. 2008b, p. 337.
Cf. Kuhn 2011, pp. 694, 730 et seqq. and pp. 734 et seqq.
Kuhn 2011, p. 678; if the surety is a natural person, the surety declaration is required to be in
writing (Art. 493 para. 1 Swiss Code of Obligations) and, for amounts over CHF 2,000, drawn up
as a public deed. For the contract of guarantee, the principle of freedom from compliance with any
particular form pursuant to Art. 11 para. 1 Swiss Code of Obligations applies; moreover, surety
contract law overall is characterised by rules for the protection of the surety, Kuhn, p. 69.
Decision of the Swiss Federal Court BGE 113 II 438 E.2. Likewise, the distinction between the
surety contract and the joint assumption of liability is a matter of great uncertainty in practice;
alongside the said presumption, various additional criteria are referred to, see only Kuhn 2011,
pp. 678 et seqq.
Kuhn 2011, p. 679.
Decision of the Swiss Federal Court BGE 129 III 702 E.2.3.
Kuhn 2011, p. 683 citing court rulings.
Kuhn 2011, p. 684 citing court rulings.
8 Nudging in Swiss Contract Law? 147

All in all, surety contract law becomes the default option for people in particular
need of protection. Relatively frequently, these parties – often without knowing it –
will agree a true surety contract and thereby allow the protective norms put in place
by the legislator in favour of the surety to apply.
Despite this de facto non-mandatory regulation, there has been considerable suc-
cess in suppressing ill-considered accommodation surety contracts, and hence
reducing hardship for sureties.171 At the same time, the law leaves broad scope for
surety contracts and guarantees between parties experienced in business, and does
not intervene beyond the necessary extent in security promises between private
individuals. This points to the possibility of employing more non-mandatory law in
general, so as to prevent the conclusion of risky contracts or to alleviate the conse-
quences of such unreflected behaviour.
Once again, however, a caveat is appropriate here on at least two counts. Firstly,
it is important to remember that credit financing in particular is heavily dominated
by banking practice. Banks as parties to business transactions are normally legally
knowledgeable and, furthermore, possess sufficient market power to waive non-
mandatory stipulations largely in their own favour.172 Legal dealings among private
individuals do, however, fundamentally remain a field of application for nudging
via non-mandatory contract law.
Secondly, not all risk-laden or non-self-interested behaviour should be viewed as
suspect with undue haste. Even surety contracts among private persons have their
good side; they can serve as the last resort in situations where there is no other way
out. That the surety exposes himself to a certain risk and shoulders the debtor’s
burden is in the nature of the matter, and need not be evaluated purely negatively by
any means. The very act of accepting major risks to help a good friend can also be
seen as a value in itself.173

8.3 Conclusions

From a legal viewpoint, non-mandatory law is caught between the competing ideas
of autonomy and material justice. Whereas Savigny thought that non-mandatory
law should be oriented exclusively to the hypothetical will of the parties, Bülow and
Stammler called for it to be oriented exclusively to objective ideas of justice.
From the perspective of classical economics, non-mandatory contract law is
purely a means for reducing transaction costs. The law, by offering ready-made
contractual provisions for use by private individuals, releases them from taking
arduous precautions against every minuscule risk so that they can confine themselves

Kuhn 2011, p. 696.
On slippery defaults, see remarks above and Willis 2013, p. 1174.
One only need think of the positive account of the readiness for self-sacrifice and loyalty
between friends in Friedrich Schiller’s “Die Bürgschaft” (The Hostage), on which see Bergenthal
1947, pp. 45 et. seqq.
148 K. Mathis and P.A. Burri

to answering the questions that matter for their concrete contractual relationship. In
effect this theory exhibits a strong convergence with Savigny’s theory, since it simi-
larly suggests that the guideline for the content of non-mandatory contract law
should be the hypothetical result of detailed contract negotiations. This connection,
however, is a purely superficial one. The normative lodestar of economic analysis is
not the autonomy of individuals but the efficiency of the contract concluded and of
the contract-conclusion process itself.
More closely related to the theory of Bülow and Stammler are the suggestions
that can be derived from behavioural economic insights. Here, as there, a purely
laissez-faire approach is assessed critically. But while the theory of Bülow and
Stammler is directed at the normative level and expressly calls for the content of
non-mandatory contract law to be oriented to objective values, the behavioural eco-
nomic analysis emphasises the de facto influence of non-mandatory norms on con-
tracting behaviour, resulting in no clear answer to the question of appropriate
Nudging measures can have the welfare of the parties in mind, in the spirit of
libertarian paternalism, and can particularly seek to protect them from their cogni-
tive deficits. Since nudging does not force the individual into any particular behav-
iour, it constitutes a form of soft paternalism. Nevertheless, the instrument can also
be used in order to realise particular social values, which need not necessarily be in
the interests of the individual concerned. To that extent, nudging is therefore a
double-edged sword.

Acknowledgement We are grateful to Kilian T. Keller, BLaw, for his critical review of the


Alchian, Armen A. 1979. Some implications of recognition of property right transaction costs. In
Economics and social institutions, ed. K. Brunner, 233–252. Boston: Martinus Nijhoff.
Ayres, Ian, and Robert Gertner. 1989. Filling gaps in incomplete contracts: An economic theory of
default rules. The Yale Law Journal 99(1): 87–130.
Belser, Eva Maria. 2000. Freiheit und Gerechtigkeit im Vertragsrecht. Freiburg: Universitätsverlag
Freiburg Schweiz.
Bergenthal, Ferdinand. 1947. In Gericht und Gnade der Wahrheit, Vom Sinn der Meisterballaden
Schillers. Augsburg-Göggingen: Johann Wilhelm Naumann Verlag.
Bucher, Eugen. 1977. Der Ausschluss dispositiven Gesetzesrechts durch vertragliche Absprachen,
Bemerkungen zu den Erscheinungsformen dispositiver Rechtssätze. In Festgabe für Henri
Deschenaux zum 70. Geburtstag, ed. rechts-, wirtschafts- und sozialwissenschaftliche Fakultät
der Universität Freiburg Schweiz, 249–269. Freiburg: Universitätsverlag Freiburg Schweiz.
Bülow, Oskar. 1881. Dispositives Zivilprozessrecht und die verbindliche Kraft der Rechtsordnung.
Archiv für die civilistische Praxis 64: 1–109.
Cerrotini, Eric. 2013. In Commentaire du contrat de travail, Stämpflis Handkommentar, ed. Jean-
Philippe Dunand and Pascal Mahon, 367–377. Bern: Stämpfli Verlag AG.
Coase, Ronald H. 1960. The problem of social cost. Journal of Law and Economics 3: 1–44.
8 Nudging in Swiss Contract Law? 149

Cziupka, Johannes. 2010. Dispositives Vertragsrecht, Funktionsweise und Qualitätsmerkmale

gesetzlicher Regeungsmuter. Tübingen: Mohr Siebeck.
DeJoy, David M. 1989. The optimism bias and traffic accident risk perception. Accident Analysis
and Prevention 21: 333–340.
DiMatteo, Larry A. 2006. Penalties as rational response to bargaining irrationality. Michigan State
Law Review 2006: 883–924.
Dunand, Jean-Philippe. 2013. In Commentaire du contrat de travail, Stämpflis Handkommentar,
ed. Jean-Philippe Dunand and Pascal Mahon, 83–104. Bern: Stämpfli Verlag AG.
Enneccerus, Ludwig, Hans Carl Nipperdey, Theodor Kipp, and Martin Wolf. 1959. Lehrbuch des
Bürgerlichen Rechts, Volume I: Allgemeiner Teil des Bürgerlichen Rechts, Half-volume 1:
Allgemeine Lehren, Personen, Rechtsobjekte. Tübingen: Mohr.
Gauch, Peter, Walter R. Schluep, and Jörg Schmid. 2008a. Schweizerisches Obligationenrecht,
Allgemeiner Teil, Volume I, 9th ed. Zürich/Basel/Geneva: Schulthess Juristische Medien AG.
Gauch, Peter, Walter R. Schluep, and Susan Emmenegger. 2008b. Schweizerisches
Obligationenrecht, Allgemeiner Teil, Volume II, 9th ed. Zürich/Basel/Geneva: Schulthess
Juristische Medien AG.
Gruber, Jonathan. 2002. Smoking’s ‘internalities’. Regulation 25(4): 52–57.
Gutzwiller, Max. 1984. Die Wahlfreiheit zwischen Bürgschaft und Garantie. Zeitschrift für
Schweizerisches Recht 103: 121–131.
Horn, Karen. 2013. Sklavenhalter der Zukunft, Wenn Verhaltensökonomen und
Ökonomisierungskritiker gemeinsame Sache machen. In Frankfurter Allgemeine Zeitung, 1
March 2013 51: 14.
Huber, Ernst Rudolf. 1933. Bedeutungswandel der Grundrechte. Archiv des öffentlichen Rechts 23:
Huguenin, Claire. 2012. Obligationenrecht – Allgemeiner und Besonderer Teil. Zürich: Schulthess
Juristische Medien AG.
Jolls, Christine. 2009. Behavioral law and economics. In Law and economics, ed. Alain Marciano,
71–79. New York: Routledge.
Jolls, Christine, and Cass R. Sunstein. 2006. Debiasing through law. Journal of Legal Studies 35:
Jolls, Christine, Cass R. Sunstein, and Richard H. Thaler. 1998. A behavioral approach to law and
economics. Stanford Law Review 50: 1471–1550.
Kähler, Lorenz. 2010. Zur Durchsetzungskraft abdingbaren Rechts. In Wie wirkt Recht?
Ausgewählte Beiträge zum ersten gemeinsamen Kongress der deutschsprachigen
Rechtssoziologie-Vereinigungen, Luzern 4.-6. September 2008, ed. Michelle Cottier, Josef
Estermann, and Michael Wrase, 431–448. Baden-Baden: Nomos.
Kähler, Lorenz. 2012. Begriff und Rechtfertigung abdingbaren Rechts. Tübingen: Mohr Siebeck.
Kahneman, Daniel, and Amos Tversky. 1979. Prospect theory: An analysis of decision under risk.
Econometrica XVLII: 263–292.
Kahneman, Daniel, Jack L. Knetsch, and Richard H. Thaler. 1991. Experimental tests of the
endowment effect and the coase theorem. Journal of Political Economy 98: 1325–1348.
Kirchgässner, Gebhard. 2008. Homo Oeconomicus. Das ökonomische Modell individuellen
Verhaltens und seine Anwendung in den Wirtschafts- und Sozialwissenschaften, 3rd ed.
Tübingen: Mohr Siebeck.
Korobkin, Russel. 1998. The status quo bias and contract default rules. Cornell Law Review 83:
Korobkin, Russel. 2000. Behavioral economics, contract formation and contract law. In Behavioral
law and economics, ed. Cass Sunstein, 116–143. Cambridge/New York/Port Melbourne/
Madrid: Cambridge University Press.
Kötz, Hein. 2003. Der Schutzzweck der AGB-Kontrolle – Eine rechtsökonomische Skizze. JuS 3:
Kuhn, Hans. 2011. Schweizerisches Kreditsicherungsrecht. Bern: Stämpfli Verlag AG.
Laband, Paul. 1888. Zum zweiten Buch des Entwurfes eines bürgerlichen Gesetzbuches für das
Deutsche Reich. Archiv für die civilistische Praxis 73: 161–208.
150 K. Mathis and P.A. Burri

Laibson, David. 1997. Golden eggs and hyperbolic discounting. The Quarterly Journal of
Economics 112(2): 443–477.
Listokin, Yair. 2009. What do corporate default rules and menus do? An empirical examination.
Journal of Empirical Legal Studies 6(2): 279–308.
Lüdemann, Jörn. 2007. Die Grenzen des homo oeconomicus und die Rechtswissenschaft. In Recht
und Verhalten, ed. Christoph Engel, Markus Englerth, Jörn Lüdemann, and Indra Spiecker
genannt. Döhmann, 7–59. Tübingen: Mohr Siebeck
Luth, Hanneke A. 2010. Behavioural economics in consumer policy, the economic analysis of
standard terms in consumer contracts revisited. Antwerp: Intersentia.
Mathis, Klaus. 2009. Efficiency instead of justice? Searching for the philosophical foundations of
the economic analysis of law. Dordrecht/Heidelberg/London/New York: Springer.
Mathis, Klaus, and Alain Anderhub. 2009. Die Kosten der Einzelfallgerechtigkeit. Praktikabilität
des Rechts aus rechtlicher und ökonomischer Sicht. Schweizerische Zeitschrift für
Sozialversicherung und berufliche Vorsorge 53: 301–330.
Mathis, Klaus, and Fabian Diriwächter. 2011. Is the rationality of judicial judgements jeopardized
by cognitive biases and empathy? In Efficiency, sustainability, and justice to future genera-
tions, ed. Klaus Mathis, 55–73. Dordrecht/Heidelberg/London/New York: Springer.
Mathis, Klaus, and Ariel David Steffen. 2015. From rational choice to behavioural economics,
theoretical foundations, empirical findings and legal implications. In European perspectives on
behaviorual law and economics, ed. Klaus Mathis, 31–48. Cham/Heidelberg/New York/
Dordrecht/London: Springer.
McFadden, Daniel. 1999. Rationality for economists? Journal of Risk and Uncertainty 19: 73–105.
Möslein, Florian. 2010. “Governance by Default”, Innovation und Koordination durch dispositives
Recht. In Unternehmen, Markt und Verantwortung. Festschrift für Klaus J. Hopt zum 70.
Geburtstag am 24. August 2010, vol. 2, ed. Stefan Grundmann et al., 2861–2880. Berlin/New
York: De Gruyter.
Posner, Eric A. 2006. There are no penalty default rules in contract law. Florida State University
Law Review 33: 563–587.
Raiser, Ludwig. 1935. Das Recht der Allgemeinen Geschäftsbedingungen. Hamburg: Hanseatische
Samuelson, William, and Richard J. Zeckhauser. 1988. Status quo bias in decision making. Journal
of Risk and Uncertainty 1: 7–59.
Schäfer, Hans-Bernd, and Claus Ott. 2012. Lehrbuch der ökonomischen Analyse des Zivilrechts,
5th ed. Berlin/Heidelberg: Springer Verlag AG.
Schmid, Jörg, and Hubert Stöckli. 2010. Schweizerisches Obligationenrecht, Besonderer Teil.
Zürich/Basel/Geneva: Schulthess.
Schnellenbach, Jan. 2012. Weicher Paternalismus? Normative Implikationen beschränkter indivi-
dueller Rationalität. Wirtschaftswissenschaftliches Studium (WiSt) 41(11): 594–600.
Schönenberger, Wilhelm, and Jäggi, Peter. 1973. Zürcher Kommentar zum schweizerischen
Zivilgesetzbuch, vol. V, Obligationenrecht, Subvolume V 1a, 3rd edn. Zürich: Schulthess.
Schwenzer, Ingeborg. 2012. Schweizerisches Obligationenrecht. Allgemeiner Teil. Bern: Stämpfli
Verlag AG.
Stammler, Rudolf. 1886. Der Garantievertrag. Archiv für die civilistische Praxis 69: 1–141.
Sunstein, Cass R. 2002. Switching the default rule. New York University Law Review 77:
Sunstein, Cass R., and Richard H. Thaler. 2003. Libertarian paternalism is not an oxymoron. The
University of Chicago Law Review 70: 1159–1202.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge. Improving decisions about health, wealth,
and happiness. New Haven/London: Yale University Press.
Tversky, Amos, and Daniel Kahneman. 1974. Judgment under uncertainty: Heuristics and biases.
Science 185: 1124–1131.
Tversky, Amos, and Daniel Kahneman. 1979. Prospect theory: An analysis of decision under risk.
Econometrica 47(2): 263–292.
8 Nudging in Swiss Contract Law? 151

Tversky, Amos, and Daniel Kahneman. 1981. The framing of decisions and the psychology of
choice. Science 211: 453–458.
Unberath, Hannes, and Johannes Cziupka. 2009. Dispositives Recht welchen Inhalts? Antworten
der ökonomischen Analyse des Rechts. Archiv für die civilistische Praxis 209: 37–83.
von Savigny, Friedrich Carl. 1840. System, des heutigen römischen Rechts. Berlin: Veit.
Weigel, Wolfgang. 2003. Rechtsökonomik. München: Franz Vahlen Verlag GmbH.
Willis, Lauren E. 2013. When nudges fail: Slippery defaults. The University of Chicago Law
Review 80: 1115–1229.
Windscheid, Bernhard. 1887. Lehrbuch des Pandektenrechts. Frankfurt am Main: Rütten &
Chapter 9
Designing Disclosures
Testing the Efficacy of Disclosure in Retail Investment

Geneviève Helleringer

Abstract This chapter explores the psychology of disclosure, as a nudging instru-

ment, in the context of advised financial investment decisions. It investigates how
information about the financial advisor’s potential conflict of interest impacts on the
participants’ judgement, whether as to attitudes or actions. Contrary to what is com-
monly suggested, problems of conflict of interest in the financial investment context
may efficiently be treated by means of disclosures, provided that such disclosures
explicit the consequences of the conflict for the decision maker. It remains to anal-
yse how explanatory disclosures can be drafted so that their influence on the trust in
the advice also translates into modified action.

9.1 Introduction

Important work has been conducted on the ways in which people deal poorly with
information, facts and numbers, opening the door to ‘nudging’ policies that com-
pensate for such limitations or even leverage them.1 Disclosure requirements,
alongside default rules and simplification, are among the concrete manifestations of
a low-cost, choice-preserving, behaviourally informed approach to regulatory
problem, also known as ‘nudge’.2 However, to realize their potential as behav-
iourally informed regulatory tools and helpfully contribute to shape the background
against which decisions are made, disclosures need to be take into account insights

Thaler and Sunstein 2008; Calo 2014.
Thaler and Sunstein 2008; Sunstein 2013.
G. Helleringer (*)
Essec Business School, av Bernard Hirsch, CS 50105 Cergy,
95021 Cergy-Pontoise Cedex, France
Institute of European and Comparative Law, Oxford University,
St Cross Road, OX1 3UL Oxford, UK

© Springer International Publishing Switzerland 2016 153

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_9
154 G. Helleringer

about how information is processed or they are likely to lack efficacy.3 The present
chapter focuses on whether people are good at evaluating the source of the informa-
tion.4 Individuals often make decisions after consulting with, or being influenced by,
others. ‘Judge-advisor system’ controlled experiments have enabled to analyse dif-
ferent aspects of the effect of advice: whether people act upon the advice they are
given or discount it; how confident advisors and decision makers are; how accurate
the final decision, etc.5 The ‘judge’ in this context refers to the person who receives
the advice, and must decide how to treat it to make a decision. In order to model the
decision-making structures in an advice-giving/advice-taking context, expertise dif-
ferences between judges and advisors, individual pre-advice decision and other
independent variables have been manipulated. Research in this area has also focused
on the potential incongruence between decision-maker’s goals and the advisors’,
i.e. the ‘agency problem’ in the terminology of agency theory,6 which raises impor-
tant legal issues.7 Experiments have been conducted to explore specific relation-
ships in which advice might be biased because of a conflict of interest,8 such as the
patient-doctor relationship. In this context, it has been demonstrated that even
advice that is known to be biased may nevertheless influence a patient’s judgement.9
This persistent bias is the result of four main factors: first, the difficulty of counter-
ing bias,10 and anchoring in particular11; secondly, the inability to unlearn false or
misleading information12; thirdly, the fact that advice from a biased advisor is not
sufficiently discounted, since people assume that an honest advisor will disregard
his own interest13; and fourthly, the fact that advice is not necessarily discounted
more when the advisor’s conflict is voluntarily disclosed than when it is not dis-
closed, the due to a specific perverse effect of disclosure: the ‘insinuated anxiety
effect’.14 While disclosure can decrease the advisee’s trust in the advice, it can
simultaneously increase pressure to comply with the same advice. It is believed to
result from the fact that advice recipients feel obliged to help satisfy their advisor’s
personal interests after such advisor has disclosed them. This work adds to the scep-
ticism towards mandatory disclosures to treat conflict of interests15 and extends such

Ben-Shahar and Schneider 2010.
For a review of work about how people evaluate advice, see Bonaccio and Dala 2006; see also the
seminal experimental data on advertising, Harris 1977.
Harvey and Fischer 1997 and Yaniv 2004.
Moore et al. 2010.
Issacharoff 2010.
Moore et al. 2010.
Loewenstein et al. 2012 and 2013.
Mussweiler and Strack 2000.
Tversky and Kahneman 1974; Strack and Mussweiler 1997; Cain et al. 2004.
Mussweiler 2003; Camerer 2003.
Ross and Nisbett 1991.
Sah et al. 2012 and 2013.
Ben-Shahar and Schneider 2010, 2014.
9 Designing Disclosures 155

scepticism towards voluntary disclosures. It reinforces the common idea that disclo-
sure can burden those it is intended to protect.
Other asymmetric relationships remain relatively underexplored: for example,
whether people make good use of information and advice in the context of financial
decision-making has been little researched.16 A recent study17 using trading and
survey data from a sample of customers of a large German bank, found that over
half of the surveyed customers stated that they consistently relied on the advice of
their professional advisor. A survey18 showed however that more than half of the
individual investors in retail finance products thought that the financial advisers
gave completely independent advice or information. It seems therefore that inves-
tors are largely ignorant of conflicts of interest. The present study focuses on the
extent to which people allow for the information they receive about the goals of the
financial advisor and whether the advisor’s goals are congruent with their own. Do
they assess the objective reliability of a financial advice in the opinion they form
about the advice and in the investment decisions they make? In attempting to answer
this question, this paper explores the psychology of receiving and making use of
The question is particularly relevant in the context of retail investment decisions,
since professional advice is ubiquitous19 and advisors are subject to competing
motives. The professional has a duty to give objective and impartial advice to the
client, but his interest in maximising his own gain may be in conflict with this duty.
In particular, the advisor’s fee structure will exacerbate the conflict if the fee is
calculated by reference to the sum invested rather than the gain made by the client
or as a flat fee.20
The present chapter offers insight into the following question: is mandatory dis-
closure of professional conflicts of interest a desirable policy approach to deal with
the problem of conflicts of interest? More specifically, to what extent which people
take into account information about the goals of the financial advisor and whether
the advisor’s goals are congruent with their own: can one successfully assess the
objective reliability of a financial advice? In Sect. 9.2., I present background infor-
mation relating to this question and describe the design and result of an original
experiment that was designed to assess whether making explicit divergence of
objectives has an influence on the investors ‘behaviour. In Sect. 9.3, I present the
main findings. Finally in Sect. 9.4, I highlight public policy implications before
concluding in Sect. 9.5.

For a thorough presentation of the role, practices, and related issues, of intermediaries, see Judge
Hacketal et al. 2009.
Chater et al. 2010.
Judge 2015.
For a review, see Chater et al. 2010.
156 G. Helleringer

9.2 Testing Disclosure Designs

9.2.1 Background

Legal disclosures regarding conflicts of interest can be strengthened in various

ways, up to the point of imposing a compulsory ‘health warning’ that spells out why
the advisor is not, and cannot be, fully impartial and objective.21 However, such
disclosure requirement assumes that people will be able to take advantage of the
disclosure to make better decisions. It assumes that they will not only pay attention
to the disclosure, encode and remember it, but also actively recall and use the infor-
mation when making their decisions.22 There is an ambiguity on these issues. Some
experimental data provide evidence that an advisee can follow such advice, namely
if the advisee believes that benefits will outweigh costs.23 For instance, decision
makers are expected to make an effort to adjust from their pre-advice decision in
response to an advice they received, if the stakes for making a correct decision are
high and if the advisor is known to be trustworthy. However, other studies provide
evidence that advisees do not follow their advisors’ recommendations as much as
they should have to benefit from them and discount the advice instead. Among the
commonly evidenced mechanisms are the fact that discounting of advice increases
as the distance between the decision maker’s opinion and the advisor’s recommen-
dations increases,24 which is an effect that is similar to the effect identified in classic
studies of attitude change.25 Also decision makers tend to discount recommenda-
tions all the more when they are very different from the recommendations of other
advisors.26 It was suggested27 that over-reliance by advisees on their initial opinion
and the corresponding discount of advice could be a function of how the informa-
tion is encoded (e.g., the information is relative to the pre-advice opinion or it is a
constant), the complexity of the information (e.g., how many piece of advice were
received), how the information is processed (e.g., in a step-by-step manner or advi-
see have to wait until after all of the advice has been received).

For experimental data, see Chater et al. 2010.
Issacharoff 2011.
Payne et al. 1993; Harvey and Bolger 2001; Schrah et al. 2006.
Yaniv 2004.
Bochner and Insko 1966; Scherif and Hovland 1961.
Harries et al. 2004.
Bonnacio and Dalal 2006.
9 Designing Disclosures 157

9.2.2 Experiment Presentation and Hypotheses

The reported experiment aimed at exploring the way in which compulsory disclo-
sure of potential biases in the advice influences the judgement of an individual
investor in two respects: first, how does it affect trust in the received advice? And
secondly, how does it affect the perceived likelihood that he will actually act on the
The experiment was computer-based: participants were shown fictional adver-
tisements accompanied by investment advice from a named advisor. They were
asked, in each case, to rate their trust in the advisor and their willingness to invest
in the advertised product. The experiment manipulated (i) the strength of language
for the advice to invest provided by the advisor, (ii) the existence of a conflict of
interest for the advisor deriving from their fee structure, and (iii) the salience of the
conflict of interest via an ‘explanatory disclosure’.
The experiment was designed to assess whether drawing the investor’s attention
to the conflict of interest in writing would impact on participants’ ratings of trust or
perceived willingness to invest. As the experiment did not require participants to
make actual investments, the data reflect attitudinal answers, i.e. the participants’
stated willingness to invest and their stated trust in the advisor, rather than behav-
ioural answers.28
There were good reasons to expect that decision makers would discount financial
advice to a greater extent if they know, or suspect, that their advisor’s goals differ
from their own. There is empirical evidence that such is the case in different con-
texts.29 There were however also good reasons to expect that decision makers might
still follow the financial advice if they knew of the conflict of interest pursuant to
the insinuated anxiety effect described above. There is empirical evidence support-
ing such predictions in different contexts.30 Also, trust in the advice reflects an atti-
tude, whereas willingness to invest reflects an action and perceived willingness to
invest, the perception of the probability to act. Hence trust and willingness to invest,
or perceived willingness to invest, are potentially dissociable variables.31 It was
therefore difficult to firmly predict whether, in a financial context, investors’ trust in
the advice and willingness to invest would be negatively related to the existence of
a conflict of interest.

For details about the collected data and statistical analyses, see Helleringer, Trust Me I Have a
Conflict of Interest, forthcoming.
Sniezek and Van Swol 2001.
E.g., patient – doctor relationship: Sah et al. 2012, 2013.
Hewstone et al. 2012.
158 G. Helleringer

9.3 Main Findings

The experiment was designed to assess whether drawing the investor’s attention to
the conflict of interest in writing would impact on participants’ ratings of trust or
willingness to invest. It was however difficult to firmly predict whether, in a finan-
cial context, investors’ trust in the advice and willingness to invest would be nega-
tively related to the existence of a conflict of interest.
The experiment set out to experimentally test the impact of conflicts of interest
and disclosure of such conflicts in a simplified context involving advice for invest-
ment. The experiment mirrored a real-life situation: information search by would-
be retail investors is not in-depth and does not involve shopping around for the best
deal. It relies on financial websites or magazines and, above all, staff of banks,
insurance or other financial institutions for advice.32 However such financial advice
may not always be disinterested. Advisors’ monetary incentives are apt to create
interests that are not aligned with the investors’ interests. The experiment manipu-
lated the disclosure to the investor of how many his goals and those of the advisor
were incongruent because of the fee structure. The experiment thereby gave some
insights into the potential impact of a regulatory tool e.g., mandatory explanatory
disclosures of the concrete implications of the advisor’s fee structure for the deci-
sion makers.
Participants provided a rating of their trust in the advice out of a scale of one to
seven after reading a financial advertisement, which displayed a handwritten note
making a recommendation to invest. The advertisement also set out the fee structure
of the advisor. In half of the cases, the advertisement included an additional note,
one version of which explained that the interests of the advisor were aligned with
those of the participant, and the other version of which explained that their interests
were not so aligned.

9.3.1 Limited and Conditional Effectiveness of Disclosures

and Advice

Disclosures are more likely to be effective when they include an explanatory disclo-
sure that spells out for the advice receiver the ‘take home message’ from the disclo-
sure. Beyond simplified disclosures, explained disclosures with an explicit
description of implications for the advice receiver are therefore more effective. One
needs however to bear in mind that the relationship is dyadic: previous research
suggests that experts who have disclosed their biases may give biased advice more
freely and indulge their private interest.33 Hence the beneficial caution derived from

Chater et al. 2010.
Cain et al. 2005.
9 Designing Disclosures 159

warning disclosures might, in practice, be offset by the fact that the advice they get
is correspondingly even more biased.
Explanatory or ‘Warning-health’ types of disclosure were found to be effective,
whereas impersonal information about the fee structure was not. The difference
between the two types of information was that the first type made explicit a point
that was only implicit in the fee structure information. Hence it appears that busi-
ness school students are pretty bad at seeing what this implication of the fee struc-
ture is. One can expect that results would only be weaker in a more diverse sample.
This finding comes as a support for numerous works on information overload in the
financial context34 and in relation with advisors’ commission in particular.35 It calls
for future research on how to convey information in this context.

9.3.2 Differences Between Trust and Willingness to Invest

Two theoretical observations support the hypotheses that collected data show a dif-
ference in the effect of disclosure between willingness to invest and trust. First,
there is a discrepancy between what people think (in theory) and how they act (in
practice). It has been shown that professional investors rate harshly the quality of
rating agencies’ grading. Nevertheless, their pattern of investment collectively
closely follow the agencies’ ratings.36 Similarly, the present data suggests that a
sharp decrease in trust conditions is not matched by the same drop in willingness to
invest. This might be explained by the fact participants are mainly business school
students, who are likely to believe they can independently assess how interesting
the investment opportunity might be. The fact that they spent a rather long time
reading each advertisement supports this hypothesis. It would be interesting to test
whether the decoupling between trust and willingness to invest is a feature of pro-
fessional investors and investors trained in business school. Testing a different
sample, e.g. arts students, might be of interest. It would also be interesting to test
whether the cultural background – high or low interpersonal trust culture – makes a
Second, and more substantively, willingness to invest should logically depend
both on how much one trusts the advisor and on one’s judgement of how good the
investment is. The difference between the two measures might be greater for a busi-
ness school students sample than for other groups, because business students might
place more weight on their independent analysis and rely relatively less on the rec-
ommendation of an advisor. Such a hypothesis would require to be tested.
Third, the question of how much participants trust the advice could be described
as an abstract assessment, because it could be corrected based on the outcome of the
investment. For this reason, the collected score captured the participants’ trust ‘in’

Marotta-Wurgler 2010; Mercer et al. 2010, Beshears et al. 2009.
Lacko and Pappalardo 2004.
see e.g. Hill 2009.
160 G. Helleringer

the advisor, i.e. an assessment of an internally based attribute, based on beliefs as to

what motivates the other or the other’s values. Trust ‘in’ differs from trust ‘that’ the
other will have or had a certain behaviour, i.e. a prediction regarding the other’s
A follow-up would include a factor analysis to identify latent variables explain-
ing the variable ‘trust’ and how it is structured (e.g. how does it relates to expertise,
past experience, etc.).

9.3.3 Contrast Between One-Shot Judgment and Upon

Multiple Comparison Judgment

Finally, there is a difference between the result found in the within-subject analysis
and the between-subject analysis. There are two possible interpretations of the
within vs. between differences: interesting psychological insight or demand
A substantial explanation may be suggested to explain that, although willingness
to invest and trust were correlated, they were impacted in an asymmetric manner by
disclosure. In the ‘between-subject’ design, data captured what happens in the first
round: that is before participants had had a chance to experience other conditions.
By contrast, the ‘within-participants’ analysis is critically about what happens when
subjects get to experience a range of conditions as they make judgements. The dif-
ference in the results may reflect a difference between the way in which people
make decisions regarding stated attitude (whether to trust the advice), and stated
behaviour (whether to invest). Trusting a piece of advice reflects the confidence that
the advisor can adequately provide a judgement in the considered context.38 Effects
on this variable, or as in our experiment on the perceived trust, may be adequately
captured in a between-subject analysis. Deciding to invest, however, represents a
different psychological process from trusting and because there is a risk of loss.
This remains true at the level of perceived probability to invest that is reflected in
the present data. Even if in practice investors do little shopping around and often do
not consider many options to find the best one.39
When expressing their stated intention, they are likely not to choose the first
opportunity offered. This behaviour gives them the opportunity to compare other
options. The implication is that effects on the WTI variable may be more adequately
captured in the within-subject analysis.
However, another interpretation cannot be ruled out. In the within-design
approach, participants perform for every condition. They may try to decipher the
purpose of the experiment and behave as they feel they are asked to. In this

For this distinction, see Hill and O’Hara 2005.
Hardin 2001; Bacharach and Gambetta 2001.
Retail Investment Purchase Process European survey, in Chater et al. 2010.
9 Designing Disclosures 161

perspective, the findings of the present study have to be treated with caution. Now
that general trends have been identified, an incentivised controlled experiment
could be set up to refine the results on the basis of backed hypotheses.

9.4 Public Policy Implications

The presented study has generated insights into the mechanisms underlying the
reception of potentially biased financial advice. These insights present some rele-
vant information for policy makers for regulating financial advice: they have
increasingly been empowered to do so and need to rely on efficient tools. Typically
the US Consumer Financial Protection Bureau (CFPB) is authorised pursuant to the
2010 Dodd-Franck Act to ensure that ‘consumers are provided with timely and
understandable information to make responsible decisions about financial transac-
tions’. The CFBP shall also ensure that information is “fully, accurately, and effec-
tively disclosed to consumers in a manner that permits consumers to stand the costs,
benefits, and risks associated with the product of service, in light of the facts and
circumstances”. To be attained these goals demand that precise behavioural studies
be conducted.
The described experiment tends to support disclosure at a prescriptive level, as it
shows that some information in writing may help investor discount biased advice.
The fact that the explanatory disclaimers about the presence of a conflict of interest
did not make a significant difference in the perceived willingness to invest might be
understood as evidence that disclosures on this topic do not change anything in the
outcome and that there are therefore no reason to require them. However the fact
that there is a significant difference in the perceived trust levels leads to a different
conclusion. First, trust in financial advisor and in intermediaries in general is a com-
ponent of trust in the market, which is an important component of capitalist econo-
mies and a policy goal that is frequently stressed by legislators and regulators. EU
directives in consumer law systematically refer to it. Some studies have shown that
trust in the market does impact investors’ and consumers’ behaviour. Second, the
effect found on perceived willingness to invest in the within-subject design gives an
indication that, under certain conditions, there is some correlation between the per-
ception of a conflict of interest and the contemplated investment. Thirdly, the design
and framing of the explanatory disclaimer is a dimension that could be improved in
order to make the consequence about the conflict of interest more salient. A recent
study stressed the specific difficulty experienced by subject to gasp the problem
underlying disclosures as compared to good ability to process information provided
in such disclosures.
The described effects qualify the scepticism that has developed over half a
decade towards the effectiveness of disclosure to influence behaviours.40 It supports

Ben-Shahar and Schneider 2010; Sah et al. 2012 and 2013.
162 G. Helleringer

behavioural findings along with theoretical considerations relating to people’s

autonomy in their decision making, pursuant to which the more carefully disclosure
(e.g., by resorting to certain findings of neuroscience to increase their salience) are
delivered the more effect they have.41
Against the backdrop of ambiguous findings in other fields, the present study
aimed at ascertaining general trends in the use of disclosure and financial advice.
This approach justifies to use reasonably abstract (e.g., there was no significant
distinction between the products advertised) and simplified (e.g., the fee structure
was basic and transparent, whereas commissions are largely hidden) decision-
making scenarios. If, as currently discussed, the future new European directive on
markets in financial instruments (MiFID) were to mandate disclosure of conflict of
interest, such as commission payments to financial advisors,42 future research
should be conducted to determine the best form and format for disclosure. The find-
ings of the present study evidenced the need for explanatory disclosure. Subsequent
experimental work with more realistic financial products and with representative
samples of consumers should ideally be conducted to optimise the implementation
of any proposed policy intervention.
It is worth stressing, however, that although the experiment’s paradigm tested
sensitivity to, and use of, disclosed information in the context of financial invest-
ment, the purpose was not to replicate perfectly a real investment situation. Any
policy conclusions that may be drawn would need to take account of the potential
for optimism bias in experiments43 and in stated attitude (trust) or act (willingness
to invest) in particular. Any policy recommendation must be based on (i) a set of
assumptions about how consumers and firms behave in a considered market, and (ii)
theories of how policy interventions will interact this behaviour.44 For instance, the
present study deals with the conflict of interest between advisors and advised cus-
tomers and how disclosure of this conflict of interest may affect consumers’ reac-
tion to advice. The question for policy makers to consider is not only to what extent
consumers wary about such conflict, but also which form of disclosure will trigger
a reaction and will improve efficiency. As there is little scope to test such policy in
the real world, randomised, controlled experiments represent a valuable alternative.
Manipulating factors of interest in study a enables to directly test and to quantify, in
relative terms, the potential impact of policy measures on the behaviour of consum-
ers. Such experimental evidence can be used to determine more effective policy
remedies to improve an identified market failure. The interaction between conflict
and explanatory disclosure revealed by the analyses (within-subjects analysis)
stresses the importance of this observation.

Calo 2012.
Inderst and Ottaviani 2012.
Sharot 2011.
Chater et al. 2010.
9 Designing Disclosures 163

9.5 Conclusion

Previous research has been inconclusive and inconsistent as to the efficacy of dis-
closure and the effects of biased advice on advisees. In the doctor-patient context,
evidence has suggested that there is little value in disclosure; by contrast, research
in the judge-advisor context has indicated a positive effect. The described study
explored the psychology of disclosure in the context of advised financial investment
decisions. It investigated how information about the financial advisor’s potential
conflict of interest impacts the participants’ judgement, whether as to attitudes or
Participants were asked to read a financial advertisement that displayed a hand-
written note recommending that they should invest; they then rated their trust in the
advice on a scale of one to seven after. The advertisement also set out the fee struc-
ture of the advisor. In half of the cases, the advertisement included an additional
note, one version of which explained that the interests of the advisor were aligned
with those of the participant, and the other version of which explained that their
interests were not so aligned.
The results of the experiment reveal, first, that participants’ ratings of their will-
ingness to invest was reliably affected by the presence of conflicts of interest only
when the specific implication of the conflict was made explicit in the explanatory
disclosure; and secondly, that their rating of trust in the advice was in all cases
affected by the presence of a known conflict of interest, but more strongly so when
accompanied by an explanatory disclosure. Even then, in ratings relating to the first-
seen advertisement, the disclosure primarily affected stated trust in the advisor
rather than willingness to invest, only affecting the latter when participants had the
opportunity to compare offers, advice and disclosures across multiple
The results suggest that, from the perspective of the recipients of advice, prob-
lems of conflict of interest may not be as impossible to treat by means of mandatory
disclosure as has been commonly suggested.45 Reported failures of disclosure to
improve the quality of decision may turn out to reflect inadequate disclosure design.
Our findings suggest that what needs to be disclosed is not the existence of a conflict
but the consequences of the conflict for the decision maker.
Furthermore, it is now recognised that the only feasible solution to the problem
of conflict of interest is to focus attention on the recipient of the advice. Recent
experimental work focusing on the effect of disclosure as perceived by the advisor
has revealed that it is not sufficient for the advisor to attempt to consciously com-
pensate for biasing influences on his judgement: for psychological reasons, he may
not be able to make adequate correction for his own bias despite his best efforts. The
problems arising out of conflict of interests may be more profound that is commonly
assumed.46 In a capitalist world (in which professional advice is ubiquitous and

Ben-Shahar and Schneider 2010.
Moore et al. 2010.
164 G. Helleringer

usually given for a commission) it is not possible to eliminate conflicts of interest

by neutralising every partisan allegiance: well-designed forms of disclosure may
have to be recognised as the only practical solution. It remains to analyse how
explanatory disclosures can be drafted so that their influence on the trust in the
advice also translates into modified action.


Anderson, L., and J. Mellor. 2009. Are risk preferences stable? Comparing an experimental mea-
sure with a validated survey-based measure. Journal of Risk and Uncertainty 39: 137–160.
Bacharach, M., and D. Gambetta. 2001. Trust in signs. In Trust in society, ed. K. Cook, 148–184.
New York: Russell Sage Foundation.
Bar-Gill, O. 2004. Seduction by plastic. Northwestern University Law Review 98: 1373–1434.
Barker C, Pistrang N, and R. Elliott. 2002. Research method in clinical psychology, 2nd ed.
New York: Wiley.
Baumeister, F., K. Vohs, and D. Funder. 2007. Psychology as the science of self-reports and finger
movements: Whatever happened to actual behavior, perspectives on psychological. Science 2:
Ben-Shahar, O., and E. Schneider. 2010. The failure of mandated disclosure. University of
Pennsylvania Law Review 659: 704–728.
Ben-Shahar, O., and E. Schneider. 2011. The failure of mandated disclosure (2010). 159 University
of Pennsylvania Law Review 659: 704–728.
Ben-Shahar, O., and E. Schneider. 2014. More than you wanted to know. The failure of mandated
disclosure. Princeton/Oxford: Princeton University Press.
Berg, J., J. Dickhaut, and K. McCabe. 2005. Risk preference instability across institutions: A
dilemma. Proceedings of the National Academy of Sciences 102: 4209–4214.
Beshears, John, James J. Choi, David I. Laibson, and Brigitte C. Madrian. 2009. How does simpli-
fied disclosure affect individuals’ mutual fund choices? , NBER working paper series 14859.
Cambridge, MA: National Bureau of Economic Research.
Bochner, S., and C. Insko. 1966. Communicator discrepancy, source credibility, and opinion
change. Journal of Personality and Social Psychology 4: 614–621.
Bonaccio, S., and R.S. Dalal. 2006. Advice taking and decision-making: An integrative literature
review, and implications for the organizational sciences. Organizational Behavior and Human
Decision Processes 101: 127–151.
Button, K., Ioannidis, J., Mokrysz, C., Nosek, B., Flint, J., Robinson, E., and M. Munafò. 2013
May. Power failure: Why small sample size undermines the reliability of neuroscience. Nature
Reviews Neuroscience 14: 365–376.
Calo, R. 2012. Against notice skepticism in privacy (and elsewhere). Notre Dame Law Review
87(3): 1027–1072.
Calo, R. 2014. Code, nudge, or notice? Iowa Law Review 99(2): 773–802.
Cain, D.M., G. Loewenstein, and D.A. Moore. 2005. The dirt on coming clean: Perverse effects of
disclosing conflicts of interest. The Journal of Legal Studies 34: 1–25.
Cass Sunstein, R. 2013. Simpler. New York: Simon and Schuster.
Chater N, Huck S, and R. Inderst. 2010. Consumer decision-making in retail investment services:
A behavioural economics perspective. Report to the European Commission SANCO.
Chugh, D., M. Bazerman, and M. Banaji. 2005. Bounded ethicality as a psychological barrier to
recognising conflicts of interests. London: Cambridge University Press.
Cialdini, R. 2001. Influence: Science and practice. London: Allyn and Bacon.
9 Designing Disclosures 165

Camerer, Colin F. 2003. Psychology and economics. Strategizing in the brain. Science (New York,
N.Y.) 300(5626): 1673–1675.
Cohen, J. 1988. Statistical power analysis for the behavioural sciences, 2nd ed. New York:
Academy Press.
Dulleck, U., Fell, J., and J. Fooken. 2011 October. Within-subject intra- and inter-method consis-
tency of two experimental risk elicitation methods. NCER Working Paper 74.
Erta, Hunt, and W. Brambley. 2013 April. Applying behavioural economics at the financial con-
duct authority. April 2013. Occasional Paper No.1. Financial Conduct Authority.
Field, A., and G. Hole. 2003. How to design and report experiments. London: Sage Publications.
Fisch, J.E., and T. Willkinson-Ryan. 2013 July. Why do retail investors make costly mistakes? An
experiment on mutual fund choice, ECGI Working Paper Series in Law, n° 220/2013.
Hacketal, A., Inderst, R., and S. Meyer. 2009. The hidden price of financial advice. New York:
Hardin, R. 2001. Conceptions and explanations of trust. In Trust in society, ed. K. Cook, 3–39.
New York: Russell Sage Foundation.
Harries, C., I. Yaniv, and N. Harvey. 2004. Combining advice: The weight of a dissenting opinion
in the consensus. Journal of Behavioural Decision making 17: 333–348.
Harris, R. 1977. Comprehension of pragmatic implications in advertising. Journal of Applied
Psychology 62(5): 603–608.
Harvey, N., and F. Bolger. 2001. Collecting information: Optimizing outcomes, screening options
or facilitating discriminations? Quarterly Journal of Experimental Psychology 54: 269–301.
Harvey, N., and I. Fischer. 1997. Taking advice: Accepting help improving judgement, and sharing
responsibility. Organizational Behavior and Human Decision Processes 70: 117–133.
Hewstone, M., Stroebe, W., and K. Jonas. 2012. Introduction to social psychology, 5th ed. Wiley.
Hey, J., A. Morone, and U. Schmidt. 2009. Noise and bias in eliciting preferences. Journal of Risk
and Uncertainty 39: 213–235.
Hill, Claire. 2009. Why did anyone listen to the rating agencies after Enron? Journal of Business
and Technology Law 4: 283.
Hill, Claire, and O’Hara Erin Ann. 2005. A cognitive theory of trust. Minnesota Legal Studies
Research Paper No. 05-51.
Inderst, R., and M. Ottaviani. 2012. Regulating financial advice. European Business Organization
Law Review 13: 237–246.
Isaac, M., and D. James. 2000. Just who are you calling risk averse? Journal of Risk and Uncertainty
20: 177–187.
Issacharoff, S. 2010. Legal responses to conflicts of interests. In Conflicts of interest: Challenges
and solutions in business, law, medicine, and public policy, ed. Moore, D., Cain, D.,
Loewenstein, G., and M. Bazerman (dir.), 189–201. New York: Oxford university Press
Issacharoff, S. 2011. Disclosure, agents and consumer protection. Journal of Institutional and
Theoretical Economics 167: 56–70.
Johnson, Devon, and Kent Grayson. 2005. Cognitive and affective trust in service relationships.
Journal of Business Research 58: 500–507.
Jolls, C., C. Sunstein, and R. Thaler. 1998. A behavioral approach to law and economics. Stanford
Law Review 50: 1471–1550.
Judge, K. 2015. Intermediary influence, 82 University of Chicago Law Review 573 (Working paper
on intermediaries).
Kahneman, D., and A. Tversky. 1974. Judgment under uncertainty: Heuristics and biases. Science
185: 1124–1131.
Korobkin, R.B. 2003. Bounded rationality, standard form contracts and inconsolability. University
of Chicago Law Review 70(4): 1203–1295.
Korobkin, R.B., and T.S. Ulen. 2000. Law and behavioural science: Removing the rationality
assumption. University of Stanford Law Review: 115–177, California Law Review 88(4):
166 G. Helleringer

Kreps, D. 1990. A course in microeconomics theory. Princeton: Princeton university Press.

Lacko, J., and J. Pappalardo. 2004. Improving consumer mortgage disclosures: An empirical
assessment of current and prototype disclosure form’s, Federal Trade Commission Bureau of
Economics Staff Report.
Lipps, T. 1880. Die Aufgabe der Erkenntnistheorie und die Wundtische Logik (The task of episte-
mology and Wundtian logic). Philosophische Monatshefte 16: 529–539.
Locked, Karen. 1996. A funny thing happened! The management of consumer emotions in service
encounters. Journal of Organisational Sciences 7(1): 40–59.
Loewenstein, G., Sunstein, C., and R. Golman. 2013. Disclosure: Psychology changes everything.
Available at SSRN 2312708.
Löhnqvist, J.-E., Verkasalo, M., Walkowitz, G., and P. Wichardt. 2014 February. Measuring indi-
vidual risk attitudes in the lab: Task or ask? An empirical comparison, Kiel working paper
1095, Feb 2014.
Madden, G., R. Bethany, C. Lagorio, Andrea M. Begotka, A. Mueller, D. Hehli, and A. Wegener.
2004. Delay discounting of potentially real and hypothetical rewards: II. Between- and, within-
subject comparison. Institution, Experimental and Clinical Psychopharmacology 12(4):
Marotta-Wurgler, F. (2010). Does disclosure matter? NYU Working Paper n°10–54.
Mason, Gillian E., Stephanie Riger, and Linda Foley. 2004. The Impact of Past Sexual Experiences
on Attributions of Responsibility for Rape. Journal of Interpersonal Violence 19(10):
Mercer, M., A. Palmiter, and A. Taha. 2010. Worthless warnings? Testing the effectiveness of dis-
claimers in mutual fund advertisements. Journal of Empirical Legal Studies 7(3): 429–459.
Moore, D., D. Cain, G. Loewenstein, and M. Bazerman. 2010. Conflicts of interest: Challenges
and solutions in business, law, medicine, and public policy. Oxford: Oxford University Press.
Mussweiler, T., and F. Strack. 2000. The ‘relative self’: Informational and judgmental conse-
quences of comparative self-evaluation. Journal of Personality and Social Psychology 79(1):
Nisbett, R., and T. DeCamp Wilson. 1977. Telling more than we can know: Verbal reports on men-
tal processes. Psychological Review 84(3): 231–259.
Paxton, J., and J. Greene. 2010. Moral reasoning: Hints and allegations. Topics in Cognitive
Science 1: 1–17.
Payne, J., J. Bettman, and E. Johnson. 1993. The adaptative decision maker. Cambridge, MA:
Cambridge University Press.
Rempel, John K., John G. Holmes, and Mark P. Zanna. 1985. Trust in close relationships. Journal
Personnel Society Psychology 49(1): 95–112.
Ross, L., and R.E. Nisbett. 1991. Theperson and the situation. Perspectives of social psychology.
New York: McGraw Hill.
Sah, S., G. Loewenstein, and D. Cain. 2012. The burden of disclosure: Increased compliance with
distrusted advice. Journal of Personality and Social Psychology 104(2): 289–304.
Sah, S., Loewenstein, G., and D. Cain. 2013. Insinuation anxiety: Increased pressure to follow less
trusted advice after disclosure of a conflict of interest. Available at
Sani, F., M. Bennett, L. Agostini, L. Malucchi, and N. Ferguson. 2000. The impact of past sexual
experiences on attributions of responsibility for rape. The Journal of Social Psychology 140(2):
Scherif, M., and C. Hovland. 1961. Social judgment: Assimilation and contrast effects in commu-
nication and attitude change. New Haven: Yale University Press.
Schrah, G., R. Dalal, and J. Sniezek. 2006. The adaptative decision-maker is not an island:
Integrative expert advice with information search. Journal of Behavioral Decision-Making 19:
Sharot, T. 2011. The optimism bias. Current biology 21(23): 1951–2032.
Sniezek, Janet A., and Lyn M. Van Swol. 2001. Trust, confidence, and expertise in a judge–Advisor
system. Organizational Behavior and Human Decision Processes 84(2): 288–307.
9 Designing Disclosures 167

Summerfield, C., and K. Tsetsos. 2012 May. Building bridges between perceptual and economic
decision-making: Neural and computational mechanisms. Frontiers in Neuroscience 6: 1–20.
Thaler, R., and C. Sunstein. 2008. Nudge: Improving decisions about health, wealth, and happines.
New Haven: Yale University Press.
Von Winterfeldt, D., and W. Edwards. 1986. Decision analysis and behavioral research. New York:
Cambridge University Press.
Yaniv, I. 2004. Receiving other people’s advice: Influence and benefit. Organizational Behavior
and Human Decision Processes 93: 1–13.
Chapter 10
Neutral Third-Party Counselling as Nudge
Toward Safer Financial Products?
The Case of Risky Mortgage Loan Contracts

Piotr Tereszkiewicz

It is a terrible thing when, for example, unsophisticated people

are inveigled into an unaffordable mortgage on unconscionable
terms. But this happens (and the mortgage crisis occurred)
despite the magnificent edifice of disclosures that was
conscientiously debated when it was adopted, that has been
repeatedly revised and expanded, and that has been supervised
by an agency that has poured effort and expertise into it
(Ben-Shahar and Schneider 2013, p. 184).

Abstract This paper discusses the idea of neutral counselling as a measure of

improving decisions of borrowers when choosing a mortgage loan. The develop-
ments regarding foreign-currency mortgage loans, popular in Central and Eastern
Europe in recent years, illustrate how incorrect mortgage choices may have detri-
mental effects on borrowers and their families. Taking a sceptical approach toward
mandated pre-contractual disclosure as a protection measure for borrowers, the
paper advocates neutral counselling of mortgage applicants in case they are offered
risky types of mortgages. Drawing on the experience of an American mortgage
counselling programme, a set of recommendations is made regarding the design of
a possible neutral mortgage counselling scheme. While the counselling solution has
imminent limitations, it displays clear potential toward improving the process of
financial decision-making.

This research has been financed by the National Science Centre, Poland within the framework of
the research project: 2015/17/B/HS5/00495.
P. Tereszkiewicz (*)
Faculty of Law and Administration, Jagiellonian University of Cracow,
Olszewskiego 7, 31-007 Cracow, Poland

© Springer International Publishing Switzerland 2016 169

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_10
170 P. Tereszkiewicz

10.1 Introduction

Mortgage credit loans are a prime example of a financial contract concluded by

consumers. The decision to take out a mortgage loan is of paramount importance
from the perspective of borrowers, for mortgage loans are taken out usually once or
twice in a lifetime. When completing a mortgage loan transaction, the borrower
enters into a relationship with a lender, which is designed to last decades. The single
act of choosing an improper mortgage loan may have a dramatic impact on the life
of borrowers and their family.
A highly relevant recent example of risky mortgage contracts are foreign-
currency indexed mortgage loans, broadly marketed in a number of central and
eastern European countries, in particular in Poland and Hungary, a few years ago.
Around a decade ago, as interest rates were much lower in Switzerland than in
Poland and other central and Eastern European countries, the Swiss franc emerged
as an attractive currency for mortgage loans. Since 2000 about 770 000 mortgage
loans denominated (indexed) in Swiss francs have been concluded in Poland. In
2010 around 64 % mortgage loans were indexed in foreign currency (mostly CHF)
with the total value of 40 billion euro.1 Other Central and Eastern European coun-
tries, including Austria, Hungary, Croatia and Romania, witnessed similar develop-
ments. In Austria, 25 % of mortgage loans in 2010 were indexed in foreign currency,
with the total value of 30 billion euro.2 In Hungary, by the second half of 2008, the
household debt was close to 40 % of the country’s GDP, and two thirds of that debt
was denominated in foreign currency, where over 90 % of the household’s foreign
currency debt was denominated in Swiss francs.3 In 2014 the total value of foreign-
currency indexed mortgage loans in 2014 was 10 billion euro.4
Both in Poland and Hungary, the most intense accumulation of foreign-currency
indexed mortgage debt resulted between 2004 and 2008, which is before the recent
financial crisis. The sharpest marketing of foreign-currency mortgage loans in
Poland took place around 2006 and 2007. In order to promote foreign-currency
indexed mortgage loans, the financial industry framed the option for a mortgage

Rzeczpospolita from 26 January 2015,,1174464-
Kim-jestes-frankowcu-.html?referer=redpol; Rzeczpospolita from 5 February 2015, http://www.,1177051-Gdzie-mieszkaja-frankowicze.html?referer=redpol; The
Economist from 15 November 2014, Forint exchange,
See the official position of the Austrian Financial Market Supervision Authority (Österreichische
Finanzmarktaufsicht, FMA) on foreign currency loans: Position der Finanzmarktaufischt zu
Fremdwährungskrediten und Informationen zur derzeitigen Lage,
Balogh and others 2011, p. 1.
The Economist from 15 November 2014, Forint exchange,
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 171

loan of 300.000 PLN for the duration of 30 years as follows: On the one hand, a
monthly instalment payable if a mortgage loan was in Polish currency would be
2.130 PLN. On the other hand, a monthly instalment payable if a loan was indexed
in Swiss francs would be 1.409 PLN, given the currency exchange rate between the
Polish and Swiss currencies at that time. Apparently, a monthly instalment of a
CHF-indexed mortgage loan was around 30 % lower compared to a mortgage loan
in Polish currency.5 It is claimed that 75 % of examined clients considered the
foreign-currency indexed mortgage loan to be cheaper.6 The first 2 or 3 years of the
mortgage contract’s duration may have justified the decision to choose it, as the cur-
rency exchange rates remained beneficial for borrowers. From 2010 on, the situa-
tion began to change abruptly. The value of Swiss franc began to rise, from around
2.1 to 2.3 francs per zloty in 2006–2007, toward 3.5 or close to 4 francs per zloty in
2014.7 This dramatic rise of currency exchange rates had a huge impact on the
extent of the monthly credit instalments in foreign-currency indexed loans which
increased automatically. The increase of debt value around 30 % means that bor-
rowers have to repay amounts which clearly exceed the value of the property
financed with the foreign-currency indexed mortgage. This also means that selling
property will not generate enough funds to repay the foreign-currency indexed
loan.8 In general, unexpectedly higher instalments weaken debtors’ ability to pay
and significantly increase the probability of default.
Some countries have reacted to the drama of struggling borrowers with far-
reaching regulatory measures. For instance, Hungary passed the law allowing bor-
rowers to convert outstanding foreign-currency mortgage loans of globally about
$11 billion from franc and euro into forint, but generally at market rates.9 By con-
trast countries like Poland, where no regulatory strategy has been implemented until
now, are witnessing a growing number of law suits by financially struggling bor-
rowers against lenders claiming misrepresentation by lenders and unfair contract
With hindsight, it is justified to claim that the lending industry marketed the
foreign-currency indexed mortgage loans on the basis of a flawed analytical

Rzeczpospolita from 2 March 2015,
Rzeczpospolita from 2 March 2015,
The development of exchange rate of franc versus PLN is accessible at the website of the Polish
central bank (Narodowy Bank Polski),
Rzeczpospolita from 2 February 2015, reports on struggles of borrowers http://www.ekonomia.,1176120-Oferta-sprzedazy-mieszkan-zadluzonych-we-frankach.
The Economist from 15 November 2014, Forint exchange,
As this contribution was submitted, the Polish Parliament was working on a draft law allowing
borrowers to convert outstanding foreign-currency mortgage loans into the Polish currency, under
specific requirements.
172 P. Tereszkiewicz

model.11 It assumed a stable currency exchange rate, and downplayed interest

rates fluctuation risks in the future. The risk of radical changes of currency
exchange rates was transferred exclusively on to borrowers who were at the time
not familiar with that type of speculative loan product.12 The lending industry did
not introduce any institutional or contractual safeguards against a radical change
of currency exchange rates, but instead chose to expose its contractual partners
to this risk. It is justified to claim that the financial sector demonstrated a lack of
any appropriate risk management regarding mortgage indexation in foreign cur-
rency, which also contributed to the financial stability risk.13
By nature, foreign-currency indexed loans were complex legal products. Mostly,
foreign-currency mortgage loan contracts did not involve paying out loans in for-
eign currencies, mostly Swiss franc, but the foreign currency was used as an index
parameter to determine the amount of the borrower’s monthly instalment. Typically,
funds were advanced and repaid in a national currency of the borrower, while a
virtual account in Swiss francs was established to provide an index parameter.14
However, the amount of monthly instalments was determined by reference to the
exchange rate for a foreign currency, which served as a denomination index. In
truth, Polish lenders did not possess enough Swiss currency to be able to extend
credit to borrowers in Swiss francs.15
To illustrate the contractual design of foreign-currency indexed mortgage loans
it is worthwhile to use the example of contract clauses judged recently by the Court
of Justice of the European Union in the first two cases brought before it regarding
this type of mortgage contracts.16 A typical contract clause stipulated that “the
amount of the loan in foreign currency will be determined at the buying rate for the
foreign currency applied by the bank on the date of advance of the funds”.
Furthermore, “after the funds have been advanced, the amount of the loan, the
related interest, the administration fees and default interest and other charges will be
determined in the foreign currency”, mostly the Swiss franc. According to another

See the interview with the former Head of the Polish Financial Supervision Authority (Komisja
Nadzoru Finansowego), Stanisław Kluza, accessible at
The Economist from 15 November 2014, Forint exchange, quoting a bank official: “Now, with
the benefit of hindsight, we can see that it was a mistake,” says a senior official at an Austrian bank
that issued lots of franc loans, “Most of the clients did not recognise the risk and in some cases we
did not inform them properly.”
Balogh and others 2011, p. 4.
Rzeczpospolita from 3 February 2015 “Dlaczego frankowicz wygra w sądzie”, http://www.eko-, quoting some typical provisions of foreign-currency indexed
mortgage loan contracts.
In order to maintain stability of the Polish financial system, the Swiss Central Bank was forced
to provide liquidity to the Polish Central Bank in the beginning of 2009, cf. Schweizerische
Nationalbank, Geschäftsbericht 2009, available at
Judgment of the Court of Justice of the European Union of 30 April 2014, Case C-26/13, Kásler
vs. OTP Jelzálogbank Zrt, ECLI EU C 2014, p. 282.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 173

typical clause “the lender is to determine the amount in [national currency of the
borrower] of each of the monthly instalments due by reference to the selling rate of
exchange for the foreign currency applied by the bank on the day before the due
date”.17 Such clauses granted lenders effectively discretion in unilaterally determin-
ing the amount of each of the monthly instalments, as the bank’s right to determine
the selling rate of exchange for the foreign currency did not involve any reference
to official exchanges rates of central banks. Furthermore, frequent were also clauses
allowing lenders to “alter the current rate of interest in the event of significant
changes on the financial markets”. This in fact allowed lenders to change the inter-
est rate in a discretionary manner.18
The complex and intransparent design of foreign-currency indexed mortgage
loans was clearly beneficial for lenders. It justified a proliferation of special fees and
charges, e.g. a risk charge, for making the credit available.19 The review of these
clauses by the Court of Justice of the European Union revealed that lenders did not
provide any service constituting consideration, which would justify these charges.
The rise and subsequent fall of foreign-currency indexed mortgage loans in
Poland, Hungary and other European countries provides justification for an analysis
on how to drive potential borrowers toward choosing safe mortgage loan contracts.
Interestingly, at least in Poland, most consumers concluding foreign-currency
indexed mortgage contracts were people between 35 and 40 years old, quite often
university graduates, with above-average income.20 As a social class, they revealed
little similarity to borrowers who were targeted with subprime lending in the
U.S. Yet they equally got committed to mortgage loans that they would most likely
not have concluded again with the benefit of hindsight. Undoubtedly, foreign-
currency indexed mortgage loans are just one of a number of risky types of mort-
gages which were broadly marketed both in the U.S. and Europe in recent years.
Other examples include interest-only loans, variable interest rate loans, loans allow-
ing negative amortization of the debt or charging pre-payment penalties.21 This
paper undertakes to advocate neutral third-party counselling of prospective borrow-
ers as a strategy toward improving financial decision-making. As a protection
means, neutral counselling is discussed as an approach different to mandated disclo-
sure. By contrast, this paper does not discuss the related issue of liability of banks

Para 22 of the Judgment, Kásler vs. OTP Jelzálogbank Zrt.
Judgment of the Court of Justice of the European Union of 26 February 2015, Case C-143/13,
Matei vs. SC Volksbank Romania, OJ C 138, 27.4.2015, par. 26.
This point is well illustrated in the Matei vs. SC Volksbank Romania case. When borrowers
questioned the fairness of the ‘risk charge’ due under the mortgage credit contract, the lender just
proposed to re-name the ‘risk charge’ to ‘Credit Management Charge’, since charging this com-
mission was expressly authorised by applicable regulations, however, without at the same time
modifying the content of terms defining the charge. Similar features in the U.S. mortgage lending
practice are described by Bar-Gill 2012, pp. 120–123.
Rzeczpospolita from 5 February 2015,,1177051-
On definitions of predatory loans see Willis 2006, pp. 718–729, Engel and McCoy 2002, p. 1284,
Stark and Choplin 2010, pp. 9–18.
174 P. Tereszkiewicz

for marketing and selling risky mortgage loans, as this is a question which merits a
separate study.22

10.2 Why Do Borrowers Choose Risky and Intransparent


There are a number of heterogeneous reasons why borrowers choose risky or

intransparent mortgage loans. For example, Stark and Choplin listed 14 cognitive
and social psychological phenomena (‘psychological barriers’) that significantly
limit the effectiveness of disclosures as the protection instrument to ensure a prop-
erly functioning market. It appears that their account mentions most of key reasons
for sub-optimal borrower choice. According to Stark and Choplin, the 14 barriers
are: (1) Inability to process user-unfriendly features of disclosure forms; (2) Lack of
contractual schemas or knowledge structures; (3) Inaccurate default assumptions of
how contractual provisions are likely to be structured and if the contract is nego-
tiable; (4) Availability heuristics; (5) Reason-based decision making; (6) Biases in
attribute estimation and evaluation; (7) Positive conformation biases; (8) Accepting
senseless explanations; (9) Argument immunization; (10) Sunk costs effects; (11)
Endowment effects; (12) Temporal and uncertainty discounting; (13) Strong moti-
vation to trust and misplaced trust in mortgage brokers; (14) Social norms and
It is impossible to discuss all of the above barriers in the context of foreign-
currency indexed loans, as they were intensely marketed a few years ago in Central
and Eastern Europe. Nevertheless, a number of factors that Stark and Choplin men-
tion, undoubtedly played a role in the decision-making process of borrowers.

10.2.1 Lack of Knowledge Structures and Availability


To begin with, borrowers may not grasp the risk (speculative nature) put into the
clauses of a mortgage contract, be it an adjustable credit rate or denomination of a
loan in a foreign currency. For instance, the U.S. experience shows that the majority
of the borrowers that were presented with a floating rate loan were surprised to find
out from a third-party counsellor that the loan they were about to conclude was an

For a comprehensive discussion of lenders liability for violating duties to provide information or
advice see Tereszkiewicz 2015.
Stark and Choplin 2010, p. 71, Table 1: Fourteen cognitive and social psychological factors that
cause disclosure forms to be ineffective.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 175

adjustable rate loan.24 Consumers usually assess the risk associated with contract
provisions by the ease with which they can imagine possible consequences of those
provisions.25 However, due to their limited experience consumers are frequently
unable to foresee harmful outcomes that are quite probable. One could legitimately
claim that during the boom of foreign-currency indexed loans around 2006 and
2007 in Poland and other Central European countries, consumers generally failed to
appreciate to what extent such loans could lead to unforeseeable risks. At the same
time, banks did not fulfil their role of professional risk managers, shifting all cur-
rency exchange risks on to borrowers under the contractual framework of mortgage
contracts. At present, following a broad coverage of over-indebtedness in the media,
the risks resulting from foreign-currency indexed loans have become better-known
and available to average consumers.

10.2.2 Reason-Based Decision-Making

Furthermore, it is claimed that people simplify the choice by looking for a reason to
favour one option over another instead of weighing all advantages and disadvan-
tages of an option.26 This decision-making strategy evidently played a role in the
choice of foreign-currency indexed loans. Many borrowers, prompted by intense
marketing, seized on one salient feature such as low initial monthly payment, ignor-
ing the risk of floating currency rate.27 This tendency was aggravated by the framing
of offers that lenders commonly used, as mentioned above – an apparently evident
choice between a ‘cheaper’ foreign-currency indexed loan and a more expensive
mortgage loan in a local currency.

10.2.3 Endowment Effects

Equally relevant are endowment effects, when borrowers form an illusion, begin-
ning to feel as if they already owned the house that they would like to purchase.28
Should the transaction never materialize, consumers will regard this as a loss. It
follows that consumers are willing to accept riskier options to avoid such a ‘loss’
than they would accept to acquire a gain.29 Providing information or warnings at a
late stage is likely to have a weaker effect than if such communications had been
made early on.

Housing Action Illinois, Findings from the GB 4050 Predatory Lending Database Pilot Program
2007, quoted after Stark and Choplin 2010, p. 20.
Stark and Choplin 2010, p. 22, drawing on Tversky and Kahneman 1980, pp. 207–232.
Shafir et al. 1993, pp. 11–36.
Willis 2006, pp. 780–781.
Stark and Choplin 2010, p. 26.
See Kahneman and Tversky 1979, pp. 263–291.
176 P. Tereszkiewicz

10.2.4 Temporal and Uncertainty Discounting

Perhaps the most important human frailty is the tendency to fail to appreciate the
likelihood of uncertain, but probable costs. As the future seems distant, negative
outcomes are delayed. It is claimed that the degree to which people discount future
expenses usually exceeds the amount that can be rationally justified.30 An example
of Homer Simpson, used by Thaler and Sunstein, provides a very good illustration
of this phenomenon.31 Lenders can take advantage of this psychological inclination
by downplaying the extent of future risks or offering initial teaser rates. This ‘worst-
case’ scenario seems to have materialised in the case of foreign-currency indexed
loans. The currency exchange rates remained relatively attractive for borrowers dur-
ing the two or three first years of mortgage loan duration. Subsequently, however,
currency rates have risen dramatically, fundamentally changing the extent of bor-
rowers’ liabilities. Only then did the full extent of risks resulting from foreign-
currency indexed mortgage loans become apparent to most borrowers.

10.2.5 The Impact of a Confidence Relationship with a Lender

or a Broker

The history of many over-indebted consumers powerfully illustrates the role that
personal trust placed in a lender or a broker has played.32 Lenders and brokers
eagerly assume the image of consumer’s ‘advisors’ and ‘helpers’ to create an
impression that they offer a particularly good, individually tailored deal to a con-
sumer.33 Most mortgage loan contracts are concluded in one-on-one settings, where
consumers place too much trust in salespeople who usually call themselves ‘advi-
sors’. Furthermore, many borrowers mistakenly believe that lenders or brokers have
an obligation to obtain the best loan and interest rate for them. Mortgage selling
practices, which occurred both in the U.S. and in some European states, could prob-
ably be described with ‘direction and deflection’ model.34 The borrower’s attention
was directed, for instance, on the currently favourable indexation rate of Swiss
francs versus Polish zloty, but at the same time deflected from the fact that the
indexation rate may abruptly rise and extend the debt value. Furthermore, common
was creating an impression that the indexed loan proposed was the best and safest
deal for the borrower. The American research claims that many clients believed that
their loan originator (lender) was under an obligation to obtain the best loan for

Stark and Choplin 2010, p. 27, drawing on Frederick, Loewenstein and O’Donoghue 2002,
pp. 351–401.
Thaler and Sunstein 2009, p. 140.
See Willis 2006, pp. 798–806.
Regarding a similar pattern in ‘self-presentation’ of insurers see Feinman 2009, p. 553.
Stark and Choplin 2010, p. 17.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 177

them.35 There is no reason to believe that the borrowers’ perception in Europe was
significantly different.

10.3 Possible Legal Measures Against Suboptimal Choices

10.3.1 Financial Education

There is a broad consensus that consumer financial literacy regarding complex and
risky financial contracts, such as mortgages, is inadequate.36 Yet it remains unclear
whether financial education programmes can be an effective means of addressing
this concern. It is submitted here that one should not rely on financial education as
preventive means of improving decision-making regarding complex financial prod-
ucts. As Willis persuasively argues, financial education would have to be extensive,
frequent and personalized to raise the probability of correct consumer decisions.37
Furthermore, given the reality of financial markets, one could be sceptical as to
whether financial education would really enhance the autonomy of consumers.38
The European experience with foreign-currency loans supports the sceptical
approach towards financial education. The data shows that typical lenders were
people with medium or high education.39 Furthermore, whatever educational mea-
sures are applied, they may not manage to keep up with the pace with which risky
financial innovations become accessible to financial consumers in the marketplace.

10.3.2 ‘Smart’ Disclosure of Contract Terms

The most intensely applied legal measure of our times is mandated disclosure.40 The
targeted use of ‘smart’ disclosure is advocated also in the context of mortgage loan
credits as a measure of improving consumer choices. Sunstein and Thaler describe
it as “the timely release of complex information and data in standardized, machine
readable formats in ways that enable consumers to make informed decisions”.41 In
the simplified form, mortgage lenders would be required to report lending costs in

Stark and Choplin 2010, p. 17.
Lusardi 2008; Meier and Sprenger 2013, pp. 159–174.
Willis 2011, pp. 431–432.
Willis 2011, pp. 431–432, claims that financial education would have “paradoxical effects on
Rzeczpospolita from 2 March 2015,
This topic attracted important literature in recent years. A major critique of mandated disclosure
stems from Ben-Shahar and Schneider 2013. For a more positive approach toward disclosure see
A. Schwartz 2015, p. 1374. See also Bar-Gill 2012, pp. 32–43.
Sunstein and Thaler 2009, p. 146.
178 P. Tereszkiewicz

two categories: fees and interest. The second category would be a machine-readable
detailed RECAP report, incorporating all the fees and interest rate provisions,
including teaser rates, what the variable-rate changes are linked to, caps on the
changes per year, and so forth.42 It is essential that this information could allow
independent third parties to offer much better advice. Sunstein and Thaler even
assume that third-party services would emerge to compare lenders. Undoubtedly,
this proposal goes some way to counter the inability to process user-unfriendly fea-
tures of disclosure forms. But, as it seems, Sunstein and Thaler are not convinced
that even ‘smart disclosures’ will be effective in the case of average, unsophisticated
consumers. It is essential to take into account that the need for simple language
sometimes conflicts with the need for legal precision, while the need for concise-
ness of disclosures may conflict with the need to provide information that may be
relevant for different categories of consumers.43 These considerations appear to
reinforce the case for a neutral third-party advisor who could perform the function
of translating the technical disclosure to the consumer.

Disclosure Regarding Mortgage Contracts Under the European Union

Mortgage Credit Directive

A brief survey of a recent piece of European legislation on mortgage credit reveals

what role disclosure can fulfil with respect to risky mortgage loans. At the European
level the Mortgage Credit Directive44 (abbreviated as MCD) was enacted in 2014
with a view to providing a high level of consumer protection in respect of loan
mortgage credits.45 It introduces a European framework of conduct rules for firms
selling mortgages. The Mortgage Credit Directive obliges lenders and intermediar-
ies to communicate to consumers a broad catalogue of information regarding the
proposed transaction (Art. 13–16 MCD). Pre-contractual information ought to be
provided in a standardised format, using the European Standardised Information
Sheet (ESIS).46 However, in case of more complex mortgage loan, such as foreign-
currency indexed mortgage loans or variable rate mortgage loans, the total amount
payable by the borrower cannot be communicated in advance. Furthermore, the
Mortgage Credit Directive mandates that where there is no provision in the credit

Sunstein and Thaler 2009, pp. 146–147.
Craswell 2006, p. 576.
Directive 2014/17/EU of the European Parliament and of the Council of 4 February 2014 on
credit agreements for consumers relating to residential immovable property and amending
Directives 2008/48/EC and 2013/36/EU and Regulation (EU) No 1093/2010, OJ L 2014, L 60,
p. 34.
The academic literature on the Mortgage Credit Directive is still rather limited. An interesting
introduction is provided by the British Financial Conduct Authority in its Consultation Paper
“Implementation of the Mortgage Credit Directive and the new regime for second charge mort-
gages”, CP 14/20, September 2014, available at
Consideration 7 of MCD.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 179

agreement to limit the exchange rate risk to which the consumer is exposed to a
fluctuation in the exchange rate of less than 20 %, the ESIS shall include an illustra-
tive example of the impact of a 20 % fluctuation in the exchange rate (Art. 23 (6)
It is worthwhile to quote the provisions of the European Standardized Information
Sheet (ESIS) regarding the scope of information in case of foreign-currency loans:
3. Main features of the loan:
Amount and currency of the loan to be granted: [value] [currency]
(Where applicable) This loan is not in [national currency of the borrower].
(Where applicable) The value of your loan in [national currency of the borrower] could
(Where applicable) For example, if the value of [national currency of the borrower] fell
by 20 % relative to [credit currency], the value of your loan would increase to [insert
amount in national currency of the borrower]. However, it could be more than this if the
value of [national currency of the borrower] falls by more than 20 %.
(Where applicable) The maximum value of your loan will be [insert amount in national
currency of the borrower]. (Where applicable) You will receive a warning if the credit
amount reaches [insert amount in national currency of the borrower]. (Where applicable)
You will have the opportunity to [insert right to renegotiate foreign-currency loan or right
to convert loan into [relevant currency] and conditions].
6. Amount of each instalment
(Where applicable) The value of the amount you have to pay in [national currency of the
borrower] each [frequency of instalment] could change. (Where applicable) Your payments
could increase to [insert maximum amount in national currency of the borrower] each
[insert period]. (Where applicable) For example, if the value of [national currency of the
borrower] fell by 20 % relative to [credit currency] you would have to pay an extra [insert
amount in national currency of the borrower] each [insert period]. Your payments could
increase by more than this.


It could be held that the disclosure relating to foreign-currency risks under the
Mortgage Credit Directive, as designed above, is of little use. It does not convey a
precise message on the extent of borrower’s liabilities. Thus it cannot fulfil the
requirements of ‘the smart disclosure’ concept advocated by Sunstein and Thaler.
The amount of the mandatory text to be communicated is large, but the major mes-
sage it conveys is that the precise extent of borrower’s liability may depend on a
number of factors and remains unspecified. The above disclosure is of little use to
an average borrower, if it is not accompanied by further measures, for instance, a
neutral explanation. Furthermore, it even carries the danger of being harmful. As
Willis rightly points out, extensive disclosures give the ‘veneer of legality and
authority’ to the loan process, even if it involves highly risky products.47 Pre-contract

Willis 2006, p. 794.
180 P. Tereszkiewicz

material or even lender’s explanations show a limited potential of overcoming

informational failures of a borrower. Furthermore, the warnings and explanations by
a lender usually do not provide the borrower with a sense of his or her alternatives.48
In this context it is doubtful whether the mandatory information on the potential
impact of exchange rate fluctuation, required by the Mortgage Credit Directive, will
serve as an important consideration for a borrower. There is an imminent risk that
this standardised warning will be crowded out by the extensive catalogue of other
mandatory disclosures.49 Moreover, the pre-contract communications by a lender,
usually focusing on a specific product, are not likely to encourage a critical and
detached reflection on the part of the borrower which could help him or her see the
unusual dangers of a ‘risky’ type of mortgage contract as compared to standard
ones. It follows that pre-contract disclosure and explanation cannot be regarded as
a suitable tool of enhancing the autonomy of borrowers. Besides, lenders as sellers
of mortgage products are typically in a conflict of interest position, for they derive
profits from mortgage contracts concluded with their clients. Therefore they cannot
be expected to provide an impartial assessment of mortgage products they expect to
earn profits from. On the whole, it is justified to assume that disclosures and expla-
nations by lenders have limited efficacy in correcting most important informational

10.3.3 Mandatory Neutral Advising

Some of leading scholars of consumer law contend that certain risky types of mort-
gages, which would include foreign-currency indexed mortgage loans, should be
banned.51 Nevertheless, if a legislator does not follow that position, as is the case at
the European Union level under the Mortgage Credit Directive, measures of protec-
tion other than disclosure should be contemplated. Neutral (independent) counsel-
ling can be considered to be one of such measures. In some common law countries,
like England or Australia, independent advice is regarded as a major device the law
uses to correct contract failure in family financial dealings, in particular in non-
professional surety contracts.52 Lenders are obliged to insist that non-professional
sureties (guarantors) be independently advised by lawyers on the transaction that
they intend to conclude with a bank.53 This paper proposes to apply this idea to
mortgage loan contracts.

Rightly Trebilcock and Elliot 2001, p. 75.
On this danger Craswell 2006, pp. 576–580.
Trebilcock and Elliot 2001, p. 75.
Willis 2006, p. 823.
Trebilcock and Elliot 2001, p. 65; see also Tereszkiewicz 2011, pp. 851–873.
Beatson et al. 2010, p. 361.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 181

Apparently, advice as a form of informational behaviour resembles disclosure.54

Nevertheless, it actually offers a different kind of help, aiming to answer the ques-
tion: “How a given measure is likely to meet your objective?”. A neutral advisor is
not an agent for the discloser. Rather, it is someone whose function is to substitute
the disclosure and assist in the decision-making process.
The idea of neutral (independent)55 advice is promising in many respects. It
appears that neutral (independent) advice may contribute to reducing the conse-
quences of both cognitive incapacities and temporary distorting states, avoiding the
shortcomings of protection by means of disclosure. First, a neutral advisor (counsel-
lor) is in a better position to inform and advise a borrower about the risks and costs
of an offered mortgage than a self-interested lender (seller). As a result the borrower
should be better informed about the content and the risks of the transaction to be
concluded and also less susceptible to being misled by a lender. As opposed to
mandated extensive disclosure, neutral counsellor is likely to use simple, non-
technical language to describe complex issues. In contrast to disclosure and expla-
nation by a lender, neutral advice offers a better chance of altering the structure of
a financial choice by a borrower, including re-thinking the decisions about taking a
mortgage. Neutral counsel may also inform the borrower about alternative mort-
gage products that are accessible to him or her in the marketplace, making specific
recommendations that simplify the complex process of comparing often incompa-
rable mortgage products of different lenders in the market place.
Furthermore, neutral advice has the potential of correcting emotional incapaci-
ties which may be temporary, some of which were mentioned before, like emotional
pressure from a lender, endowment effects, strong desire to urgently secure funding
for a specific property.56 Generally, neutral advice is much better suited to enhance
borrower’s internal autonomy, stimulating him or her to reflect on the reasons for
choosing a specific risky mortgage offer and to what extent this mortgage form
would meet her goal. It is common wisdom that a person lacking material informa-
tion or being under a major misapprehension regarding her future options may
make a choice against her actual preferences or needs. As John Stuart Mill wrote,
preventing a person from crossing a bridge that she mistakenly believes to be safe
is not a real interference with her autonomy.57 An intervention in the form of coun-
selling may lead to outcomes that a party would have chosen if he or she had not
been ignorant of material circumstances or unduly pressured to make a decision.58
One should agree with Trebilcock and Elliot, that even if mandatory counselling is

I use the term ‘informational behaviour’ drawing on the illuminating analysis of Craswell 2006,
p. 566.
Terms ‘neutral’ and ‘independent’ are used here largely as synonyms, as are terms ‘advice’ and
Trebilcock and Elliot 2001, p. 76, Stark and Choplin 2010, pp. 26–27.
Mill 1947, p. 89.
Trebilcock and Elliot 2001, p. 69.
182 P. Tereszkiewicz

regarded as a paternalistic measure, it is in such circumstances respectful of agent’s

true preferences.59
Moreover, recent criticism of disclosure highlights yet another important consid-
eration concerning neutral counselling. Ben-Shahar and Schneider stress that peo-
ple commonly prefer advice to data when faced with making unfamiliar and
complex decisions.60 Choosing a right mortgage loan contract is undoubtedly one of
the most complex financial choices to be made during a lifetime. In light of the solu-
tion advocated here neutral counsellors can be regarded primarily as professional
information intermediaries who largely substitute the disclosure and translate its
content to borrowers.

The Concept of Advising (Counselling) Under the Mortgage Credit


Regulatory Approach

The above-mentioned Mortgage Credit Directive dates from 2014 and thus could
have been an opportunity to introduce an innovative counselling scheme, as a regu-
latory option for Member States. This is why a brief overview of the Directive’s
approach to regulating borrower advice is justified. In an attempt to improve the
protection allowed by lenders’ mandatory disclosures to borrowers, the Mortgage
Credit Directive introduces provisions regarding advising (counselling) borrowers
with a view to a prospective mortgage loan contract. The Mortgage Credit Directive
adopts the position that providing advice in the form of a personalized recommen-
dation to a consumer is a distinct activity, which may but need not be combined
with granting or intermediating credit.61 Therefore, creditors and intermediaries
should explicitly inform consumers whether advisory services are or can be pro-
vided to the consumer in the context of a given transaction (Art. 22 (1) MCD).
Furthermore, specific rules are designed to make consumers aware of what range of
products lenders and intermediaries take into account in their recommendations.
Lenders and intermediaries are obliged to inform consumers whether the recom-
mendation will be based on considering only their own product range or a wide
range of products from across the market (Art. 22 (2) a MCD).
It is questionable whether this information duty goes far enough toward warning
a consumer that lenders may offer him or her a limited range of products. With
respect to abstract products such as mortgage contracts, most consumers will have

Trebilcock and Elliot 2001, p. 69.
Ben-Shahar and Schneider 2013, p. 13, p. 190, proposing a broad concept of ‘information inter-
mediaries’, comprising consultants, aggregators, and even savvy shoppers.
Consideration 63 of MCD.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 183

difficulties distinguishing between ‘untied’ and ‘tied’ intermediaries.62 In an attempt

to provide for more specific measures of protection at the national level, the
Mortgage Credit Directive provides Member States with two regulatory options.
First, they may prohibit the use of the term ‘advice’ and ‘advisor’ or similar terms
when the advisory services are being provided to consumers by creditors, tied credit
intermediaries or appointed representatives of tied credit intermediaries (Art. 22 (4)
MCD). Alternatively, where Member States do not prohibit the use of the term
‘advice’ and ‘advisor’, they shall impose the following conditions on the use of the
term ‘independent advice’ or ‘independent advisor’ by creditors, credit intermediar-
ies or appointed representatives providing advisory services:
(a) Creditors, credit intermediaries or appointed representatives shall consider a
sufficiently large number of credit agreements available on the market; and
(b) Creditors, credit intermediaries or appointed representatives shall not be remu-
nerated for those advisory services by one or more creditors (Art. 22 (4) MCD).
Furthermore, the Mortgage Credit Directive provides an option clause, whereby
Member States may impose more requirements in relation to the use of the terms
‘independent advice’ or ‘independent advisor’ by creditors, credit intermediaries or
appointed representatives, including a ban on receiving remuneration from a credi-
tor (Art. 22 (5) MCD).


Undoubtedly, the Mortgage Credit Directive provisions on advice go beyond the

model of ‘advised sales’, where a seller (lender) is providing advice, but only on the
basis of his or her own mortgage products.63 Theoretically, the Mortgage Credit
Directive could contribute to creating a market model, where lenders will advise
their clients, taking into account a sufficiently large number of mortgage products
available on the market. Imposing a duty on lenders to take into account a market-
representative catalogue of products while advising on products can also provide
incentives for lenders to optimize their own products from the perspective of con-
sumer needs.
On the other hand, it could be claimed that the Mortgage Credit Directive does
not go far enough toward establishing a model of independent advice for consum-
ers. There remains an imminent risk that self-interest of lenders will play a crucial
role in the recommendations that they provide to consumers. It is thoroughly

Cf. Financial Services Authority (UK), Consultation Paper 11/31, Mortgage Market Review:
Proposed package of reforms, par. 1.109: “Most customers believe that if they speak to an interme-
diary, they have been given ‘advice’, no matter how many times they may be told that they are not
being given advice and whatever form of service disclosure they are given confirming this posi-
tion”; available at
On the concept of ‘advice’ under Directive 2004/39/EC on markets in financial instruments
(MiFID I) see Moloney 2010, p. 200.
184 P. Tereszkiewicz

possible to take into account a sufficiently large number of mortgage agreements

available on the market, but at the same time still present one’s own offer as particu-
larly advantageous and ‘unique’ for a consumer. Given the need to protect unso-
phisticated consumer borrowers, the measures aimed at taming self-interest of
lenders appear insufficient. Advice provision under the concept of the Mortgage
Credit Directive cannot be regarded as genuinely ‘neutral’.
Assessing the model of the Mortgage Credit Directive, it is necessary to bear in
mind the inevitable political constraints of the European legislation. More elabo-
rated mandatory counselling schemes for consumers cannot be realistically intro-
duced at a European level, but their creation could be encouraged. A pan-European
uniform solution in this respect does not currently appear feasible. Introducing and
developing legal rules on counselling schemes for mortgage borrowers remains a
legal competence of, and a political challenge for every single Member State in the
process of providing high level of mortgage borrower protection.

10.4 Neutral Counselling as a Mandatory Institution

It seems that the model of neutral counselling should be mandatory in order to have
any meaningful effect. A softer measure would be to require lenders to recommend
prospective borrowers that they should seek independent advice before concluding
a mortgage transaction.64 A major risk of such a solution, as was showed by Burrows,
is that it might provide an incentive to lenders to make recommendations in a way
that minimises the likelihood of the independent advice being taken.65 This in turn
encourages active persuasion on the part of lenders that borrowers should take up
their offers. A requirement that counselling should be mandatory eliminates that
risk and forces lenders to cooperate with advisors. Undoubtedly, this is a more
paternalistic measure than an optional counselling scheme. Nevertheless, if pre-
contract neutral counselling is supposed to provide a substitute for mandated disclo-
sure, its mandatory character appears justified.

10.4.1 Illinois Predatory Lending Database Pilot Program

as an Illustration
The Scope and Aims of the Illinois Pilot Program

A notable example of a mortgage counselling programme is a pilot programme run

in Cook County, Illinois, between September 2006 and January 2007. Originally,
the sponsors of the bill introducing the pilot programme wanted the law to apply

This position was part of English law of private sureties, see Trebilcock and Elliot 2001, p. 75.
Burrows 1995, p. 140.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 185

state-wide, but were not able to get it passed by the state legislature.66 Consequently,
the legislation was reduced to a pilot programme, with the target area chosen on the
basis of higher foreclosure rates and subprime lending.67 Essentially, the programme
required ‘high-risk’ mortgage applicants acquiring or refinancing properties in ten
Chicago zip codes to submit loan offers from state-licensed lenders to review by
loan counsellors certified by the Department of Housing and Urban Development
(HUD).68 This requirement also applied to applicants choosing certain mortgage
products deemed risky by legislator. More specifically, the pilot programme man-
dated financial counselling for mortgage credit applicants whose credit scores were
low enough to regard them as high-risk borrowers (the low-FICO borrowers).69 The
mid-FICO borrowers could be subject to counselling in case they chose specific
categories of high risk mortgage products, such as interest-only loans, loans with
interest rate adjustments within 3 years, loans underwritten on the basis of stated
income and repeat re-financings within the last 12 months. Furthermore, counsel-
ling was obligatory regardless of the consumer’s creditworthiness (FICO) score in
case they intended to take out loans with prepayment penalties, loans allowing neg-
ative amortization, or loans with closing costs in excess of 5 % of the loan.70
The programme rules mandated that borrowers falling under its scope attend a
counselling session with a certified loan counselling agency. Technically, brokers
or lenders were obliged to submit information on the prospective loan to the Illinois
Department of Professional and Financial Regulation (IDPFR) within 10 days of
the offer, following which eligible borrowers received notice of the requirement for
counselling from the IDPFR.71 A central database with all mortgage offers was held
by the IDPFR. Borrowers were required to contact the counselling agency of their
choice and attend the 1 h counselling session.72 The aim of such sessions was to
discuss the terms of the specific loan offer for a home purchase or refinancing and
to explain their meaning and consequences to the prospective borrower. The coun-
sellors were essentially supposed to warn borrowers against most frequent risks.
Concluding a session, the counsellor was obliged to record a number of key findings
about the loan, in particular whether the fees or interest rates were excessive,
whether the consumer understood the contract or whether he could afford the loan.
As the report’s authors emphasize, the interview with the borrower and an inde-
pendent collection of data on his or her financial situation enabled counsellors to

Bates and Van Zandt 2007, p. 39.
“Pilot program area means areas designated as such by Department due to the high rate of fore-
closure on residential home mortgages that is primarily the result of predatory lending practices.
The Department shall designate the pilot program area within 30 days after the effective date of
this amendatory Act of the 94th General Assembly.” See Bates and Van Zandt 2007, p. 18.
Agarwal and others 2009, p. 2.
Agarwal and others 2009, p. 2.
Bates and Van Zandt 2007, p. 16. This led to the high-FICO borrowers being counselled condi-
tional on their mortgage contract choice.
All mortgage loan applications were entered into a central database.
Bates and Van Zandt 2007, p. 23.
186 P. Tereszkiewicz

form an assessment of borrower creditworthiness that exceeded the findings of

lenders. In this manner, the counsellors relied on private information from borrow-
ers which were often not used by lenders in the decisions about granting loans. Such
information also became a matter of public record, as counsellors entered their rec-
ommendations for borrowers in the state-maintained database.73
At the end of the counselling session counsellors made possible recommenda-
tions as to whether the loan should be approved or not and for what reason.
It is worthwhile to list the possible recommendations, as they point to the most
typical – in the legislator’s view – reasons for suboptimal financial decisions:
Possible recommendations included:
1. Loan should not be approved due to indication of fraud;
2. Loan should be approved, no problems noted;
3. Borrower cannot afford loan;
4. Borrower does not understand transaction;
5. Borrower does not understand costs;
6. Borrowers income and expenses have been reviewed;
7. Loan rate is above market;
8. Borrower should seek competitive bid;
9. There are discrepancies between borrower’s understanding and originator’s
submitted form;
10. Borrower is precipitously close to not being able to afford the loan;
11. Borrower understands the true cost of debt consolidation and need for credit
card discipline;
12. Information provided by borrower has been amended by originator.74
The scope of possible recommendations shows that a counsellor assesses a num-
ber of factors which could be generalized as follows: (a) the lender’s offer and
conduct toward the borrower; (b) the borrower’s understanding of transaction’s
legal nature and financial implications; (c) the borrower’s creditworthiness, and (d)
the fairness of loan rates.
Counsellor’s recommendation were not binding on the borrower who could
decide to take up the loan offer obtained despite a recommendation to the contrary.
Furthermore, the Illinois Pilot Programme regulations did not assign liability to
either the lender or the counsellor.

Agarwal and others 2009, p. 8.
See Bates and Van Zandt 2007, p. 39, Figure 1, on the basis of 38 Illinois Administrative Code
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 187

Major Findings

The major finding from the Illinois Pilot Program is that the counselling scheme
resulted in substantially lower ex post default rates and somewhat better loan
choices among some of the counselled borrowers that remained in the market.75
Specifically, the 18-month default rate among borrowers with low creditworthiness
scores was about 4.5 % lower than that among similar borrowers in the control
group. This decline could be seen as very serious.
Furthermore, stronger evidence for indirect effects of the counselling require-
ment on mortgage origination and mortgage decision making was found. First, the
study documents a spike in rejection rates of mortgage applications by lenders who
are subject to the legislation during the treatment period, with rejection rates return-
ing to their normal level as soon as the Illinois Pilot Program is rescinded.76 A pos-
sible explanation of this change is the temporary exit of lenders with loose screening
practices from the area subject to the Illinois Pilot Program, coupled with tighter
screening by lenders remaining in the market. Secondly, the study finds a fairly
large decline in the amount of low-documentation mortgages.77 This effect can be
attributed to demands by counsellors that prospective borrowers should bring their
income documentation to the counselling session. Both of the results point toward
the oversight effect - the conclusion that third-party counselling, involving a collec-
tion and recording of lender information, results in more thorough screening.78
Thirdly, it is reported that mid and high creditworthiness score borrowers who
could avoid the mandatory counselling programme by choosing less risky mortgage
products, in fact did so.79 A possible reason for this choice was to minimise interac-
tion with counsellors. Fourthly, the study reports that borrowers who had been
counselled rejected fewer mortgage offers.80 Fifthly, it was found that the financial
counselling requirement improved ex post default rates for the low creditworthiness
counselled borrowers relative to similar borrowers outside the Illinois Pilot Program
area -the effect of the Illinois Pilot Program on default is said to be “impressive in
its economic magnitude”.81
Most importantly, the study’s authors emphasise that the threat of oversight of
lending practices and the imposition of transaction and compliance costs of coun-
selling, rather than the information contained in counselling sessions, were the pri-
mary catalyst for improvement in the decision making process of borrowers, leading
to lower default rates.82 It is claimed that the mere presence of the regulator in the
marketplace and the third-party review of mortgage contracts appear to have a large

Agarwal and others 2009, p. 3.
Agarwal and others 2009, p. 4.
Agarwal and others 2009, p. 4.
Agarwal and others 2009, p. 4.
Agarwal and others 2009, p. 31.
Agarwal and others 2009, p. 31.
Agarwal and others 2009, p. 23.
Agarwal and others 2009, pp. 4–5.
188 P. Tereszkiewicz

effect on the quality of mortgage contract. On the other hand, it must be acknowl-
edged that counselling is perceived as a burden by borrowers who either stay away
from the market or choose products that do not trigger counselling.


The findings from the Illinois Predatory Lending Database Pilot Program are inspir-
ing, as they support the proposition that neutral counselling schemes may contribute
to better financial decision-making by consumers. Given a very specific and narrow
spatial scope of the Illinois Pilot Program, these findings cannot be overly general-
ized. Its results may not be applicable to other areas of the state of Illinois itself, not
to mention its relevance to analysing the position in other U.S. states or European
A possible major critique of the Illinois Pilot Program concerns the focus on
specific classes of individuals who were subject to the programme instead of extend-
ing protections to most or all home loans. The Illinois Pilot Program protections
were triggered by the credit score of the borrower, regardless of whether the mort-
gage loan had unusual terms. This debatable solution, made undoubtedly with good
intent, was precisely the reason why the Illinois Pilot Program was doomed to fail.83
Notwithstanding the questionably narrow scope of the Illinois Pilot Program, the
idea of impartial counselling that it put forward should not be discarded.

10.5 A Possible Design of an Optimal Counselling


10.5.1 Triggering Mandatory Counselling

As the experience of the Illinois Pilot Program shows, criteria for counselling
should not refer to the borrower’s characteristics. Borrower-related criteria, how-
ever designed they might be, will inevitably be regarded as discriminatory. The
counselling triggers should be product-related. It is of course debatable whether all
types of mortgage contracts should require counselling or only those that are deemed
risky. A conceivable strategy may be to apply a lower protection threshold to all
home mortgage loans, i.e. an interactive computer counselling programme, whilst a
higher protection threshold, personalised counselling, would apply only to ‘expen-
sive’ and ‘risky’ mortgage loans.84
It appears that counselling should be trigged by both the complexity of the mort-
gage contract and the extent of financial liabilities assumed. Under Stark and

See Bates and Van Zandt 2007, p. 39.
Stark and Choplin 2010, p. 53.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 189

Choplin’s proposal, mandatory counselling would apply only to: (a) a ‘high cost
home loan’; (b) an ‘unaffordable home loan’, or (c) a loan ‘with potentially preda-
tory features’. According to Stark and Choplin, a loan with “potentially predatory
features” should be defined as a loan that has any one of the following features: (i)
a prepayment charge, (ii) a loan that permits the payment of interest only or nega-
tive amortization of the debt; and (iii) a loan that accrues interest at a floating rate.85
Some of these ‘risky’ mortgage contracts are indirectly recognized by the Mortgage
Credit Directive, in that the Directive provides specific rules for certain types of
mortgages: foreign-currency loans, interest-only loans and variable rates credits.
Needless to say, any list of loans with ‘potentially predatory features’ should be
subject to a regular review, as a reaction to new forms of market failure.
Designing a relatively broad catalogue of ‘high-cost’ and ‘risky’ mortgage loans
as trigger for counselling is of particular importance. Two major considerations are
at stake here. First, designations of mortgage contracts as ‘risky’ tend to become
publicly known and may therefore constitute a credible signal for potential borrow-
ers to avoid such mortgage products. The long-term consequences of an adequate
designation may be the decline of choice of risky products by borrowers. Secondly,
as was proven by the Illinois Pilot Program, smart triggers for counselling may
result in mortgage products selection being driven by the desire to avoid counsel-
ling and its associated costs.86

10.5.2 The Requirements Regarding the Person of Counsellors

An underlying assumption behind the idea of counselling programmes is that mort-

gage counsellors should be independent of mortgage lenders and intermediaries
(brokers). Counsellors will only be able to critically review the mortgage loans pro-
posals if they have no links to mortgage originators or brokers, and thus derive no
direct or indirect benefits from a decision of a borrower to take a specific mortgage
Furthermore, counselling programmes can have the desired effect if counsellors
are professionally trained: they have to be able to determine whether a loan is too
risky or too expensive for a borrower. The author of this paper has not attempted to
precisely calculate an estimate of the total costs of such training. Sources quoted by
Stark and Choplin indicated that it was estimated it would take approximately 50 h
to train a person to learn mortgage finance concepts, applicable legal regulations
and counselling methods.88 Training for future mortgage advisors should be

Stark and Choplin 2010, p. 55.
Agarwal and others 2009, p. 17.
Similarly, in the context of family suretyships, Trebilcock and Elliot 2001, p. 80.
Stark and Choplin 2010, p. 57.
190 P. Tereszkiewicz

provided by a specialist public administration office.89 Costs of such training will

undoubtedly vary from country to country and cannot be considered here in more

10.5.3 The Counselling Process

In order to ensure that the counselling process is as efficacious as possible, certain

standards and procedures have to be met. Borrowers would obtain the draft loan
agreements from lenders and have them reviewed by a counsellor. The objection
against counselling solution that borrowers do not have loan agreements and final
statements in front of them before closing on the transaction would not hold.90
Counsellors should be able to express clear recommendations about the suitability
of the mortgage loan for a borrower on the basis of the complete documentation
made accessible by lenders.
Given the assumption that mortgage counselling should be triggered by offers of
risky types of mortgages, a mortgage counsellor’s inquiry should aim at clarifying
why the applicant is interested in a given risky type of mortgage and what benefits
he or she expects from that the decision. Furthermore, the counsellor should explain
the particular risks of the mortgage contract offered to the applicant as compared
with a simple fixed interest rate mortgage loan as the safest way to reduce the extent
of borrower’s liability. Moreover, safer and less complex alternatives should be
proposed to the applicant. Finally, the mortgage counsellor should be able to offer
assistance to the applicant in obtaining other less risky mortgage loan offers. This
last feature could play a significant role in determining whether a counselling
scheme can have any meaningful effect.91
Having completed a counselling session, prospective borrowers would in prin-
ciple be free to take the financial decisions they prefer. As Trebilcock and Elliot
emphasize, “with the assistance of an advisor people should be free to arbitrate their
own preferences”.92
Given that a counsellor’s recommendation should be unbinding, there is a risk
that some borrowers will choose not to seek a lower priced or less risky loan due to
a relationship of trust that they have developed with the lender or broker. As a
means of countering the reliance on personal trust, Stark and Choplin propose that
the borrower should perform an interactive computer counselling program before
the borrower is allowed to work with the broker or lender. This computer counselling

Relating to the U.S., Stark and Choplin 2010, p. 56, recommended that the mortgage counsellors
should be trained by the United States Department of Housing and Urban Development.
This concern is raised by Engel and McCoy 2002, p. 1310.
Similarly Stark and Choplin 2010, p. 56, who claim that without this assistance many borrowers
may return to the original mortgage lender or broker they were working with due to psychological
factors discussed in the first part of this essay.
Trebilcock and Elliot 2001, p. 79.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 191

program would be interactive, educational, and would provide the borrower with:
(i) the APR figure they should be seeking based on his or her credit score, the debt
service ratio, and loan to value ratio based on the loan amount the borrower is seek-
ing, (ii) a calculation of the borrower’s gross monthly income and loan amount that
would yield a prudent debt service ratio, and (iii) a script for what to say to the
mortgage broker or lender.93 Once borrowers become aware that lenders are under
no duty to act in their best interest and may have financial incentives to offer them
risky loans, they may become less prone toward sub-optimal financial decisions.

10.5.4 The Costs of the Counselling Sessions

The central aspect regarding the counselling schemes relates to their funding and
costs. It is sometimes contended that creating a massive mortgage counselling
scheme would be very expensive.94 This argument is undoubtedly relative, for the
financial aspect depends on the design of the counselling schemes, the scope and
intensity of counselling provided to individual consumers. As already emphasized,
neutral mortgage counselling schemes are not meant as a form of general financial
education for consumers. They would not aim at providing long counselling ses-
sions with a view to educating consumers in financial matters, which would gener-
ate considerable costs but rather uncertain returns. Instead, counselling schemes
would provide assistance with regard to assessing specified draft contract (offers)
made to a borrower by a lender.
Under the Illinois Pilot Program a rule was adopted that the $300 cost of the
counselling session was to be borne by the mortgage originator and not the bor-
rower. A similar solution is also advocated by Stark and Choplin.95 Counselling on
issues, which would not trigger mandatory counselling, like the net economic ben-
efit issue, would have to be funded by borrowers themselves.
This solution proves controversial, for the financial sector would regard itself
unduly financially burdened. Nevertheless, there are two major justifications in
favour of lenders or brokers bearing the cost of counselling. First, one could justify
it by claiming that if lenders or brokers bring risky products to the marketplace
which are likely to yield financial gains to them, they should also bear the cost of
counselling consumers who may buy such products and bind themselves for decen-
nia with an obligation to repay. Secondly, one could rightfully claim that the global
costs of mortgage counselling, however high they might prove to be, are still limited
in comparison to the costs incurred in order to correct the market failings due to
misselling of mortgages. Much has been written about the dramatic consequences

Stark and Choplin 2010, p. 68.
Engel and McCoy 2002, pp. 1309–1311. More generally Trebilcock and Elliot 2001, pp. 82:
“The benefits of procedural protection need to be balanced against the increased cost of finance
that those protections entail”.
Stark and Choplin 2010, footnote 141.
192 P. Tereszkiewicz

of the subprime lending in the U.S. for the society at large.96 Similar consequences
are reported in Europe.97
There is no doubt that imposing the cost of counselling sessions on lenders will
lead to some unintended consequences. As regards the Illinois Pilot Program, it is
reported that brokers attempted to pass the $300 counselling fee to the borrowers in
the form of higher closing costs or administrative charges.98 Undoubtedly, manda-
tory counselling programmes will lead to some increase in costs of credit. However,
if costs of mandatory counselling sessions can be kept at a relatively low level
through regulation, such costs can be regarded as a negligible consequence of intro-
ducing counselling programmes. The expected overall improvement of decision-
making by borrowers would likely offset these costs.

10.5.5 The Limits of the Counselling Solution

The assumption underlying the mortgage counselling scheme is that if a mortgage

counsellor determines that a borrower could obtain the same loan at a better price or
safer terms, and makes such a recommendation to a borrower, that the borrower
should act rationally to decline the loan and seek a better-termed loan, in particular
where a counsellor assists with the process of obtaining a better-priced loan.99
Moreover, if specific types of loans are risky or predatory in nature to some con-
sumers, but not all of them, counselling may help to preserve the ability of consum-
ers to make informed choices in light of their individual circumstances. Certain
types of loans, e.g. with a floating interest rate, might be beneficial for a narrow
class of more sophisticated consumers and the process of counselling may promote
making such welfare-enhancing transactions.
The position of a counsellor merits close attention. The responsibility of counsel-
lors for enabling borrowers to make better choices regarding mortgage products
bears some resemblance to a gatekeeper strategy. A gatekeeper strategy entails
imposing liability on a private party who is not the cause of misconduct nor a ben-
eficiary, but who is in a position to prevent misconduct or wrongdoing by refusing
cooperation or support.100 It is not an objective of this essay to discuss a potential

Bar-Gill and Warren 2007, pp. 159–162.
For Hungary Balogh and others 2011, for Poland Rzeczpospolita from 2 February 2015, reports
on struggles of borrowers,1176120-Oferta-sprzedazy-
Bates and Van Zandt 2007, p. 23 with further references. See also Trebilcock and Elliot 2001,
pp. 69–70, warning that having to bear advice costs could be onerous for lower income
Stark and Choplin 2010, p. 61.
Trebilcock and Elliot 2001, pp. 67–70, discuss the gatekeeper duties of lenders in the context of
family sureties. Generally on a gatekeeper strategy Kraakman 1986, pp. 53–104.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 193

liability of counsellors toward borrowers.101 Rather, this essay underlines the poten-
tial of impartial third parties to detect and correct flawed mortgage choices and
enhance borrowers’ autonomy.
The counselling programmes, as advocated here, aim at improving decisions
about mortgage contracts. They are not meant to provide protection against factors
in loan delinquency which are not predictable or controllable by the parties when
the contract is made. Such factors are job loss by borrowers or health problems,
which are regarded as the main causes of defaults by debtors.102 Risks relating to
default and over-indebtedness resulting from events occurring during the lifecycle
of a long-term mortgage contract should be dealt with applying different legal mea-
sures, such as consumer bankruptcy or various financial aid programmes.
Finally, counselling programmes, as advocated here, should not be regarded as
the exclusive instrument of protection against risky mortgage loans. Mandatory
counselling introduces a measure of paternalism going beyond disclosure protec-
tion, but it should be conceived as a protection measure complemented among oth-
ers by usury laws and unfair contract terms legislation.103

10.6 Conclusions

These considerations relating to counselling programmes for mortgage products

reveal the multi-layered nature of regulatory nudges relating to financial products.
In theory, counselling programmes for mortgage borrowers could provide a useful
function of largely substituting mandated disclosure to consumers, by translating
the offered deals to them.104 From this perspective, introducing counselling pro-
grammes could go some way toward addressing major objections against mandated
disclosure as a regulatory tool in consumer law. Effective counselling programmes
could enable the borrower to find the lowest rate and cost loan available. At the
same time, significant dangers remain. If counselling programmes prove to be
unsuccessful, their mere existence will, paradoxically, benefit lenders from a civil
liability perspective. Lenders would commonly invoke a defence that a consumer
obtained independent professional advice before concluding a mortgage loan con-
tract. In this way civil liability of lenders as a regulatory tool could be seriously
weakened. This should reinforce the argument that the advocated mortgage

The principles and scope of such liability would depend among others on the status of counsel-
lors in a given legal order and also on the legal nature of their mandate.
See in general Ramsay 2007, pp. 578–580.
In a similar vein Ben-Shahar and Schneider 2013, p. 194: “Sometimes the choice between pater-
nalism and libertarianism simply can’t be avoided, and substantive regulation with no opting out is
Ben-Shahar and Schneider 2013, p. 190, list some of different kinds of information
194 P. Tereszkiewicz

counselling programme should not be regarded as the sole protection measure

against risky home lending.
The Illinois Pilot Program shows that a limited ability of borrowers to act on
counselling information has to be acknowledged. Nevertheless, this does not consti-
tute a reason against introducing counselling programmes. It is thinkable that regu-
lators achieve the goal of risk reduction by the mere threat of counselling and not by
the content of counselling sessions. No regulator can foresee all possible conse-
quences of norms that it adopts. When contributing to better mortgage choices is
recognized as a legitimate goal of a specific regulatory intervention, a legislative
instrument cannot be dismissed only because this legitimate goal was achieved
mainly as a result of unintended, but still legitimate consequences of the interven-
tion. Further empirical studies of counselling programmes regarding risky financial
products are necessary to discover the limits of their positive contribution, but there
is enough evidence to acknowledge their potential.
At a more general level, it appears justified to claim that a certain extent of pater-
nalism in designing the protection of mortgage borrowers is necessary. In the after-
math of recent experiences in the U.S. and in Europe, it must be recognised that
choices of borrowers do not always reflect their underlying preferences and contrib-
ute to contract failure in the mortgage market. Drawing on Trebilcock and Elliott,
paternalistic analysis begins by identifying the forms of contract failure that fre-
quently occur in a given transactional setting, and goes on to designing protections
aimed at efficiently responding to these typical patterns of failure.105 The protective
measures advocated here should be regarded as relatively ‘soft’ paternalism, with
the objective of assisting people in becoming better choosers and avoiding contract
failure.106 It is submitted here that neutral counselling constitutes a legal device,
which should be conceived as an alternative to mandated disclosure, avoiding a
number of its drawbacks.


Agarwal, Sumit, Amromin, Gene, Ben-David, Itzhak, Chomsisengphet, Souphala, and Douglas D.
Evanoff. 2009. Do financial counseling mandates improve mortgage choice and performance?
Evidence from a legislative experiment. Working Paper 2009-07. Chicago: Federal Reserve
Bank of Chicago.
Balogh, Csaba, Gereben, Áron, Karvalits, Ferenc, and Pulai, György. 2011. Foreign currency ten-
ders in Hungary: A tailor – made instrument for a unique challenge. Bank for International
Settlements Papers No 73. Available at:
Bar-Gill, Oren. 2012. Seduction by contract. Law, economics and psychology in consumer mar-
kets. Oxford: Oxford University Press.

Trebilcock and Elliot 2001, p. 84.
Waldron 2014, Section 4.: “I wish, though, that I could be made a better chooser rather than
having someone on high take advantage (even for my own benefit) of my current thoughtlessness
and my shabby intuitions”.
10 Neutral Third-Party Counselling as Nudge Toward Safer Financial Products? 195

Bar-Gill, Oren, and Elisabeth Warren. 2007. Making credit safer. University of Pennsylvania Law
Review 157: 101–201.
Bates, Lisa K., and Van Zandt, Shannon. 2007. Illinois’ new approach to regulating predatory
lending: Unintended consequences of borrower triggers and spatial targeting. Working Paper
2007-02. Urbana-Champaign: University of Illinois, Spatial Policy Analysis Research
Beatson, Jack, Andrew Burrows, and John Cartwright. 2010. Anson’s law of contract, 29th ed.
Oxford: Oxford University Press.
Ben-Shahar, Omri, and Carl E. Schneider. 2013. More than you wanted to know. The failure of
mandated disclosure. Princeton: Princeton University Press.
Burrows, Paul. 1995. Contract discipline: In Search of principles in the control of contracting
power. European Journal of Law and Economics 2: 127–147.
Craswell, Richard. 2006. Taking information seriously: Misrepresentation and nondisclosure in
contract law and elsewhere. Virginia Law Review 92: 565–632.
Engel, Kathleen C., and Patricia A. McCoy. 2002. A tale of three markets: The law and economics
of predatory lending. Texas Law Review 80: 1255–1348.
Feinman, Jay. 2009. The insurance relationship as relational contract and the ‘Fairly Debatable’
rule for first-party bad faith. San Diego Law Review 46: 553–572.
Financial Services Authority. 2011. Consultation paper 11/31, Mortgage market review: Proposed
package of reforms. Available at:
Frederick, Shane, George Loewenstein, and Ted O’Donoghue. 2002. Time discounting and time
preference: A critical review. Journal of Economic Literature XL: 351–401.
Mill, John Stuart. 1947. On liberty. Illinois: Northbrook.
Kahneman, Daniel, and Amos Tversky. 1979. Prospect theory: An analysis of decision under risk.
Econometrica 47: 263–291.
Kraakman, Reinier H. 1986. Gatekeepers: The anatomy of a third-party enforcement strategy.
Journal of Law, Economics, and Organization 2: 53–104.
Lusardi, Annemarie. 2008. Household saving behavior: The role of financial literacy. Information,
and financial education programs, National Bureau of Economic Research: Working Paper
No. 13824. Available at:
Meier, Stephan, and Charles D. Sprenger. 2013. Discounting financial literacy: Time preferences
and participation in financial education programs. Journal of Economic Behavior &
Organization 95: 159–174.
Moloney, Niamh. 2010. How to protect investors. Lessons from the EC and the UK. Cambridge:
Cambridge University Press.
Österreichische Finanzmarktaufsicht. 2015. Position der Finanzmarktaufischt zu
Fremdwährungskrediten und Informationen zur derzeitigen Lage. Available at: https://www.
Ramsay, Iain. 2007. Consumer law and policy. Oxford: Hart.
Schwartz, Alan. 2015. Regulating for rationality. Stanford Law Review 67: 1373–1410.
Schweizerische Nationalbank. 2009. Geschäftsbericht 2009. Available at
Shafir, Eldar, Itamar Simonson, and Amos Tversky. 1993. Reason-based choice. Cognition 49:
Stark, Debra Pogrund, and Jessica M. Choplin. 2010. A cognitive and social psychological analy-
sis of disclosure laws and call for mortgage counseling to prevent predatory lending.
Psychology, Public Policy, & Law 16: 85–131. Available at
Sunstein, Cass, and Richard Thaler. 2009. Nudge: Improving decisions about health, wealth, and
happiness. London: Penguin.
Tereszkiewicz, Piotr. 2011. Schutz des Angehörigenbürgen durch Aufklärung und Beratung im
Entwurf des Gemeinsamen Referenzrahmens (Draft Common Frame of Reference). Zeitschrift
für Europäisches Privatrecht 19: 851–873.
196 P. Tereszkiewicz

Tereszkiewicz, Piotr. 2015. Obowiązki informacyjne w umowach o usługi finansowe. Warszawa:

Wolters Kluwer Poland.
Trebilcock, Michael J., and Steven B. Elliott. 2001. The scope and limits of legal paternalism.
Altruism and coercion in family financial arrangements. In The theory of contract law. New
essays, ed. Peter Benson, 45–85. Cambridge: Cambridge University Press.
Tversky, Amos, and Daniel Kahneman. 1980. Availability: A heuristic for judging frequency and
probability. Cognitive Psychology 4: 207–232.
Waldron, Jeremy. 2014 October. It’s all for your own good. The New York Review of Book 61: 9.
Willis, Lauren. 2006. Decisionmaking and the limits of disclosure: The problem of predatory lend-
ing: Price. Maryland Law Review 65: 707–840.
Willis, Lauren. 2011. The financial education fallacy. American Economic Review 101: 429–434.
Chapter 11
The Potential Use of Visual Packaging
Elements as Nudges
An Analysis on the Example of the EU
Health Claims Regime

Kai Purnhagen, Erica van Herpen, and Ellen van Kleef

Abstract Regulators legislate businesses’ use of claims on product packaging by

mostly focusing on textual claims and the extent to which they could potentially
mislead consumers. Interpreting textual claims generally requires consumers to
engage in extensive and deliberate processing (so-called type 2 processing in dual
processing models in social psychology). In real life competitive choice settings,
however, businesses preferably make use of colours and other visual elements to
successfully attract consumers. In case this approach supports consumers in making
healthier food decisions, it can be seen as a form of nudging, as this approach largely
appeals to consumers’ reactive, intuitive modes of decision making (i.e., system 1
processing). Nevertheless, visuals may mislead consumers to a larger extent than
textual claims in that they overpromise health benefits of consuming the product. In
order to effectively regulate health claims in the EU, we claim that regulation has to
devote attention to the regulation of pictorial claims. We will first illustrate such a
mechanism and its potential for effective nudging on the example of health and
nutrition claims in the EU. We will then investigate on the example of the differ-
ences between textual and pictorial claims whether the EU health claims regulation
is effective in making sure that sales techniques of food companies are being regu-
lated as effective nudging instead of misleading marketing tactics.

K. Purnhagen (*)
Law and Governance Group, Wageningen University,
Hollandseweg Leeuwenborch (building 201), 6706 KN Wageningen, The Netherlands
E. van Herpen • E. van Kleef
Marketing and Consumer Behavior Group, Wageningen University,
Hollandsweg, 6706 Wageningen, The Netherlands

© Springer International Publishing Switzerland 2016 197

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_11
198 K. Purnhagen et al.

11.1 The Health and Nutrition Claims Example

Businesses make use of claims in order to gain a comparative advantage over other
companies stimulating consumer choice.1 A good example are health and nutrition
claims, which with reference to Art. 2 1. 4) and 5) Regulation No 1924/2006 on
nutrition and health claims made on foods (hereinafter ‘Regulation’) state, suggest
or imply that a food has a particular beneficial nutritional property or that a relation-
ship exists between a food category, a food or one of its constituents to health.
The Regulation applies to all nutrition and health claims. It aims to protect con-
sumers not only from potentially misleading information but also seeks to assist
them in making healthy informed food choices.2 To do so, EU regulation mostly
focuses on the regulation of potentially misleading textual claims. In European
health claims regulation this has also been recognised by Recital 10 Regulation:
Foods promoted with claims may be perceived by consumers as having a nutritional, physi-
ological or other health advantage over similar or other products to which such nutrients
and other substances are not added. This may encourage consumers to make choices which
directly influence their total intake of individual nutrients or other substances in a way
which would run counter to scientific advice. To address this potential undesirable effect, it
is appropriate to impose certain restrictions as regards the products bearing claims. In this
context, factors such as the presence of certain substances, or the nutrient profile of a prod-
uct, are appropriate criteria for determining whether the product can bear claims.

Interpreting this type of information generally requires consumers to engage in

extensive and deliberate processing (so-called type 2 processing in dual processing
models in social psychology). In real life competitive choice settings, however,
businesses preferably make use of colours and other visual elements to tap into
consumers’ intuitive processing modes (i.e., system 1 processing). These visual ele-
ments generally attract more consumer attention than textual information. While
textual information could, besides being a marketing practice, serve as a classical
form of consumer information to overcome the information asymmetry, businesses
preferably use pictorial claims, which are much closer to nudging techniques, to
enhance market share. One of the reasons why businesses move to pictorial health
claims may be found in Porter’s hypothesis that legal barriers incentivize businesses
to innovate and overcome them.3 In health claims regulation this has an ironic side
effect: Because of regulation which shall enable consumers to make a better choice,
businesses increasingly do not make use of textual claims for the benefit of pictorial
claims. As these tap into type-2-processing they are a better fit to create market
share but also may carry a bigger risk to mislead consumers in terms of the prod-
uct’s health benefit than textual claims. In order to effectively regulate health claims
in the EU, we claim that regulation has to devote attention to the regulation of picto-
rial claims.

Bremmers et al. 2013, p. 162.
See Sect. 11.2.1 below.
Porter 1991. Porter’s hypothesis has been transferred earlier to health claims requirements by
Bremmers et al. 2013, p. 164.
11 The Potential Use of Visual Packaging Elements as Nudges 199

We will hence first confront the highly regulated textual claims regime with
insights on the effects of claim wording, to assess the extent to which existing leg-
islation is effective in reaching the envisioned level of protection for consumers.
Second, we will show on the example of nutritional and health claims that the
approach taken by the EU regulator in practice to focus primarily on textual claims
without sufficiently taking into account the “pictorial, graphic or symbolic
representation”4 attached to the textual claim might not reach the goal envisaged by
the Regulation to ensure a high level of protection for consumers.

11.2 The Regulation of Health and Nutrition Claims

Contrasted to the Findings of Behavioural Sciences

In this section, we will first illustrate how health and nutrition claims are regulated
in the context of internal market regulation (Sect. 11.2.1) before we move on to
contrast these findings with empirical findings based on data from nudging theory
and behavioural sciences. We will discuss insights in behavioural sciences regard-
ing verbal nutrition and health claims (Sect. 11.2.2) as well as the impact of visual
packaging elements in this context (Sect. 11.2.3).

11.2.1 The Regulation of Nutrition and Health Claims

in the Context of Internal Market Regulation

Information regulation traditionally lies at the heart of consumer protection in the

EU. It is a common belief of traditional internal market law that consumers are
adequately protected if the Union provides information of potentially dangerous
products to consumers (information-related regulation) and does not prohibit or
limit the design of consumer products (content-related regulation), the so-called
information paradigm.5 Recently, however, EU internal market regulation is in flux.
The Lisbon-agenda and the newly introduced concept of the internal market,6 the
growing overlap of fundamental rights and fundamental freedoms,7 and the increase
of more intrusive secondary legislation on consumer products, endorsed by the
Court of Justice of the European Union (hereinafter CJEU)8 already advent a trans-

Art. 2 (2) 1 of the Regulation.
Franck and Purnhagen, ‘Homo Economicus’ 2013, pp. 334–340; Purnhagen, ‘Virtue’ 2014,
pp. 329–332; Steindorff 1996, pp. 195–196; Usher 2001, pp. 152–153.
Barends 2009, pp. 123 et seqq.; Riley 2007, pp. 703 et seqq.
Trstenjak and Beysen 2013, pp. 293 et seqq.
Unberath and Johnston 2007, p. 1238 “(O)n the whole the Court has been at pains to grant the
consumer protection directives a wide scope of application and – crucially- to give its provisions
the most interventionist reading.” See particularly with respect to consumer information in foods
200 K. Purnhagen et al.

formation of consumer protection in the EU. This development culminates in recent

judgments of the CJEU stipulating that the aim of consumer protection at EU level
is not only the adjustment of the ‘level of knowledge’ (information-regulation) of
the consumer, but also his ‘bargaining power’ (potentially including also content-
related regulation) ‘vis-à-vis the seller or supplier’.9
The health-and-nutrition-claims regime at EU level as laid down in the Regulation
was issued somewhat in the middle of the quarrel between content-related and
information-related regulation, as it regulates the quality of information as such.10
In this sense, it starts and accepts the presumptions of the information paradigm as
being the core of consumer protection at EU level: Recital 1 of the Regulation
makes clear that sovereign consumers shall and should be the ones to decide on the
level of protection of internal market products. These sovereign decisions, however,
need to be made in an environment provided for by the regulator, mainly through
the provision of the protective function of fundamental rights11: The right to physi-
cal integrity as enshrined in Art. 3 (1) Charta of Fundamental Rights may be inter-
preted as obliging the EU legislator to protect individual consumers from making
decisions that are contrary to their well-being. In this sense, particularly in the area
of EU health and nutrition claims regulation, the benchmark of the ‘average con-
sumer’ has changed to incorporate a consumer that is more vulnerable to the pitfalls
from choice architecture. Relevant in this respect is that Art. 5 (2) Regulation12
explicitly states that “the use of nutrition and health claims shall only be permitted
if the average consumer can be expected to understand the beneficial effects as
expressed in the claim”.
Taking all this into account, the Regulation lays “down specific provisions con-
cerning the use of nutrition and health claims concerning foods to be delivered as
such to the consumer.”13 The main regulatory goal behind this very much lies in line
with the information paradigm in internal market law: it may ensure, on the one
hand, “a high level of protection for consumers and to facilitate their choice,”14

particularly the judgment in Karl Berger v Freistaat Bayern (C-636/11), and the respective annota-
tion by Purnhagen 2013, pp. 71 et seqq.
C-508/12, Vapenik vs. Thurner, para 26.
Purnhagen, ‘precautionary principle’, 2015.
The recent judgment of the CJEU in Weintor emphasised the importance of interpreting the
health claim in the light of fundamental rights, see Deutsches Weintor eg v Land Rheinland-Pfalz
(C-544/10), nyr. Weatherill cites this judgment as “a helpful example of the structure and style of
the analysis one may anticipate (and hope for)”, see Weatherill, p. 415.
For a general account from the perspective of EU food law see also Recital 41 of Regulation
(EU) No 1169/2011 of the European Parliament and of the Council of 25 October 2011 on the
provision of food information to consumers, amending Regulations (EC) No 1924/2006 and (EC)
No 1925/2006 of the European Parliament and of the Council, and repealing Commission Directive
87/250/EEC, Council Directive 90/496/EEC, Commission Directive 1999/10/EC, Directive
2000/13/EC of the European Parliament and of the Council, Commission Directives 2002/67/EC
and 2008/5/EC and Commission Regulation (EC) No 608/2004, 2011, L 304/18; for a more thor-
ough assessment see Purnhagen and van Herpen 2014.
Recital 3 Regulation (EC) No 1924/2006.
Recital 1 Regulation (EC) No 1924/2006.
11 The Potential Use of Visual Packaging Elements as Nudges 201

thereby largely refraining from regulative value judgments on how consumers shall
exercise their choice. On the other hand, the Union regulator acknowledges that
such a consumer choice presupposes a regulatory environment, which acknowl-
edges that “(a) varied and balanced diet is a prerequisite for good health.”15 By this,
the regulator acknowledges that consumers’ free choice can be limited by internal
market regulation when it comes to the composition of the consumers’ diet.
According to this, regulatory intervention into the general provision of free choice
of consumers is exceptionally permitted in order to ensure that consumers choose
for a ‘varied and balanced diet.’ For this matter Art. 5 of the Regulation prohibits
the use of nutritional and health claims, which do not satisfy some ‘general condi-
tions’ substantiated further in EU law.
To cope with the uncertainties accompanied with these requirements,16 the
Regulation stipulates a positive list of textual nutritional claims in the
ANNEX. Although the Regulation itself encompasses non-textual claims (pictorial,
graphic or symbolic representations are explicitly mentioned), no claim on this list
focuses on pictorial claims. Likewise, the health claims that have been authorised
by the Commission so far embrace only textual claims.17 The respective table,
which represents the EU Register on nutrition and health claims is also designed in
a way that it will be hard to incorporate claims which involve pictures. The few
Court decisions available on the subject focus on textual claims only. Just to men-
tion a few: In Weintor the CJEU had to judge on the question whether the label
‘easily digestible’ on a wine bottle of a German wine cooperation would qualify as
a claim in the sense of the Regulation.18 In Monsterbacke, the CJEU was confronted
with the question whether the Slogan “As important as a daily glass of milk!” falls
under the EU health claims regime.19 In Green Swan, it was subject to the CJEU’s
decision whether the phrase “[t]he preparation also contains calcium and vitamin
D3, which help to reduce a risk factor in the development of osteoporosis and frac-
tures” classifies as a health claim.20 The German Supreme Court BGH decided that
the use of the trademarks ‘Praebiotik® + Probiotik®’ as well as the slogan “Praebiotik®
+ Probiotik® with natural lactic acid cultures – Praebiotik® for the support of a
healthy intestinal flora” qualify as health claims.21 In determining whether these
claims are health claims or are misleading, none of these judgments took into
account any non-textual claims and not even the pictorial environment these claims
were embedded in.
If we look into scholarly work on health claims and nutrition, the picture is simi-
lar. A majority of papers in the consumer food science field focuses on consumer

Recital 1 Regulation (EC) No 1924/2006.
Bremmers et al. 2013, p. 163.
See the EU Register on nutrition and health claims, available at
C-544/10 Deutsches Weintor eg v Land Rheinland-Pfalz, nyr.
C-609/12 Ehrmann AG v Zentrale zur Bekämpfung unlauteren Wettbewerbs eV, nyr.
C-299/12 Green Swan, nyr.
BGH, 26.02.2014 – I ZR 178/12 = NJW-RR 2014, 1129.
202 K. Purnhagen et al.

response to textual claims.22 This has spurred the distinction between differently
worded claims and their effects,23 but attention for non-textual claims is scarce.
This short analysis illustrates that the main focus of attention of the misleading
potential of health and nutrition claims is on textual claims. With the focus on text,
the EU regulator lies in line with traditional legal scholarship and practice that pays
little attention to non-textual forms of written advocacy.24 It is hence no wonder that
textual claims so far have been the main target for regulators in the EU. This is
despite of the fact that the concept of ‘claim’ according to Art. 2 (2) 1 of the
Regulation explicitly covers also ‘pictorial, graphic or symbolic representation’. In
the next section, we will discuss behavioural insights into textual health and nutri-
tion claims, before moving to insights related to these visual elements.

11.2.2 Textual Claims: The EU’s Regulation in Contrast

to the Insights Gained from Behavioural Sciences

With respect to textual claims, studies examining consumer behaviour and decision
making have focused on consumer understanding of these claims, and the potential
that consumers are misled by such claims, individual differences in how consumers
understand and respond to these claims, and the effect of these claims on consumer
choice. These three topics are pertinent for EU regulation, and we will discuss these
in turn.

Consumer Understanding of Claims

Comprehension of health and nutrition claims is a requirement for correct use of

such claims and a key cognitive process in decision making. One study topic is the
wording of claims and how these may be difficult and confusing to consumers, even
when attention is specifically directed to them.25 For many consumers, there is too
much information present on product packages. Consequently, the use of words
such as ‘may’ (e.g., “a healthy diet low in saturated and trans fat may reduce the risk
of heart disease”) tend to be ignored by consumers when interpreting the meaning
of a claim, even though this word represents an important nuance for policy makers
and scientists.26
Understanding of health claims can be examined using both qualitative and
quantitative methods, and contrasted against objectively correct information. For

Grunert and Wills 2007, pp. 385–399, Van Kleef, Dagevos 2015, pp. 291–303; Lähteenmäki
2013, pp. 196–201.
Van Trijp and van der Lans 2007, pp. 305–324.
Porter 2014, p. 1752.
Nocella and Kennedy 2012, pp. 571–580.
Harris 1977, pp. 603–608.
11 The Potential Use of Visual Packaging Elements as Nudges 203

instance, Grunert and colleagues27 investigated a yogurt product with a health claim
(‘helps strengthening the body’s natural defences’) by exposing German respon-
dents to the product and corresponding advertisement. Answers to open-ended
questions such as “after seeing this pack and commercial, if you had to tell a friend
what this product is, what would you say?” were content-analyzed and coded as
safe (in line with scientific evidence on claim), risky (inferences beyond scientific
evidence) or other value (irrelevant or vague notions). Results showed that 67 % of
the respondents’ responses could be classified as safe and 21 % of the interpreta-
tions were risky. A positive attitude towards functional foods was the best predictor
of risky answers. This implies that some groups of consumers are particularly vul-
nerable to exaggerated understanding of claims.
A similar approach on understanding how consumers may (over)interpret claims
was carried out by Bilman and colleagues.28 In contrast to the findings of Grunert
and colleagues, they found that the majority of respondents in four European coun-
tries correctly interpret satiety-related claims (e.g., ‘feel full for longer’) in that they
do not expects benefits beyond their literal meaning (e.g., ‘lose weight’). For satiety-
related claims, consumers understand that additional personal efforts are required to
obtain potential weight-control benefits. This suggests that perhaps the extent to
which claims are correctly understood differs across types of claims.
When confronted with a health claim, consumers will make inferences based on
the claim and often these inferences are biased. Inference making occurs because
people typically process health claims with low levels of involvement. Consequently,
they do not cognitively elaborate on the meaning of the claim, but instead make
inferences about what is missing from readily available information. How individu-
als draw inferences in situations of incomplete information or limited knowledge is
often explained by schema-based deduction. Schemas are associative networks of
information items based on prior knowledge. For example, a logo or word (e.g.,
‘low fat’) can trigger beliefs or associations between learned concepts such as ‘low
fat is less tasty’, and ‘low fat equals less calories’.29 These associations fill in the
gaps in knowledge when an individual tries to make sense of a product label. A
wide variety of cues can be used to activate the mental links in the schema. The type
of cues used depends on the accessibility (salience) and diagnosticity (relevance) of
the information.30
Several biased inferences may occur: overgeneralization of the claim to general
health benefits, distorted inferences about other product aspects, and incorrect
inferences about competing products. Regarding the first type of biased inferences,
a serious concern is that consumers over interpret health and nutrition claims in that
they ascribe inappropriate health benefits to them. Consumers may overgeneralize

Grunert et al. 2011, pp. 269–277
Bilman et al. 2012, pp. 912–920.
Chandon 2013, pp. 1–6.
Kardes et al. 2004, pp. 230–256.
204 K. Purnhagen et al.

from specific claims (e.g., ‘no cholesterol’) to general health (e.g., ‘healthy’).31 This
is also referred to as the ‘magic bullet effect’.32
Another type of biased inference occurs when claims about one aspect of a prod-
uct lead to distorted inferences about various other aspects of the product. The gen-
eral ‘halo effect’ is a bias uncovered in social psychological research in which a
person’s positive impression of one aspect of a stimulus extends to other aspects as
well, even though these aspects may be unrelated.33 Health halos for food products
occur when the presence of a health claim has a strong effect on expected and expe-
rienced taste and health perceptions, above and beyond the description of ingredi-
ents/nutrition content.34 In other words, the claim makes consumers perceive them
as generally superior. For example, a low cholesterol claim can lead consumers to
infer a lower fat level than is actually present, and this perception may bias estima-
tions of the calories the food contains.
Moreover, if a product prominently displays an unusual attribute for that cate-
gory, consumers may infer that competitive products do not have the attribute.
People follow basic rules when speaking to each other, called ‘conversational max-
ims’ in Grice’s theoretical accounts of the logic of conversations.35 For example, the
maxim of quantity directs speakers to say things as informative and concise as pos-
sible. Consumers exposed to a nutrition claim may falsely conclude that all infor-
mation needed to draw legitimate inferences has been provided. For example, the
promotion of a lollipop using a ‘fat free’ claim in New Zealand in 2004 was seen by
the Complaints Board as misrepresenting the product category as it may lead to the
inference that candy without fat is unique.36
Besides overgeneralizations, a ‘boomerang effect’ may occur when claims gen-
erate reactance. Similar to the ‘forbidden fruit’ idea, warnings or claims may
threaten consumers’ experienced freedom to eat a certain food. The message “In
this product, 90 % of the calories come from fat. Warning: the US Surgeon General
has determined that eating high fat food increases your risk of heart disease” made
participants in a supposed taste test want to eat the fatty product more.37
The insights regarding biased inferences resulting from health and nutrition
claims can be informative for legislation. Of particular interest is the distinction into
different types of biases, that is, the insight that consumer misunderstanding can
take different forms. Not only can consumers make biased inferences regarding the
product with a health or nutrition claim, but such a claim may also lead to incorrect
inferences about competing products (even when the claim is objectively correct, as
with the ‘fat free’ lollipop). Legislation should take both inferences about the prod-
uct in question as well as inferences about competing products into account.

Andrews et al. 1998, pp. 62–75.
Roe et al. 1999, pp. 89–105.
Nisbett and Wilson 1977, pp. 250–256.
Chandon, and Wansink 2012, pp. 571–593.
Grice 1975, pp. 41–58 and Hastak and Mazis 2011, pp. 157–167.
Hastak and Mazis 2011, pp. 157–167; Parsons and Schumacher 2010, pp. 1539–1558.
Bushman 1998, pp. 97–101.
11 The Potential Use of Visual Packaging Elements as Nudges 205

Individual Differences

In addition to studies investigating biases in consumer understanding of claims, a

large stream of literature exists in which individual differences in understanding of
claims and nutrition information has been examined. A consistent finding across
these studies is that older consumers and people with lower levels of education are
more likely to have difficulty understanding the terms used on food labels.38 For
example, older adults (over 65), people with lower levels of education and those
from lower social classes are less likely to be able to accurately interpret front-of-
pack logos. Also minority ethnic groups have difficulty interpreting them.39
Although young people may understand certain labels better, middle-aged and
elderly consumers tend to be more health-oriented and interested in nutrition infor-
mation at food labels. That is because they are more likely to experience health
problems themselves or among people in their near social environment.40
Moreover, substantial differences in how consumers look at health claims exist
across countries.41 For example, understanding of Guidelines Daily Amount (GDA)
labels was high in the UK, Sweden and Germany, and more limited in France,
Poland and Hungary.42 Limited literacy is also an important factor to consider as 75
million Europeans lack basic reading and writing skills.43 Most research on con-
sumer understanding of nutrition- and health-related information at food packages
has been done with literate and often highly educated participants. Low-literate
consumers have difficulty interpreting nutrition summary information such as per-
centage of daily values (%DVs) and have to spend considerable effort on relatively
simple reading tasks.44
The findings that there are large individual differences in understanding of health
and nutrition claims, is of direct relevance for the ongoing debate about the ‘average
consumers’ concept. Traditionally, internal market consumer regulation targeted
the normative ‘average consumer,’ who had to bear the burden of perceiving and
processing information, and also the drawbacks that may follow should they carry
out a market transaction suffering a cognitive deficit.45 Consequently, entrepreneurs
needed to provide all information available to consumers while consumers had to
bear the burden of processing this information on potentially hazardous products

Cowburn and Stockley 2005, pp. 21–28.
Malam et al. 2009.
Nocella and Kennedy 2012, pp. 571–580.
Van Trijp and Van der Lans 2007, pp. 305–324.
Grunert et al. 2010. pp. 261–277.
United Nations Regional Information Centre for Western Europe (UNRIC). EU: 75 millions
adults lack basic reading skills. Accessed at January 21 2015 at
Viswanathan et al. 2009, pp. 135–145.
Franck and Purnhagen 2014, p. 337.
206 K. Purnhagen et al.

and services.46 Such a consumer was, as the Court clarified in ‘Gut Springenheide,’47
a “reasonably well informed and reasonably observant and circumspect.”48 We find
this image of the ‘average consumer’ today in Recital 18 of Directive 2005/29/EC
as someone “who is a reasonably well informed and reasonably informed observant
and circumspect, taking into account social, cultural and linguistic factors.” To be
sure, this notion has been developed as a normative response to the specific require-
ments of internal market regulation at the time when the internal market was still in
its infancy.49 Today, as Recital 18 of Directive 2005/29/EC also clarifies, the classi-
cal concept of the ‘average consumer’ needs to be make sure that it also “contains
provisions aimed at preventing the exploitation of consumers whose characteristics
make them particularly vulnerable to unfair commercial practices.” While the ‘aver-
age consumer’ in the past ‘required’ from consumers to process all kind of informa-
tion, this new formulation opens up a leeway to determine which information
‘vulnerable consumers’ can actually process. This is particularly relevant in the area
of health claims. As stipulated earlier, Art. 5 (2) Regulation50 explicitly emphasizes
this change of the understanding of the ‘average consumer’ as “the use of nutrition
and health claims shall only be permitted if the average consumer can be expected
to understand the beneficial effects as expressed in the claim”. Adding insights from
behavioural research to the notion of the average consumer can form a good means
to do so in a scientifically robust way.51 In order to assess consumer understanding
in a more systematic way, and to better understand what misleading claims look like
and allow monitoring over time, the establishment of a standardized methodology
involving both qualitative and quantitative approaches52 preferably to be used at EU
level,53 would be desirable.

Steindorff 1996, p. 195.
Gut Springenheide and Tusky (C-210/96) [1998] E.C.R. I-4657.
Gut Springenheide and Tusky (C-210/96) [1998] E.C.R. I-4657.
Opinion of Advocate General Gulmann of 29 September 1993, Case C-315/92 of 2 February
1994, Verband Sozialer Wettbewerb eV v Clinique Laboratoires SNC and Estée Lauder Cosmetics
GmbH, 1994, ECR I-317, para 25. See Duivenvoorde for a recent summary and assessment of the
judgments on the average consumer models.
For a general account from the perspective of EU food law see also Recital 41 of Regulation
(EU) No 1169/2011 of the European Parliament and of the Council of 25 October 2011 on the
provision of food information to consumers, amending Regulations (EC) No 1924/2006 and (EC)
No 1925/2006 of the European Parliament and of the Council, and repealing Commission Directive
87/250/EEC, Council Directive 90/496/EEC, Commission Directive 1999/10/EC, Directive
2000/13/EC of the European Parliament and of the Council, Commission Directives 2002/67/EC
and 2008/5/EC and Commission Regulation (EC) No 608/2004, 2011, L 304/18; for a more thor-
ough assessment see Purnhagen and van Herpen (2014).
Sibony 2015, pp. 71–107.; Trzaskowki 2011, pp. 377–392.
Leathwood et al. 2007, pp. 474–484.; Grunert et al. 2011, pp. 269–277; On the changing internal
market environment and its impact on the average consumer see Purnhagen and van Herpen 2014.
Purnhagen and van Herpen 2014.
11 The Potential Use of Visual Packaging Elements as Nudges 207

Effects of Textual Claims on Product Choice

The strong focus on textual health claims in EU internal market regulation as well
as in studies on consumer understanding of claims seems to be in stark contrast to
research findings from data collected in behavioural sciences on product choice.
When explicitly asked, many consumers see claims as useful.54 Likewise, in con-
trolled experiments where consumers are specifically instructed to pay attention to
nutrition- and health-related claims or are sensitized to health issues, nutrition infor-
mation has been found to affect the perceptions of the healthiness of a product, and
to some extent purchase and consumption behaviour.55 However, nutrition- and
health-related claims may have little or no effect on consumers when their attention
is not specifically directed towards the information. In real life supermarket set-
tings, where consumers are confronted with a multitude of cues and information
under often occurring time pressure, most consumers seem to have little time and
motivation to engage in the processing of nutrition information.56 In a study in
which consumers were asked to think aloud while shopping for foods in a supermar-
ket, actual stated label use was very limited, both when shoppers were shopping
normally and when they were asked to shop more healthily.57 Other studies in which
consumers were unobtrusively observed in stores show similar low levels of use.58
Although some consumers are more likely to look for nutrition information while
shopping (e.g., consumers trying to control their weight59 or having specific health
concerns60), lack of attention is an important obstacle in the use of nutrition labels.61
When it comes to health and nutrition claims, the verbal statement of the claim
and how consumers may interpret this, has received ample attention both in empiri-
cal consumer research and in legislation. However, given that these verbal claims
often do not appear to influence product choice to a large extent in real life, it becomes
pertinent to look beyond the verbal statement itself and examine the packaging con-
text in which the claim is presented, including visual packaging elements.

11.2.3 Visual Packaging Elements: The Missing Piece

In addition to verbal nutrition and health claims, product packaging elements, such
as its colour and certain graphics, can also change consumers’ perceptions of the
healthfulness of products (and thereby purchase behaviours). These visual elements

Williams 2005, pp. 256–264.
van Herpen et al. 2012, pp. 22–34 et seqq.; Lähteenmäki 2013, pp. 196–201.
Grunert et al. 2010, pp. 177–189.
Raynor et al. 2001, pp. 24–31.
Grunert and Wills 2007, pp. 196–201.
Vyth et al. 2010, pp. 1882–1889.
Schor et al. 2010, pp. 22–32.
Van Trijp 2009, pp. S41–S48; van Herpen and van Trijp 2011, pp. 148–160.
208 K. Purnhagen et al.

tap into consumers’ system 1 processing and generally attract more attention than
textual information.62 Associations that are spontaneously triggered by certain
colours (e.g., blue = light) or visual images can be used by consumers as a heuristic
to draw inferences about a product. In this section, we will review evidence from
studies on consumer behaviour, which suggests that visual elements of a food pack-
age, drawing on consumers’ automatic system 1 processing, may be more influen-
tial in changing consumers’ perceptions of a food product than textual information
about a food product, which requires more systematic type 2 processing. Such
visual information is potentially misleading when it is inconsistent with the textual
information on a food package. Yet, whereas the textual nutritional information on
food packages is strictly regulated, the use of visual elements such as colours and
graphics is regulated to a much smaller extent in practice. Although the legislation
explicitly includes any “pictorial, graphic or symbolic representation, which states,
suggests or implies that a food has particular characteristics” (Art. 2 (2) 1 of the
Regulation), in practice there has been an almost exclusive focus on textual claims
(see Sect. 11.2.1). This raises the question of whether package colours and other
visual elements of food packages should be (more strictly) regulated.

The Importance of Visual Elements

Pictorial, graphic or symbolic representations on food packaging constitute visual

elements, and these elements are relevant in consumer decision making. In fact,
research in psychology has argued and shown that visual elements in general tend
to have longer lasting effects on memory than textual elements (the picture-
superiority effect63). In addition, visual elements are vivid, and therefore may attract
attention relatively easily.64 Furthermore, when people are under time pressure, they
tend to pay less attention to textual ingredient information and instead rely rela-
tively more on pictorial information.65 Thus, notwithstanding individual differences
in the preference for visual and textual information,66 visual information in general
matters greatly for consumer choice.
Visual elements are included on food packages for various reasons. Visual ele-
ments may provoke feelings and communicate affect. They can also draw attention
to products. Given the crowded and overwhelming display of products in a typical
store, capturing the consumers’ attention within a limited time can be very influen-
tial for product purchase.67 Additionally, visual elements may inform consumers

Meyers-levy and Peracchio 1995, pp 121–138; Underwood and Klein 2002, pp. 58–69.
Childers, and Houston 1984, pp. 643–654, Stenberg et al. 1995, pp. 425–441; Hockley 2008,
pp. 1351–1359.
Underwood and Klein 2002, pp. 58–68.
Pieters and Warlop 1999, pp. 1–16.
Childers et al. 1985, pp. 125–134.
Clement et al. 2013, pp. 234–239; Milosavljevic et al. 2012, pp. 67–74; Ottenbring et al. 2013,
pp. 1211–1225.
11 The Potential Use of Visual Packaging Elements as Nudges 209

about intrinsic characteristics of a product (e.g., taste) that are not directly observ-
able.68 This can be done through the use of uniquely differentiated visual elements
that are not used by competing products, but also through the use of established
symbols and colours.
The use of distinctive images can help people learn and remember quality differ-
ences between products (e.g., that the product in the purple packaging tastes
deliciously).69 Distinctive visual elements can increase the accuracy of memory, and
help consumers remember their experience with the product. Yet, visual differentia-
tion is not always helpful. In particular, the use of established visual elements
(colours, images, symbols) can allow consumers to quickly grasp what a product is
about, and identify unobservable attributes of the product. The consistent use of the
same visual elements across multiple products to communicate about intrinsic prod-
uct attributes ensures that these visual elements are diagnostic and allows consum-
ers to quickly identify important product attributes.70 For instance, in the milk
category, the distinction between whole milk, reduced fat milk, skim milk, etc., is
usually communicated through the colour of the milk bottle top, although the spe-
cific colour codes differ across countries.71 It can be beneficial for manufacturers to
use such established colours that adhere to the colour norms in a category, even
though this may diminish the visual differentiation of the packaging.72 Thus, when
most ‘light’ (low fat) versions of products in a category carry a light blue colour, it
may be advantageous to also use this colour when marketing a light product, so as
to clearly communicate this to consumers. This, in turn, will strengthen the learned
association between the colour ‘light blue’ and the attribute ‘low fat’ in people’s
mind. Over time, due to repeated joint exposures, the colour obtains meaning.
Visual elements do not merely represent reality (e.g., show what is inside the
packaging) but are also capable of sophisticated rhetorical tasks, and can be seen as
a form of information.73 Visual elements such as pictures may have a ‘chunking’
function, in the sense that these can communicate many bits of information at
once.74 Consumers can ‘read’ information about the attributes of products from
these visual elements, and this is made possible because the mass media are teaching
them a large vocabulary of images.75 Thus, symbols stand for complex ideas because
years of communication have layered them with meaning. Consumers can interpret
these symbols and other visual elements, and distract information about the product.
In this way, a heart symbol can communicate heart health.

Underwood and Klein 2002, pp. 58–68; van Herpen and Pieters 2007, pp. 77–88.
Warlop et al. 2005, pp. 27–44.
van Herpen and Pieters 2007, pp. 77–88.
Labrecque and Milne 2013, pp. 165–176.
Scott 1994, pp. 252–273.
Gardner and Houston 1986, pp. 64–78.
Scott and Vargas 2007, pp. 341–356.
210 K. Purnhagen et al.

Complicating matters, the meaning of visual elements is contextually situated,

and thus depends on other elements in the direct context,76 implying that the mean-
ing of a visual element depends on other packaging elements. Thus, whereas a heart
symbol in one case can communicate heart health, in another case it may communi-
cate (romantic) love or affection. The type of product on which the symbol occurs,
and the context of other packaging elements, clarifies which meaning is intended.
This implies that it may not be possible to regulate pictorial, graphic or symbolic
representations with general rules that apply to all product packages. After all, the
same symbol may have a very different meaning in various contexts. This potential
complexity clearly indicates that regulation, in order to do so effectively, has to find
a way to incorporate the context in which the pictorial, graphic or symbolic repre-
sentations are used. Given the diversity of potential visual elements and context
factors, traditional regulatory tools for health claims such as black or white lists of
approved visual elements similar to the current list of approved textual health claims
are unlikely to be successful. What one might consider is a regulation in analogy to
how design elements are protected under intellectual property law.

Combining Visual and Verbal Health Claims

The specific relation between visual elements on food packaging and verbal claims
related to nutrition or health has only recently received research attention. Results
of these studies consistently indicate that visual elements influence health percep-
tions and choice of products, even when verbal health claims are present. For
instance, Carrillo, Varela, and Fiszman77 examined ten enriched biscuits with dif-
ferent nutritional or health claims (e.g., with folic B and non-visible fibre: delicious
way to help care for the heart; with omega 3; high fibre content). People could taste
test the product either with or without verbal information about the products, or they
saw the packages with visual design elements. Results showed that perceived
healthiness of the biscuits differed between these conditions, suggesting that the
package as a whole could raise high expectations of healthiness for the biscuits.
Importantly, for several biscuits, perceived healthiness differed between people
who tasted the product and saw the verbal information (the nutritional fact panel
and nutrition and health claims, without design features) and people who saw the
package (containing the same information, but with visual design elements present
as well). This provides a first indication that visual elements may influence how
consumers interpret a health claim.
A more direct test78 has shown that visual elements can serve as a context in
which a verbal health claim is interpreted. The idea was tested that visual elements
directly influence product beliefs even when the verbal health claim of the package
is held constant and provides accurate product attribute information. Specifically,

Scott 1994, pp. 252–273.
Carillo et al. 2012, pp. 209–216.
Bone and France 2001, pp. 467–489.
11 The Potential Use of Visual Packaging Elements as Nudges 211

people evaluated the package of a cola which contained a caffeine claim (“This cola
contains 20 mg of caffeine. Regular colas contain between 43 and 49 mg of caf-
feine”) as well as a pictorial element depicting either a football player with a light-
ning bolt or a man lounging under a palm tree. When asked about caffeine content,
this latter picture affected consumer perceptions of caffeine content in the presence
of the verbal claim, showing that the graphical picture influenced people’s percep-
tions of the product even in the presence of correct verbal information.
Other studies have examined the use of colour on packaging with nutritional
information or claims, and have likewise shown that health perceptions depend on
visual elements in the presence of accurate verbal information. For instance, con-
sumers perceive a candy bar as more healthful when the calorie label is green rather
than red or white.79 Summarizing, the empirical results from consumer psychology
clearly indicate that visual elements on pack are of great influence on consumers,
also in the presence of accurate verbal information, and thus have the potential of
misleading consumers.

11.3 Discussion and Conclusion

We were able to show that the EU health claims regime provides an excellent set-
ting that can be used for nudging. We then investigated whether the current EU
health claim regime is fit for purpose on the example of regulating textual versus
pictorial information. We could show on the example of literature reviews that
visual elements matter in health claims. A reason for this might be found in the
theory of visual rhetoric, which argues that images are capable of communicating
subtle and complex messages.80 Furthermore, visual elements and colours allow for
rapid and automatic processing (system 1 processing), whereas textual information
generally requires more deliberate processing (system 2 processing). Thus, visual
elements can communicate relatively complex messages without necessarily rely-
ing on system 2 processing to a large extent, and can thus be of great influence.
Graphic presentations and symbols are especially beneficiary for consumers who
have difficulty processing and understanding basic verbal information, grossly
referred to as vulnerable consumers in internal market law.81 The new notion of the
‘average consumer’ in internal market law expressly seeks to incorporate such vul-
nerable consumers. However, insecurities exist regarding how to match the classical
notion of the ‘average consumer’ with the ‘vulnerable consumer’. Here, the findings
mentioned above can be of help. If studies from behavioural science show that con-
sumers are particularly sensitive to pictorial claims, a health claims regime which

Schuldt 2013, pp. 814–821.
Scott and Vargas 2007, pp. 341–356
Recital 18 UCPD.
212 K. Purnhagen et al.

seeks to assist consumers getting to ‘(a) varied and balanced diet’82 should listen to
such insights.
The studies presented herein show that pictorial claims matter and should be
taken care of by jurisprudence ‘in action’. Drawing normative conclusions beyond
this statement should be handled with care. Indeed, the studies presented here show
that claims might not have the same effect on all consumers, as, within the borders
of the methodology of the studies, perception of claims differed according to indi-
vidual consumer features. It is hence tempting to also ask from jurisprudence to
design health claim regulation along the lines of ‘vulnerability categories’ as pre-
sented in these studies. That would, to a certain extent, ease the arbitrariness of
political decision-making from determining vulnerability levels and may also give
businesses more legal certainty in terms of foreseeability when acting on the EU
Despite of these advantages one shall not forget that a single study only has the
potential to generate insights within the limits of the respective methodology.
Drawing normative conclusions for internal market regulation from studies that
have been designed to satisfy a different purpose and in a different context has to be
taken with care. Also the studies themselves can lead to mixed findings, depending
on the way these were designed. In order to really allow drawing such normative
conclusions from behavioural sciences to regulation, a standardised methodology is
needed, which can be applied across all sectors in jurisprudence.84
The implementation of the insight that pictorial claims matter needs to be han-
dled with care also from a different perspective: Visual elements on pack provide
information, but their meaning is contextually situated.85 This implies that regula-
tions cannot be targeted towards individual visual elements without taking the con-
text in which these appear into account. For instance, a heart symbol on a package
may communicate heart health (when used on a margarine package with a choles-
terol claim) or it may communicate love (when used on a package of chocolates for
Valentine’s day). Black or white lists of ‘health claim’ pictures or images such as
the ones we know from the ANNEX of the Regulation would be of little help. Here,
one might think about the lessons learned from how visual elements are protected
as a trade mark or design.
Concluding we can see from behavioural studies that, although this is not easy to
address, more attention to visual elements is vital for consumer protection in the
health claims regime. Drawing normative conclusions beyond that statement from
the evidence we provided shall be handled with care, as each of the studies did not
follow a standardized methodology. Furthermore, the effect of pictorial claims var-
ies regarding the context it is used it.

Recital 1 Regulation (EC) No 1924/2006.
Rott 2015, pp. 163–167; Sibony pp. 71–107.
Leathwood, Richardson, Strater, Todd, and Van Trijp, pp. 474–484.; Grunert, Scholderer, and
Rogeaux, pp. 269–277.
Scott 1994, pp 252–273.
11 The Potential Use of Visual Packaging Elements as Nudges 213

Acknowledgements We are grateful for the comments we received in the context of the Lucerne
conference. We would also like to thank Harry Bremmers for extensive comments on our paper.
Mistakes are our own.


Andrews, J. Craig, Richard G. Netemeyer, and Scot Burton. 1998. Consumer generalization of
nutrient content claims in advertising. Journal of Marketing 62(4): 62–75. doi:10.2307/1252287.
Barends, René. 2009. Constitutional horse trading: Some comments on the protocol on the internal
market and competition. In Views of European Law from the Mountain: Liber amicorum Piet
Jan Slot, ed. Mielle Bulterman, Leigh Hancher, Alison McDonnell, and Hanna Sevenster,
123–133. Alphen aan den Rijn: Wolters Kluwer.
Bilman, Els M., Ellen van Kleef, David J. Mela, Toine Hulshof, and Hans C.M. van Trijp. 2012.
Consumer understanding, interpretation and perceived levels of personal responsibility in rela-
tion to satiety-related claims. Appetite 59(3): 912–920. doi:10.1016/j.appet.2012.07.010.
Bone, Paula F., and Karen R. France. 2001. Package graphics and consumer product beliefs.
Journal of Business and Psychology 15(3): 467–489. doi:10.1023/A:1007826818206.
Bremmers, Harry, Bernd van der Meulen, and Kai Purnhagen. 2013. Multi-stakeholder responses
to the European health claims requirements. Journal on Chain and Network Science 13(2):
161–172. doi:10.3920/JCNS2013.1006.
Bushman, Brad J. 1998. Effects of warning and information labels on consumption of full-fat,
reduced-fat, and no-fat products. Journal of Applied Psychology 83(1): 97–101.
Carillo, Elizabeth, Paula Varela, and Susana Fiszman. 2012. Effects of food package information
and sensory characteristics on the perception of healthiness and the acceptability of enriched
biscuits. Food Research International 48(1): 209–216. doi:10.1016/j.foodres.2012.03.016.
Chandon, Pierre. 2013. How package design and package-based marketing claims lead to overeat-
ing. Applied Economic Perspectives and Policy 35(1): 1–6. doi:10.1093/aepp/pps028.
Chandon, Pierre, and Brian Wansink. 2012. Does food marketing need to make us fat? A review
and solutions. Nutrition Reviews 70(10): 571–593. doi:10.1111/j.1753-4887.2012.00518.x.
Childers, Terry L., and Michael J. Houston. 1984. Conditions for a picture-superiority effect on
consumer memory. Journal of Consumer Research 11(2): 643–654. doi:10.1086/209001.
Childers, Terry L., Michael J. Houston, and Susan E. Heckler. 1985. Measurement of individual
differences in visual versus verbal information processing. Journal of Consumer Research
12(2): 125–134. doi:10.1086/208501.
Clement, Jesper, Tore Kristensen, and Kjell Grønhaug. 2013. Understanding consumers’ in-store
visual perception: The influence of package design features on visual attention. Journal of
Retailing and Consumer Services 20(2): 234–239. doi:10.1016/j.jretconser.2013.01.003.
Cowburn, Gill, and Lynn Stockley. 2005. Consumer understanding and use of nutrition labelling:
A systematic review. Public Health Nutrition 8(1): 21–28. doi:10.1079/PHN2004666.
Duivenvoorde, Bram B. 2015. The consumer benchmarks in the unfair commercial practices directive.
Cham/Dordrecht/Heidelberg/New York/London: Springer. doi:10.1007/978-3-319-13924-1_2.
Franck, Jens-Uwe, and Kai Purnhagen. 2013. Homo economicus, behavioural sciences, and eco-
nomic regulation: On the concept of man in internal market regulation and its normative basis.
In Law and Economics in Europe, ed. Klaus Mathis, 329–365. Dordrecht: Springer.
Gardner, Meryl P., and Michael J. Houston. 1986. The effects of verbal and visual components of
retail communications. Journal of Retailing 62: 64–78.
Grice, Herbert P. 1975. Logic and conversation. In Syntax and semantics 3: Speech acts, ed. Peter
Cole and Jerry L. Morgan, 41–58. New York: Academic Press.
Grunert, Klaus G., and Josephine M. Wills. 2007. A review of European research on consumer
response to nutrition information at food labels. Journal of Public Health 15(5): 385–399.
214 K. Purnhagen et al.

Grunert, Klaus G., Josephine M. Wills, and Laura Fernández-Celemín. 2010a. Nutrition knowl-
edge, and use and understanding of nutrition information on food labels among consumers in
the UK. Appetite 55(2): 177–189. doi:10.1016/j.appet.2010.05.045.
Grunert, Klaus G., Laura Fernández-Celemín, Josephine M. Wills, Stefan Storcksdieck genannt
Bonsmann, and Liliya Nureeva. 2010b. Use and understanding of nutrition information on
food labels in six European countries. Journal of Public Health 18(3): 261–277. doi:10.1007/
Grunert, Klaus G., Joachim Scholderer, and Michel Rogeaux. 2011. Determinants of consumer
understanding of health claims. Appetite 56(2): 269–277. doi:10.1016/j.appet.2011.01.009.
Harris, Richard J. 1977. Comprehension of pragmatic implications in advertising. Journal of
Applied Psychology 62(5): 603–608. doi:10.1037/0021-9010.62.5.603.
Hastak, Manoj, and Michael B. Mazis. 2011. Deception by implication: A typology by truthful but
misleading advertising and labeling claims. Journal of Public Policy and Marketing 30(2):
157–167. doi:10.1509/jppm.30.2.157.
Hockley, William E. 2008. The picture superiority effect in associative recognition. Memory &
Cognition 36(7): 1351–1359. doi:10.3758/MC.36.7.1351.
Kardes, Frank R., Steven Posavac, and Maria L. Cronley. 2004. Consumer inference: A review of
processes, bases and judgment context. Journal of Consumer Psychology 14(3): 230–256.
Labrecque, Lauren I., and George R. Milne. 2013. To be or not to be different: Exploration of
norms and benefits of color differentiation in the marketplace. Marketing Letters 24(2): 165–
176. doi:10.1007/s11002-012-9210-5.
Lähteenmäki, Liisa. 2013. Claiming health in food products. Food Quality and Preference 27(2):
196–201. doi:10.1016/j.foodqual.2012.03.006.
Leathwood, Peter D., David P. Richardson, Peter Strater, Peter M. Todd, and Hans C.M. van Trijp.
2007. Consumer understanding of nutrition and health claims: Sources of evidence. British
Journal of Nutrition 98(3): 474–484. doi:10.1017/S000711450778697X.
Malam, Sally, Sue, Clegg, Sarah, Kirwan, Stephan, McGinigal in association with Monique, Raats,
Richard, Shephaerd, Julie, Barnett, Victoria, Senior, Charo, Hodgkins, and Moira, Dean. 2009.
Comprehension and use of UK nutrition signpost labelling schemes. British Market Research
Bureau. Prepared for Food Standards Agency. At
uk/20131104005023/ Accessed 16
Jan 2015.
Meyers-levy, Joan, and Laura Peracchio. 1995. Understanding the effects of color: How the cor-
respondence between available and required resources affects attitudes. Journal of Consumer
Research 22(2): 121–138. doi:10.1086/209440.
Milosavljevic, Milica M., Vidya Navalpakkam, Christof Koch, and Antonio Rangel. 2012. Relative
visual saliency differences induce sizable bias in consumer choice. Journal of Consumer
Psychology 22(1): 67–74. doi:10.1016/j.jcps.2011.10.002.
Nisbett, Richard E., and Timothy D. Wilson. 1977. The halo effect: Evidence for unconscious
alteration of judgements. Journal of Personality and Social Psychology 35(4): 250–256.
Nocella, Giuseppe, and Orla Kennedy. 2012. Food health claims: What consumers understand.
Food Policy 37(5): 571–580. doi:10.1016/j.foodpol.2012.06.001.
Otterbring, Tobias, Poja Shams, Erik Wästlund, and Anders Gustafsson. 2013. Left isn’t always
right: Placement of pictorial and textual package elements. British Food Journal 115(8): 1211–
1225. doi:10.1108/BFJ-08-2011-0208.
Parsons, Andrew G., and Christoph Schumacher. 2010. Advertising regulation and market drivers.
European Journal of Marketing 46(11/12): 1539–1558. doi:10.1108/03090561211259970.
Pieters, Rik, and Luk Warlop. 1999. Visual attention during brand choice: The impact of time pres-
sure and task motivation. International Journal of Research in Marketing 16(1): 1–16.
11 The Potential Use of Visual Packaging Elements as Nudges 215

Porter, Michael E. 1991. America’s green strategy. Scientific American 264(4): 96–101.
Porter, Elisabeth. 2014. Taking images seriously. Columbia Law Review 114(7): 1687–1782.
Purnhagen, Kai. 2013. Beyond threats to health: May consumers’ interests in safety trump funda-
mental freedoms on information on foodstuffs? European Law Review 38: 711–719.
Purnhagen, Kai. 2014. The virtue of cassis de Dijon 25 years later – It is not dead, it just smells
funny. In Varieties of European economic law and regulation Liber Amicorum Hans Micklitz,
ed. Kai Purnhagen and Peter Rott, 315–342. Dordrecht: Springer.
Purnhagen, Kai. 2015. The EU’s precautionary principle in food law is an information tool!
European Business Law Review, 26: 903–921.
Purnhagen, Kai, and Erica, van Herpen. 2014. Can bonus packs mislead consumers? An empirical
assessment of the ECJ’s mars judgment and its potential impact on EU marketing regulation.
Wageningen Working Papers Series in Law and Governance 2014/07, Wageningen University.
Raynor, Michael, Annette Boaz, and Cathy Higginson. 2001. Consumer use of health-related
endorsements on food labels in the United Kingdom and Australia. Journal of Nutrition
Education 33(1): 24–30. doi:10.1016/S1499-4046(06)60006-7.
Riley, Alan. 2007. The EU reform treaty and the competition protocol: Undermining EC competi-
tion law. CEPS Policy Brief 142: 703–754.
Roe, Brian, Alan S. Levy, and Brenda M. Derby. 1999. The impact of health claims on consumer
search and product evaluation outcomes: Results from FDA experimental data. Journal of
Public Policy & Marketing 18(1): 89–105.
Rott, Peter. 2015. Der ‘Durchschnittsverbraucher’ – ein Auslaufmodell angesichts zielgerichteten
Marketings?. Verbraucher und Recht: 163–167.
Schor, Danielle, Shelley Maniscalco, Michele M. Tuttle, Sarah Alligood, and Wendy R. Kapsak.
2010. Nutrition facts you can’t miss: The evolution of front-of-pack labelling. Providing con-
sumers with tools to help select foods and beverages to encourage more healthful diets.
Nutrition Today 45(1): 22–32. doi:10.1097/NT.0b013e3181cb4561.
Schuldt, Jonathon P. 2013. Does green mean healthy? Nutrition label color affects perceptions of
healthfulness. Health Communication 28: 814–821. doi:10.1080/10410236.2012.725270.
Scott, Linda M. 1994. Images in advertising: The need for a theory of visual rhetoric. Journal of
Consumer Research 21(2): 252–273. doi:10.1086/209396.
Scott, Linda M., and Patrick Vargas. 2007. Writing with pictures: Toward a unifying theory of
consumer response to images. Journal of Consumer Research 34(3): 341–356.
Sibony, Anne-Lise. 2015. Can EU consumer law benefit from behavioural insights?: An analysis
of the unfair practices directive’. In European perspectives on behavioural law and economics,
ed. Klaus Mathis, 71–106. Cham/Dordrecht/Heidelberg/New York/London: Springer.
Steindorff, Ernst. 1996. EG-Vertrag und Privatrecht. Baden-Baden: Nomos.
Stenberg, Georg, Karl Radeborg, and Leif R. Hedman. 1995. The picture superiority effect in a
cross-modality recognition task. Memory & Cognition 23: 425–441. doi:10.3758/BF03197244.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge. London: Penguin books.
Trstenjak, Verica, and Erwin Beysen. 2013. The growing overlap of fundamental freedoms and
fundamental rights in the case law of the CJEU. European Law Review 3: 293–315.
Trzaskowki, Jan. 2011. Behavioural economics, neuroscience, and the unfair commercial practices
directive. Journal of Consumer Policy 34(3): 377–392. doi:10.1007/s10603-011-9169-2.
Unberath, Hannes, and Angus Johnston. 2007. The double-headed approach of the ECJ concerning
consumer protection. Common Market Law Review 44(5): 1237–1284.
Underwood, Robert L., and Noreen M. Klein. 2002. Packaging as brand communication: Effects
of product pictures on consumer responses to the package and brand. Journal of Marketing
Theory and Practice 10(4): 58–69.
216 K. Purnhagen et al.

Usher, John A. 2001. Disclosure rules (information) as a primary tool in the doctrine on measures
having an equivalent effect. In Party autonomy and the role of information in the internal mar-
ket, ed. Stefan Grundmann, Wolfgang Kerber, and Stephen Weatherill, 151–161. Berlin: de
van Herpen, Erica, and Rik Pieters. 2007. Anticipated identification costs: Improving assortment
evaluation by diagnostic attributes. International Journal of Research in Marketing 24(1):
77–88. doi:10.1016/j.ijresmar.2006.09.005.
van Herpen, Erica, and Hans C.M. van Trijp. 2011. Front-of-pack nutrition labels. Their effect on
attention and choices when consumers have varying goals and time constraints. Appetite 57(1):
148–160. doi:10.1016/j.appet.2011.04.011.
van Herpen, Erica, Ellen Seiss, and Hans C.M. van Trijp. 2012. The role of familiarity in front-of-
pack label evaluation and use: A comparison between the United Kingdom and The Netherlands.
Food Quality and Preference 26(1): 22–34. doi:10.1016/j.foodqual.2012.03.003.
Van Kleef, Ellen, and Hans Dagevos. 2015. The growing role of front-of-pack nutrition profile
labelling: A consumer perspective on key issues and controversies. Critical Reviews in Food
Science and Nutrition 55(3): 291–303. doi:10.1080/10408398.2011.653018.
Van Trijp, Hans. 2009. Consumer understanding and nutritional communication. Key issues in the
context of the new EU regulation. European Journal of Nutrition 48(1): 41–48. doi:10.1007/
Van Trijp, Hans, and Ivo A. van der Lans. 2007. Consumer perceptions of nutrition and health
claims. Appetite 48(3): 305–324. doi:10.1016/j.appet.2006.09.011.
Viswanathan, Madhubalan, Manoj Hastak, and Roland Gau. 2009. Understanding and facilitating
the usage of nutritional labels by low-literate consumers. Journal of Public Policy & Marketing
28(2): 135–145. doi:10.1509/jppm.28.2.135.
Vyth, Ellis L., Ingrid H.M. Steenhuis, Jessica A. Vlot, Anouk Wulp, Meefa G. Hogenes, Danielle
H. Looije, Johannes Brug, and Jacob C. Seidell. 2010. Actual use of a front-of-pack nutrition
logo in the supermarket: Consumers’ motives in food choice. Public Health Nutrition 13(11):
1882–1889. doi:10.1017/S1368980010000637.
Warlop, Luk, S. Ratneshwar, and Stijn M.J. van Osselaer. 2005. Distinctive brand cues and mem-
ory for product consumption experiences. International Journal of Research in Marketing
22(1): 27–44. doi:10.1016/j.ijresmar.2004.02.001.
Weatherill, Stephen. 2014. Why there is no principle of mutual recognition in EU law (and why
that matters to consumer lawyers). In Varieties of European economic law and regulation, ed.
Kai Purnhagen and Peter Rott, 415. Dordrecht: Springer.
Williams, Peter. 2005. Consumer understanding and use of health claims for foods. Nutrition
Reviews 63(7): 256–264. doi:10.1111/j.1753-4887.2005.tb00382.x.
Chapter 12
Nudging: Information, Choice Architecture
and Beyond
Theory and Applications in Financial Markets Law

Rainer Baisch

We would sign our lives away without reading,

so people need protection from a lot that is legal.
(Kahneman 2013, p. 1340.)

Abstract The traditional disclosure-paradigm is based on the assumption that

transparent and effectively processed information will enable the investor to make
well-founded investment decisions. Having the type of a «homo oeconomicus» in
mind, financial market laws used to be designed for responsible and knowledgeable
actors. Analysing human flaws regarding the way we process information in the
light of behavioural findings, the provision of data for consumers of financial prod-
ucts should be further optimised. Despite the information overload the provision of
information must not be condemned; this would mean throwing the baby out with
the bath water. Disclosure can be used and function’s also as a nudge; thus the way
information is framed matters a lot. However, inundating customers with too much
information does not help. Valueing the positive outcome of simple rules of thumb
might complement to an optimised cognition and perception of human behaviour as
it should be imputed to individuals when designing regulation. Enhanced preven-
tion based on smarter information requirements will give consumers a better chance
to avoid human flaws. Additionally, education could, at least partially, compensate
negative outcomes of human flaws. However, nudging should avoid weakening the
motivation of individuals to actively inform themselves. Combined with a strict
Code of Conduct at the point of sale, a Suitability Test including the real risk percep-
tions of the consumer leads to an enhanced level of investor protection. Nudging in
combination with some more paternalistic elements can de-bias detrimental human
shortcomings while distending paternalism should be avoided.

R. Baisch (*)
Rechtswissenschaftliche Fakultät, University of Zurich,
Rämistrasse, 8001 Zurich, Switzerland

© Springer International Publishing Switzerland 2016 217

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_12
218 R. Baisch

12.1 Introduction

12.1.1 Point of Departure

Neo-classical economic theory advocates investor protection aside from market

abuse only if market failures – often based on information asymmetries – occur.
However, additionally there are (i) various behavioural biases that influence the
financial decisions of individuals, combined with (ii) individual’s disability to man-
age information properly and (iii) the latent information overload; those facts call
for an appropriate regulatory facet.1 On top of that, (iv) financial intermediaries
might be tempted to use common biases to sell products which are not in the best
interest of the client.2
Financial innovation led to “an ever-widening set of financial options from an
expanding set of firms and accompanied by a sometimes dizzying amount of
information”.3 While the complexity of the supply side confuses the consumer,4
the demand side is more and more becoming a ‘do-it-yourself’ activity.5 Many
consumers are not aware of the fact that their performance regarding investment
decisions is rather poor. Additionally, the supply side capitalises on human limi-
tations by taking advantage of the way we process information.6 The question
why consumers of financial services products often make relatively poor deci-
sions has been analysed by behavioural scientists; they blame cognitive short-
comings based on the use of heuristics or various subconscious biases.7 Accepting
this cognitions as being true and having a negative impact on investment deci-
sions (or at least, that there is a certain, latent and unwished menace for consum-
ers) it should be discussed how that issue can be adequately addressed. Such a
solution could also benefit from the concept of nudging as introduced by Thaler
and Sunstein.8

For a (critical) overview of the emergence of behavioural law and economics as a counterpart to
orthodox law and economics with reference to the main literature see Mitchell 2014, pp. 167–168.
See also Langevoort 1992 for a detailed analysis regarding the efficient market hypotheses and the
influence of behavioural research.
The paper carries forward and enhances findings that were presented at the 3rd Law and
Economics Conference in Lucerne: Behavioural Law and Economics in April 2014; Baisch and
Weber 2015.
Campbell et al. 2011, p. 91.
Libertarians prefer to talk about “investors”; the term “consumer” suggests and implies the need
of a certain degree of protection. See Kingsford Smith 2010; Moloney 2012, and Sarra 2013.
Ryan et al. 2011, p. 462.
Akerlof and Shiller 2015.
See also World Bank Group, World Development Report 2015, Mind, Society and Behavior,
December 2015.
Thaler and Sunstein 2008; Sunstein 2014a.
12 Nudging: Information, Choice Architecture and Beyond 219

Looking at retail investors the regulator must abandon the idea of decision-
making based only on rational choice assumptions. Therefore, disclosure is neces-
sary but will alone not lead to optimal investment allocation.9 In reality, behaviourally
impairments influencing investment decisions are accompanied by herding, bubble-
building and wrongly incentivised staff at the point-of-sale (POS); therefore, any
kind of information provision as well as rules for a Code of Conduct must accept
findings from behavioural research. Keeping in mind that the professionals from the
financial industry are also aware of these effects, it cannot be ignored that those
biases might be capitalised. By looking at product flyers with attractive recent return
curves or highlighted assurance that the capital is protected, it is not farfetched to
assume that at least some behavioural weaknesses are exploited (framing). Based on
this premise, a closer look at the POS-regulation is inevitable.

12.1.2 Regulatory Activities

In recent years the findings of behavioural finance research10 have influenced the
traditional understanding of the processes and patterns that allow the functioning of
financial markets. In the meantime regulatory perceptions which used to be based
on the assumptions of rational behaviour partially allow for human flaws.11 For
instance, the new PRIIP regulation12 defining the KID13 which is going to “be fur-
ther harmonised through Level 2 instruments that take into account existing and
on-going research on consumer behaviour”.14

Ulen 2013 delivers a good overview of behaviourally informed investor protection regulation and
also advocates augmented disclosure obligations.
A good collection of early key papers can be found in Kahneman et al. 982; see also Kahneman
and Tversky 2000; legal implications stemming from behavioural findings are described by
Korobkin and Ulen 2000. See also Moloney 2010, 67–74, regarding the irrational and uninformed
Cunningham 2002, p. 836, resumes that behavioural finance is “a marriage of cognitive psychol-
ogy and the financial economics of market inefficiency”. See FCA papers: Kristine Erta et al.,
Applying Behavioural Economics at the Financial Conduct Authority, Financial Conduct Authority,
Occasional Paper No. 1, April 2013,
occasional-paper-1.pdf; Paul Adams/Stefan Hunt, Encouraging Consumers to Claim Redress:
Evidence from a Field Trial, Financial Conduct Authority, Occasional Paper No. 2, April 2013,
Proposal for a Regulation of the European Parliament and of the Council on key information
documents for packaged retail and insurance based investment products, PRI(I)Ps; http://register.
Key Investor Document.
Discussion Paper of the Joint Committee of European Supervisory Authorities (the ESAs: EBA
European Banking Authority, EIOPA European Insurance and Occupational Pensions Authority,
and ESMA European Securities and Markets Authority), Key Information Documents for Packaged
Retail and Insurance-based Investment Products (PRIIPs), 17 November 2014, p. 17.
220 R. Baisch

The assessment of the prospect, options, scope and effectiveness of a regulatory

approach based on the outcome of behavioural research in the area of financial mar-
ket law through techniques like nudging is a growing field of research.15 The fact
that some actors in the financial industry have shown exploitative activities in their
interaction with customers of financial services products is another reason why
regulators see themselves forced to provide preventive measures based on behav-
ioural research.16 Market behaviour that cannot be subsumed under market abuse or
fraud but which still has to be regarded as cheating the customers because it capital-
ises on human flaws is an issue on the regulative agenda.17
Opponents of this change lament a loss of freedom accusing such kind of policy
and regulation as too paternalistic.18 However, in principle, “nudges are in many
cases not paternalistic at all, but instead largely cases of rational persuasion”.19
Admittedly, already some years ago behavioural economics were rightly considered
as a field that naturally has its limits20:
As policymakers use it to devise programs, it’s becoming clear that behavioural economics
is being asked to solve problems it wasn’t meant to address. Indeed, it seems in some cases
that behavioural economics is being used as a political expedient, allowing policymakers to
avoid painful but more effective solutions rooted in traditional economics.

In the U.S. the Consumer Financial Protection Bureau is authorized to ensure

that “consumers are provided with timely and understandable information to make
responsible decisions about financial transactions”.21 More precisely, the bureau
may prescribe rules to ensure that the features of any consumer financial product or ser-
vice, both initially and over the term of the product or service, are fully, accurately, and
effectively disclosed to consumers in a manner that permits consumers to understand the

A good overview is delivered by Reiss 2013, pp. 278–302. See also Bovens 2009.
However, it is also possible that “sometimes firms might not know that their customers are mak-
ing mistakes. What looks like deliberate exploitation may actually just be firms responding to
observed consumer demand without realising that it is driven by biases.” See FCA, Occasional
Paper No.1, Kristine Erta, Stefan Hunt, Zanna Iscenko and Will Brambley, Applying behavioural
economics at the Financial Conduct Authority, April 2013, 7 and 21;
Add-on-insurance sale; see FCFCA, Market Study, General insurance add-ons: Provisional find-
ings of market study and proposed remedies, March 2014;
Some authors criticise the approach from Thaler and Sunstein; e.g. White 2013, xiii, states that
policies based on nudges are “ineffective, unethical, and dangerous to individual choice, interests,
and autonomy”. See also Whitman and Rizzo 2007, Rebonato 2012. Sunstein 2005, p. 201 seq.,
already addresses some major lines of criticism: (i) slippery slope (resist the beginnings – nip it in
the bud), (ii) mistrust in the abilities of the planner or choice architect and (iii) enthusiastic pater-
nalists dislike the non-mandatory nature of nudging. See also Conly 2013.
Hausman and Welch 2010, p. 136.
George Loewenstein and Peter Ubel, Economics Behaving Badly, The New York Times, July 14,
Dodd-Frank Act, 2010; 12 U.S. Code § 5511 – Purpose, objectives, and functions.
12 Nudging: Information, Choice Architecture and Beyond 221

costs, benefits, and risks associated with the product or service, in light of the facts and

In Europe, the UK installed a similar consulting team and especially the FCA
(see later Sect. 12.4.2) avows oneself to consider behavioural findings; recently
Germany also takes similar steps in this direction.23

12.2 Regulatory Design

Nudges are ways to make it easy for people (…)

compatible with their values and (…) with their desires.24

12.2.1 Paternalism

Paternalism25 is widely accepted to protect lives (mandatory use of seat-belt) or

minors (sale of alcohol). However, a latent phobia can be observed amongst market
liberals being afraid of too much paternalism with respect to the outcomes of behav-
ioural research. According to common understanding, paternalism limits the liberty
or autonomy of individuals but does so for that person’s own good – based on its
Latin origin – in a fatherly manner but restricting individual freedom and responsi-
bilities; however, the addressee is not a child anymore.26 The assumption that a
father is normally acting in the best interest of his child might not be true in the
regulatory world where no family ties come into play; in fact there is no guarantee
that policy makers are acting altruistic and/or in the best interest of consumers –
whether by accident or design, unconsciously or deliberately. In addition, initially
as a precondition the insight and comprehension have to be acquiesced, we as indi-
viduals need such kind of assistance. Polemically phrased an important cognition
based on behavioural research is a widening of the category of idiots requiring
protection. To avoid a proliferation of paternalism, good regulation will impose
minimal restrictions on those who behave rationally, but will guide the others.27

Dodd-Frank Act, 2010; 12 U.S. Code § 5532 – Disclosures.
Similar to the U.S. (Social and Behavioral Sciences Team) and Great Britain’ Nudge-Unit
(Behavioural Insights Team) also the German Chancellery is setting up an interdisciplinary team
to apply empirical evidence from behavioural economics in public policy making. See Steinbeis,
Maximilian: ‘Nudging’ arrives in Germany, VerfBlog, 27 August 2014, http://www.verfassungs-
Kahneman 2013, p. 1340.
See Dworkin 2013.
Sunstein 2014c addresses the ethical aspects of nudging and related criticism. Trout 2005, 408,
argues that biases damage peoples autonomy not institutional remedies.
Camerer et al. 2003, p. 1218.
222 R. Baisch

Thus, paternalistic regulation will not be in conflict with constitutional limits like
the right to freedom of action and self-determination in a liberal state.28
One of the elementary topics regarding financial market regulation culminates in
the question as to which degree paternalistic emphasis would be inevitable.29 Is
there just the necessity to have criminal law against fraud in place? What kind of
private law design is best suited for settlement and compensation? Are people enti-
tled to make their own choices despite the fact that they err and, thus, might influ-
ence the general welfare in a negative way? Therefore, in some areas, remedies
representing a form of regulatory paternalism to protect customers from their own
mistakes appear to be the better alternative avoiding that households become too
dependent on their own devices and, in consequence, have to learn from their own
mistakes. Anyhow, unfortunately learning effects might come into fruition only
very rarely for many consumers, since financial decisions are undertaken seldom
and their outcomes are delayed, sometimes for decades (endowment insurance).
Regarding long-term investments, consumers might not even realise how the invest-
ment is performing or whether they have been mis-sold a financial product until
many years have been passed. Anyhow, consumers often will not be able to learn
from experience given the infrequency with which they often act as investors.30
Due to other traditions in Europe it might not be necessary to add euphemistic
attributes ending up with Libertarian or Asymmetric Paternalism. Because any kind
of limitation regarding choices or restriction of alternatives interfering with some-
one’s liberty must be regarded as paternalistic, such kind of regulation also should
be named paternalistic31; without any negative undertone however, public and man-
datory pension insurance systems which are common in Europe are clearly pater-
nalistic like all other compulsory insurances. Any kind of tax deduction incentives
or tax concessions are at least a soft paternalistic nudge. According to Thaler and
Sunstein32 paternalism does not always have to involve coercion and could also be
based on coaxing reactions out of individuals. For them, even the sequence in which
food is presented in a cafeteria could be (softly) paternalistic.

Van Aaken 2016, explores the legal limits of paternalistic nudging under the German Constitution
and discusses the (typical German) proportionality assessment which consists of a quadrinominal
test: (i) legitimate aim, (ii) suitability, (iii) neccessity (least restrictive means tests) and (iv) propor-
tionality. It is important to distinguish whether a nudge targets preferences and autonomy (end
paternalism) or just cognition (means paternalism); while the first needs to be subjected to greater
scrutiny the latter is true for informational nudges.
Libertarians prefer the freedom of choice, while condemning paternalism; see e.g. Boaz 1997; on
the other side, paternalists mistrust freedom of choice and argue against some occurrences of lib-
ertarianism; see e.g. Goodin 1991.
See ESA-DP (fn. 15), p. 17.
Blumenthal-Barby 2013.
Thaler and Sunstein 2008, p. 11
12 Nudging: Information, Choice Architecture and Beyond 223

Fig. 12.1 Matrix classifying nudges (Mullane and Sheffrin 2012)

12.2.2 Nudging

Thaler and Sunstein developed their concept of nudging33 which has its roots in
their understanding of Libertarian Paternalism.34 A nudge, as they use the term,
is any aspect of the choice architecture that alters people’s behaviour in a predictable way
without forbidding any options or significantly changing their economic incentives. To
count as a mere nudge, the intervention must be easy and cheap to avoid. Nudges are not
mandates. Putting fruit at eye level counts as a nudge. Banning junk food does not.35

Following this definition, it is not really convincing (or coercive) to call the con-
cept of nudging paternalistic because such kind of potentially helpful treatment or
interaction is not restricted only within the relationship to a father figure.
However, the right combination of regulative measures from paternalistic to lib-
ertarian is a challenging task. It should not be questioned that nudges may be benefi-
cial in many regulative areas. But there are important challenges regarding the best
way of implementation. Therefore, nudges belong in the “toolkit for regulation, but
with a considered humility. They are not a panacea.”36 Besides disclosure as a
nudge, nudges like default rules (e.g. opting in/out as professional or retail client)
could work well in financial regulation.
Nudging should not happen in secrecy. Figure 12.1 also identifies non-transparent
types of nudging. Such interventions should come with some kind of information;

Thaler and Sunstein 2008.
Sunstein and Thaler 2003.
Thaler and Sunstein 2008, 6. Regarding choice architecture see also Thaler et al. 2013; Johnson
et al. 2012.
According to Hansen and Jespersen 2013, pp. 20–23; based on the dual process theory which
underpins Thaler and Sunstein 2008, pp. 19–22, and is also fundamental to Kahneman 2011.
224 R. Baisch

e.g. a rearranged cafeteria could be tagged with a respective indication somewhere

at the entrance.
Looking to other fields of regulation, e.g. food labelling obligations, it must be
conceded that there is not a real effect against obesity. There might be a need to
consider taxes on unhealthy food, but instead of changing the price of junk food, the
focus should be on consumer behaviour. Are there similar facts and implications
regarding financial regulation? Because there are (i) some doubts whether nudging
consumers is the best way to improve their situation, and there is also (ii) a call for
more traditional legal and economic solutions, this topic generates interesting
research questions.

12.2.3 Example: Product Ban vs. Product Warning

To highlight the difference in the regulatory approach based on paternalism from

the point of view of nudging, the following example shall help to clarify the under-
standing as adapted in this paper. In the first use of its new consumer protection
powers, the UK’s Financial Conduct Authority (FCA) restricted firms from distrib-
uting contingent convertible securities (CoCos) to the mass retail market.37 However,
based on the UK regulation professional, high-net-worth and sophisticated investors
as well as individual investors via execution-only dealing platforms are still able to
buy CoCos.38
Given such kind of exceptions, would it not be better just to make assertive warn-
ings compulsory?39 A ban might be too rigid, while product warnings could be suf-
ficient. On the other hand, it is also still permitted to buy shares and plain bonds of
financial institutions. The problem is that with CoCos, investors do not have control
over if and when this type of bond converts to an equity; however, in comparison to
a share, an interest is paid and the share-value will fall similarly in the case of a trig-
ger event. Mandatory suitability tests and restrictions to allow the distribution of

FCA Press release, 5 August 2014,
cocos-to-retail-investors. CoCos are hybrid bonds that currently offer attractive yields but their
value can be written down or converted into equity if the issuing bank’s capital drops below a
certain minimum level.
CoCos may still be sold to investors for whom they are more likely to be suitable, including
professional investors or retail investors who are high net worth or sophisticated. Sophisticated
investors are defined by the FCA as retail clients with extensive investment experience and knowl-
edge of complex instruments, able to understand and evaluate the risks and potential rewards of
unusual, complex and/or illiquid investments. To be classed as high-net-worth, investors need to
have an annual income of more than £100,000 or have investable net assets of more than £250,000;
see Temporary product intervention rules, Restrictions in relation to the retail distribution of con-
tingent convertible instruments, August 2014,
FCA started a consultation in October 2014 (CP14/23: Restrictions on the retail distribution of
regulatory capital instruments) and published rules in June 2015 permanent rules are in place since
1 October 2015 when the temporary product intervention rules for CoCos expired (PS15/14).
12 Nudging: Information, Choice Architecture and Beyond 225

CoCos only to customers who use advisory or portfolio management services are
the better option. Nudging by means of information and warnings are preferable.

12.3 Behavioural Insights

Disclosure requirements should be designed for homo sapiens,

not homo economicus.40

12.3.1 Information: Capacitance, Processing, and Evaluation

Based on the research from behavioural economics on cognitive failings the artifi-
cial construct of a “reasonable investor” – resulting in a lack of protection for inves-
tors – is at stake. Today, even in countries with a tradition of reliable social systems,
unsophisticated retail investors are forced to invest in the market to save for their
retirement; the long phase and partially ongoing zero interest-rate policy in impor-
tant economic regions led to an additionally reinforcing effect. Consumers have to
care for their retirement despite their cognitive failings and their lack of training for
the task; therefore, regulators have to widen their perspective.41 Obviously, in
Europe similar to the U.S. “disclosure is not the panacea that drafters of federal
securities laws may have thought it to be”.42 Additionally, it is not really question-
able that nudges may be beneficial in many regulative areas. And there are impor-
tant challenges regarding the best way of implementation.
One of the key pillars undergirding the federal securities laws – that investors effectively
process information – is under pressure. An extensive literature shows that investors and
other securities market participants are subject to cognitive biases and, because of bounded
rationality, adopt heuristics in making investment decisions.43

The traditional disclosure-paradigm is based on the assumption that transparent

and effectively processed information will enable the investor to make well-founded
investment decisions. Having the type of a «homo oeconomicus» in mind, financial
market laws used to be designed for responsible and knowledgeable actors. After
the financial crisis culminating in the collapse of Lehman Brothers we should not
condemn information; this would mean throwing the baby out with the bath water.
Failures resulting from consumer biases or suboptimal information lead to similar
regulatory challenges, like shown in the regulatory framework in Fig. 12.2.44

Sunstein 2013, p. 9.
See Davidoff and Hill 2013.
Black 2013, p. 1507.
Paredes 2003, p. 484.
OECD 2007, Roundtable of Economics for Consumer Policy: Summary Report, 26 July 2007,
p. 13.
226 R. Baisch

Fig. 12.2 OECD 2007, decision tree

Disclosure can function also as a nudge.45 However, customers should not be

inundated in too much information. It can be criticised that the design of legal mea-
sures for transparency does not sufficiently take into consideration whether an
investor is capable and (under the behavioural aspects) willing to assimilate the
information made available to him. Therefore, it must be assumed that the provision
of information alone is not sufficient, since the capacity and willingness to handle
information might be limited. Swamping disclosure requirements used to misjudge
the abilities of retail investors. The new European regulation leads to the mandatory
provision of basic leaflets. It remains to be seen whether such kind of information is
properly processed by the consumers. Behavioural finance exemplifies that many
biases cannot be prevented through disclosure. “Disclosure has considerable attrac-
tions as a retail market tool, but the challenge for regulators is to resist the tempta-
tion to make disclosure the panacea for investor protection.”46 What is the target of
regulation, the empowered investor or the irrational and uninformed investor? The
fact that even experienced investors are prone to irrational decisions should deliver
some food for thought.

Sunstein 2014b, p. 727.
Opening statement of Steven Maijoor, Chair of ESMA, at the ESMA Investor Day, 12 December
2012, Paris;
12 Nudging: Information, Choice Architecture and Beyond 227

12.3.2 Inertia, Overconfidence and Emotions

Not only information but other factors such as inertia, overconfidence and emotions
substantially influence decisions and whole markets.47 Based on emotions, humans
do not like to regret a former decision, overconfidence48 influences the assessment
of probabilities and due to inertia49 decisions are refused. Influenced by reference
biases and the manner how information is framed and anchored, non-rational fac-
tors are having a major impact on the way information is processed. The interrela-
tion of inertia50 and the regret aversion bias51 hinders investors to act. In consequence
goods (or investments) which are already in someone’s possession (Thaler worded
it “included in the individual’s endowment”) are valued higher than those for which
money would have to be spent.52
Sunstein identifies four categories (Fig. 12.3) in which central findings with par-
ticular importance to regulatory policy of recent social science research can be
related to for purposes of regulation.53
1. Default rules, rules that determine the result if people are too lazy or even refuse
to make a choice at all, are quite common to encounter inertia. Especially, in the
domain of retirement savings default rules can be extremely important to over-
come the power of inertia.54 Default rules can balance the lack of self-control.
2. Widely recognised is the finding that people can be influenced by how informa-
tion is presented or framed.55 Emotions instead of rational behaviour come into
play. For any kind of information, salience greatly matters. If key facts like fees

The role of psychology in economics was already discussed by Keynes 1936, p. 315, referring to
animal spirits (162) and unrealistic optimism or pessimism as causes for booms and busts.
Overconfidence could have many positive effects. Looking at retirement savings the expectation
to live longer than average should motivate individuals to higher savings. Baisch and Weber 2015,
pp. 167–171, with further references.
The debate regarding akrasia (the Greek ἀκρασία means a lack of command over oneself) is
quite an old one: Socrates argued in Plato’s Protagoras (Plato, 380 B.C., Platon: Protagoras 345
d9-e4) that akrasia does not exist, because no one acts willingly against one’s better judgment.
Aristotle countered more empirically and acknowledged that individuals behave intuitively
(Aristoteles, The Nicomachean Ethics VII 3, 1145 b22–24).
Gal 2006.
Regret aversion bias is the tendency to avoid making a decision due to the fear of the later prob-
ability to experience the pain of regret; in consequence, people will avoid taking decisive actions
because they fear that, in hindsight, whatever action they may have selected could prove less than
Thaler 1980, p. 44.
Sunstein 2011, p. 1350.
Just one of many: Madrian and Shea 2001. See also EIOPA-report, Good practices on informa-
tion provision for DC schemes, Enabling occupational DC scheme members to plan for retirement,
24 January 2013.
Sunstein 2011, p. 1353.
228 R. Baisch

Fig. 12.3 Sunstein’s four categories

or risks are not sufficiently salient to people, this results in limited attention.56
Therefore, vivid displays are more effective than abstract presentations of statis-
tical risks.
3. The third category, social influences, is less relevant for financial market regula-
tion but could be seen as partially interesting in view of compliance to code-of-
conduct-regulations shown by advisors. Social norms can help to create a
phenomenon of ‘compliance without enforcement’ like wearing a helmet while
biking or skiing even if it is not mandatory to do so. A similar phenomenon is
that people are willing to cooperate with one another or to contribute to a solu-
tion, despite the fact that standard economic theory predicts other optimal
4. Finally, the optimism bias or overconfidence based on difficulties assessing
probabilities, combined with the ignorance regarding not known facts and other
influences like the confirmation bias must always kept in mind when it comes to
decision making.58

12.3.3 Intermediate Conclusion: Suitability as a Cornerstone

of Investor Protection

According to IOSCO, suitability requirements aim at ensuring that financial inter-

mediaries only make suitable recommendations and that customers have the neces-
sary expertise, knowledge and financial capacity to trade in the respective financial

The problem of shrouded attributes comes into play; Gabaix and Laibson 2006; see also Sunstein
2015, p. 316.
Sunstein 2011, p. 1357.
Sunstein 2011, pp. 1358 et seq.
12 Nudging: Information, Choice Architecture and Beyond 229

instruments and to understand associated risks given their investment objectives;

thereby the risk tolerance including the risk of loss of capital must also be
In Europe the introduction of suitability requirements in 2004 with MiFID60
changed the retail investor protection regime significantly, setting a new “high-
water mark of the extent to which the new regulatory strategy for the retail markets
intervenes paternalistically in the investor’s decision”.61 But obviously the financial
crises tsunami surpassed that high-water mark and the dam is now heightened not
only by strengthening the requirements regarding the investor’s individual risk-
tolerance within the suitability assessment. Today the mandatory assessment regard-
ing the individual appropriateness and suitability of financial investments at the
POS is widely accepted.62 In the U.S. with the standards set by the industry, suit-
ability requirements are established already a long time ago (since 2012, the FINRA
Rule 211163 replaces NASD Suitability Rule 2310).64
It was quite common to accept some of the behavioural findings, however, often
the traditional approach preferring education – often named investor empower-
ment – was followed instead of enhancing investor protection.65
Investor education rather than investor regulation is probably the best way to respond to the
increasing recognition of the substantial role that cognitive biases play in investor behav-
iour. Nevertheless, the insights from behavioural finance do suggest a couple of areas where
existing legal rules should change.66

IOSCO report, Suitability Requirements With Respect To the Distribution of Complex Financial
Products, January 2013, p. 6 et seq.
Market in Financial Instruments Directives, 2004/19EC of the European Parliament and of the
Council of 21 April 2004 on markets in financial instruments, 30.4.2004, OJ L 145/1-44 MiFID II
2014/65/EU of 15 May 2014 repealing Directive 2004/39/EC, 12.6.2014, OJ L 173/349–496.
Moloney 2008 (2nd ed.), 614; Moloney 2014 (3rd ed.) gives a good overview of the discussion
considering issuer-disclosure, the Efficient Market Hypothesis and behavioural insights, pp. 54–59.
Already Rapp 1998, p. 189, stated in relation to financial industry standards of conduct that none
is “more frequently cited, and least objectively applied, than the ‘suitability’ requirement”.
Rule 2310 is the former (since 1939) Article III, §2 of the NASD Rules of Fair Practice.
Trout 2005, p. 433, repudiates this approach: “Anyone who assumes the adequate efficiency of
debiasing through individual training is either ignoring the magnitude of institutional intervention
required for such educational programs, or ignoring the cognitive costs to the individual of cor-
recting such biases. Either way, these strategies of individual training do not take seriously the best
available science.”
Cunningham 2002, p. 797. At p. 788: “It also means that such investor education must include
not only tutelage in the principles of finance and insights from behavioral finance, but also expla-
nation of how these axioms may collide.”
230 R. Baisch

This statement reflects this cognition but the need for some rules to be adjusted
was already recognised with reference to NASD 2310.67
With the adoption of MiFID II in Europe, the potential of shortcomings which
ESMA has found regarding the implementation of the MiFID I suitability require-
ments is going to be reduced.68 Based on supervisory experience of the provisions
included in Article 35 of the MiFID Implementing Directive, ESMA suggested
expanding these provisions in a number of key areas reflecting the expectations on
financial intermediaries already communicated in the ESMA suitability guide-
lines.69 The financial intermediary does often evaluate in his own interest whether a
customer is able to bear the risk of an investment. Whether a certain consumer is
really aware of the risk of a particular investment or his portfolio and, in conse-
quence, actually is really willing to tolerate this risk, is much harder to assess. A
specific often missing level of financial literacy might be inevitable to allow for
such kind of evaluation.
For instance, ESMA makes demands on robust processes for assessing the risk a
client is willing and able to take (including their ability to bear the investment risk).
The suitability process should also respect that based on behavioural research
rational-choice-presumptions are to be challenged. The various biases regarding the
perception of information and the validating of facts influencing investment-
decisions must be taken into account: (i) many investors are not focused on maxi-
mising profit but looking for security and conservation of value; (ii) the past
performance orientated marketing instruments are leading to immoderate expecta-
tions. A responsible advisor will take care of these facts and not ignore his client’s
level of risk awareness to make sure that the potential losses will be tolerated.
But how regulation should look like in a clearly non-advised environment?
Should there be regulatory space allowing for a playground attracting gamblers and
speculators? Action against abusive behaviour is by far the better approach. Even in
a paternalistic world there should be scope for a niche of online-brokerage protected
by appropriate warning and caution symbols – instead of prohibitive signs and
bans – to avoid unintended actions of investors.

NASD (National Association of Securities Dealers) is the former name of the American self-
regulatory organization for broker-dealers, which is now consolidated with the member regulation,
enforcement and arbitration functions of the New York Stock Exchange FINRA (Financial Industry
Regulatory Authority). Founded and registered with the SEC in 1939 the NASD was based on the
1938 Maloney Act amendments to the Securities Exchange Act of 1934. NASD as the one and only
Registered Securities Association supervised the conduct of its members under the oversight of the
SEC. Section 15A of the Securities Exchange Act of 1934 was enacted in 1938 to extend regula-
tory authority under the federal securities laws to the over-the-counter securities markets. NASD
created a system of cooperative regulation of broker-dealers and was responsible for the adoption
and enforcement of the rules to protect investors and the public interest.
ESMA criticised: (i) the failure to ask clients the right questions; (ii) the failure to collect the
necessary and relevant information; (iii) the failure to interpret correctly the information provided
by the client; and (iv) even where the right information is collected, the failure to recommend a
suitable investment;
ESMA, consultation paper, May 2014, p. 132 et seq.;
12 Nudging: Information, Choice Architecture and Beyond 231

12.4 Examples of Changes in Financial Market Regulation

This implies that providing all relevant legal and

technical information is not sufficient by itself.70

12.4.1 European Regulation


The latest procreation of Europe’s financial markets policy makers is called PRIIP-
KID71 (Key Information Document) and intends to improve the quality and compa-
rability of information provided to retail investors in the EU.72 A mandatory
disclosure document, the ‘Key Information Document’, has to be supplied to retail
consumers before they take the decision to invest in a financial product. UCITS and
retail AIF73 that are required by national law to provide the UCITS-KIID
(Undertakings for Collective Investment in Transferable Securities, Key Investor
Information Document) are exempt from the new regulation during 5 years. The
intention is that then the PRIIP KID is being used for all types of products.
The rules require the financial services industry to provide basic information
about their investment products, the risk and return that can be expected as well as
the overall aggregate cost that will arise in making the investment. The EU proposal
is a regulation on key information documents for packaged retail and insurance-
based investment products (PRIP). The new regulation follows the introduction of a
standardised disclosure document for investors investing in UCITS funds.
The regulation states: “… retail investors have often made investments without
understanding the associated risks and costs and have, on occasion, suffered
unforeseen losses.”74 For products which are not simple or which are difficult to
understand, the KID must also include a comprehension alert which reads: “You are
about to purchase a product that is not simple and may be difficult to understand.”75

EIOPA (fn. 56), p. 7.
PRIIP: Packaged retail investment and insurance-based investment products. On 9 December
2014, the Regulation (EU) No 1286/2014 of the Europe an Parliament and of the Council of 26
November 2014 on key information documents for packaged retail and insurance-based invest-
ment products (PRIIPs) has been published in the Official Journal of the European Union.
The regulation covers structured deposits, structured products, insurance products with an invest-
ment element (but not pension products) and funds; Regulation (EU) No 1286/2014 of the
European Parliament and of the Council of 26 November 2014 on key information documents for
packaged retail and insurance-based investment products (PRIIPs), OJ L 352/1, 09.12.2014.
UCITS: Undertakings for Collective Investment in Transferable Securities; Directive 2001/107/
EC and 2001/108/EC; AIFMD: The Alternative Investment Fund Managers Directive 2011/61/EU.
PRIIPs regulation, pn. (1).
PRIIPs regulation, pn. (18): “A product should be regarded as not being simple and as being
difficult to understand in particular if it invests in underlying assets in which retail investors do not
commonly invest, if it uses a number of different mechanisms to calculate the final return of the
232 R. Baisch

Risk label

low A A
risk E

for a more detailed

explanation of this official
label, please go to

Fig. 12.4 Belgian risk label (Financial products marketed in Belgium to retail clients, as from 12
June 2015, have to be assigned a standardized risk label. The technical requirements for the label
are laid down in a Regulation by the Financial Services and Markets Authority (FSMA). Royal
Decree of 25 April 2014, published in the Belgian Official Gazette (Moniteur belge/Belgisch
Staatsblad) on 12 June 2014)

The KID provides on no more than three pages the nature of the product, the
risks including performance scenarios, all initial, ongoing and exit costs and certain
other information. ESA’s Discussion Paper (ESA-DP)76 highlights many issues that
need to be addressed in order to balance the desire for conciseness and simplicity
facing the complexity of investment product’s risks and costs. An important task is
to allow for comparability. Most of the document is dedicated to the especially
tricky areas of risk and return disclosure and costs disclosure, which become even
more difficult for products that offer multiple options and/or are linked to insurance
contracts. For the presentation of risk and return the ESA-DP discusses general
types or ways of presenting information, considered in abstract rather than specific
terms. Various concrete examples implemented or on its way in different European
countries are shown to illustrate visual elements. Figure 12.4 illustrates the Belgian
The ESA’s address consumer behaviour as one of the issues for retail investors
and “in order for the KID to be easy to use by the retail investor”, they envisage that
“the KID should have a clear behavioural purpose for the retail investor”77:
Given this, a traditional approach to disclosures focused solely on information and with
little regard to its presentation, is in being superseded in policy making by an approach that
is more informed by insights into consumer behaviours. For instance, the framing of infor-
mation can be considered, so as to counter cognitive biases which may distort perceptions
and provide information in a way that is both simple to understand but also salient for the

investment, creating a greater risk of misunderstanding on the part of the retail investor or if the
investment’s pay-off takes advantage of retail investor’s behavioural biases, such as a teaser rate
followed by a much higher floating conditional rate, or an iterative formula.”
ESA-DP (fn. 15).
ESA-DP (fn. 15), p. 17 et seq.; with reference to Sunstein 2011.
12 Nudging: Information, Choice Architecture and Beyond 233

consumer (i.e. capable of drawing the consumers’ attention and appearing important for
the decision to be made). The PRIIPs Regulation reflects these considerations already at
‘level one’.

With the adoption of the MiFID II/MiFIR-package there is additionally regula-

tion in place addressing also aspects of communication with clients.78 While MiFID
II at level 1 (accompanied by level 2 implementing directives and level 3 regulatory
as well as implementing technical standards) focuses on suitability and conflicts of
interests the PRIIP regulation addresses the KID.

Risk Tolerance Assessment within the Scope of Suitability Requirements

Already the European regulatory approach within MiFID provides for such tests;
however, this was not preventing European private investors (not aware and prop-
erly informed about the issuers default risk) from losing money with Lehman cer-
tificates.79 Assessing investment suitability requires the right set of questions to
cope with diverging experience, personal investment objectives as well as with
individual risk-capabilities and risk-tolerances. International politics have picked
up these issues as well.80 The assessment of suitability shall determine, on an indi-
vidual basis, whether a client has not only the necessary expertise to trade in the
specific financial instrument, but also whether he can understand and can take the
associated risks – and all of that in line with the individual financial status and
investment objectives.81
Investors are looking for predictable returns and they hate losses. Therefore, the
downside-potential should be clarified during the conduct with the customer; under

According to Art. 24 (4) (b) MiFID II the information to be provided to client shall include guid-
ance on and warnings of the risks associated with investments in financial instruments or in respect
of particular investment strategies. ESMA’s MiFID Consulting Paper from May 2014 (311 pages,
additionally ESMA published in May 2014 a Discussion paper regarding the future technical stan-
dards with 533 pages, meanwhile the consultation paper was published in December 2014, 645
pages) specifies in Section 2.13 that Article 31 (2) MiFID Implementing Directive (2006/73/EC),
relating to the description of risks, should specifically address the risk of financial instruments.
Baisch and Weber 2015, pp. 174–180.
In the OECD G20 High-Level Principles on Financial Consumer Protection, published in
October 2011,
edsectors/48892010.pdf,) the principle Disclosure and Transparency is stressed in section 4; the
following formulation exemplifies that this principle is clearly understood in a comprehensive
manner: “Financial services providers and authorised agents should provide consumers with key
information that informs the consumer of the fundamental benefits, risks and terms of the product.
They should also provide information on conflicts of interest associated with the authorised agent
through which the product is sold.” Addressing risks it is stated: “The provision of advice should
be as objective as possible and should in general be based on the consumer’s profile considering
the complexity of the product, the risks associated with it as well as the customer’s financial objec-
tives, knowledge, capabilities and experience. Consumers should be made aware of the importance
of providing financial services providers with relevant, accurate and available information.”
According to Rapp 1998, p. 212, “the most common characterization offered is that suitability is
an amorphous concept, with no accepted definition, and bereft of case law guidance”.
234 R. Baisch

the regime of MiFID II and PRIIPS regulation this issues are addressed. Most inves-
tors will be more than happy to surrender part of the up-side potential when this can
be combined with a cap of any potential losses. The customers risk attitude matters;
therefore, the focus should not be on their risk absorbing capacity.82
The draft technical advice in ESMA’s Consulting Paper83 recommends that Art.
35 of the MiFID Implementing Directive is expanded to clarify that the investment
firm should be required to provide for a suitability report to the retail client that must
include how the recommendation provided is suitable for the retail client, including
how it meets the client’s objectives and personal circumstances with reference to
client’s attitude to risk and capacity for loss.

12.4.2 FCA and Behavioural Economics

Since the FSA (Financial Services Authority) split into the PRA (Prudential
Regulation authority) and FCA (Financial Conduct Authority), the idea of
Behavioural Economics gained in importance. The FCA publishes occasional
papers addressing also aspects of behavioural economics and argues that behav-
ioural economics can help the regulator understand the mistakes consumers make
and how firms respond to these mistakes to help FCA consider appropriate interven-
tions. The first of a series of meanwhile ten occasional papers84 is focused on and
applies behavioural economics at the FCA, summarises the main lessons from
behavioural economics for retail financial markets and describes how to use behav-
ioural economics in the regulation of financial conduct.85 Effectively, this paper
opened up the debate around the use of interventions like nudges in financial
Behavioural biases have marked effects on how people make economic decisions. These
effects are particularly pertinent in financial services where products are particularly com-
plex, where consumers may not have much opportunity to learn about products, and where
decisions often involve assessment of risk and uncertainty, trade-offs between the past and
the future and emotions such as anxiety or regret.86

Especially, in the last two years the FCA embedded behavioural economics in its
regulatory practice. This happened because of the perception that, with the help of
sophisticated analysis of consumer behaviour, problems in markets can be identi-
fied and then those findings will help to design policy interventions. Behavioural
experiments are one of the tools regulators can use to support practical behavioural
analysis. This methodology makes it possible to study how consumer behaviour

Baisch and Weber 2015, pp. 172–174 and pp. 182–187.
CP May 2014 (fn. 71), p. 132 et seq.
Available at
FCA-OP 1 (fn. 17).
FCA-OP 1 (fn. 17), p. 51.
12 Nudging: Information, Choice Architecture and Beyond 235

varies across different contexts, and to identify the causes of these variations with
far higher level of confidence than simply observing consumers’ choices in the
market. The FCA chief executive Martin Wheatley clarified the role behavioural
economics is playing in assisting the FCA advance its competition objective by say-
ing “…what we’ve had over the years is a perfect storm of conditions eroding com-
petitive pressure” and mentioning “bounded willpower and rationality” on the buy
side as well as “moral failures on the sell side – knowingly exploiting consumer
biases”. With the help of behavioural economics FCA aims to break the “link
between poor products and high financial reward”.87
Instead of the traditional FSA-style enforcement of conduct rules, the FCA has
developed a new basis for intervention in the name of consumer protection based on
a wider scepticism about the efficiency of markets. Therefore, the means of smart
disclosure accepting the insights of behavioural economics led to a revisiting of
requirements on disclosure and marketing promotions like demonstrated with the
PRIIP regulation.

12.4.3 Good Practices on Information Provision for DC


The EIOPA report88 found that substantial improvements can still be made to infor-
mation a provision that is meant to support members in their retirement planning;
the provision of sufficient legally and technically relevant information has proven to
be ineffective. Information requirements are discussed based on the exploration of
good practices in different member states as well as the translation of insights from
behavioural economics and communication science. Because of the tendency of
governments to reduce public pension schemes and the shift from DB (designed
benefit) to DC (designed contribution) employer schemes, increasingly risk and
responsibility for individual retirement planning are passed onto the consumers. In
some countries first pillar pensions (often financed pay-as-you-go) tend to become
less generous due to the age pyramid and demographic change. Simultaneously
second pillar pensions tend to become more risky (bringing less interest). The report
clearly states that the provision of information is not sufficient by itself because
people tend not to cope rationally with risks and, therefore, have great difficulty
planning for retirement.
The report postulates a shift towards the insight that a legal purpose of informa-
tion provision is inadequate because people need first and foremost key information.
The new approach to information provision is neither focusing on the perspective of

Making competition king – the rise of behavioural economics at the FCA, Speech by Martin
Wheatley, Chief Executive, the FCA, at the Australian Securities and Investments Commission
(ASIC), March 2014; available at
EIOPA (fn. 56).
236 R. Baisch

the market nor just aiming to comply with legal purposes. Legal information is
necessary, and has to be disclosed in an appropriate way, however, information
should support people in making financial decisions. Therefore, less is more and the
format is essential, because many readers
will turn away from information if they do not quickly understand how it is relevant to them
and how they should translate the information into financial decisions.89

The information provider has to realise and accept that the readers lack motiva-
tion and time.90
On the way to IORP II (institutions for occupational retirement provision) it will
have to be seen how much of these insights remain in the final text. In March 2014,
the European Commission published a proposed revision of the IORP Directive,
IORP II.91 Smart information matters regarding the PBS (pension benefit state-
ment). The proposal would introduce a standardised PBS at EU level with simple
and clear information about the individual pension entitlements to support informed
decision-making about pension adequacy (answering the question and investment
strategy. The PBS will contain both personalised and generic information about the
pension scheme according to a standard template of two pages with a view to help-
ing individuals take decisions regarding their retirement planning.

12.5 Practical Applications

Whereas it is assumed that information leads

to understanding, to the willingness to act
and subsequently to appropriate actions,
this appears most often not the case.92

12.5.1 Key Investor Document: Disclosure as a Nudge

The EU regulatory response to the insights how information should be presented to

consumers can be seen within the PRIIPS regulation (Sect. 12.4.1 above). This
development was influenced by the behavioural research. As mentioned in the
beginning nudges may make their appearance in various forms and for financial
products a well-designed leaflet is essential. Sunstein explores

EIOPA (fn. 56), p. 9.
Helleringer 2015.
Available at
EIOPA (fn. 56), p. 8.
12 Nudging: Information, Choice Architecture and Beyond 237

the uses of disclosure as a behaviourally informed regulatory tool. It is important to distin-

guish between summary disclosure, often provided at the point of purchase, and full disclo-
sure, typically provided on the Internet. A central point is that disclosure policies should be
based on an understanding of how people process information.93

Looking at necessary information for the evaluation of financial product risk and
cost jut out. The regulatory efforts in Europe foster a clear disclosure of any kind of
administrative cost in connection with the purchase of an investment product94 and
present an answer on how to present risk and return.95 Sunstein stresses the impor-
tance of “plain language, clarity, and simplicity”.96 And this is important regarding
not only risk and return, but also the cost structure of a product. However, “accurate
disclosure of information may be ineffective if the information is too abstract, vague,
detailed, complex, poorly framed, or overwhelming to be useful”.97
Whether the three-page A4 standard format KID really helps (i) to understand
and compare PRIIPs, (ii) to estimate the total cost of the investment as well as (iii)
to raise awareness of the risk-reward profile will depend on the consumer’s willing-
ness to process this information. Perhaps the most challenging task will be to find a
solution to the question how the complex structures that may sit behind PRIIPs can
be broken down and explained in order to correctly state their likely return and risks
in a consistent manner. However, this is being left to the Level 2 or 3 measures.
Presumably, the KID will optimise the documentation for financial products sold to
retail investors and, hopefully, this is the way how to regulate disclosure avoiding
information overload but using it as a nudge. However, analyses have revealed that
there is still a large fraction of investors that is not able to understand the informa-
tion presented.98 In the face of information overload and the widespread unwilling-
ness of investors to take note of information, EU-regulators introduced the KID. The
scepticism remains: will a retail investor really process that information? And based
on the behavioural findings: may things go wrong anyway?99

Sunstein 2014b, p. 727.
See Article 33 of the level 2 MiFID-ID (MiFID Implementing Directive, Commission Directive
2006/73/EC of 10 August 2006 implementing Directive 2004/39/EC of the European Parliament
and of the Council as regards organisational requirements and operating conditions for investment
firms and defined terms for the purposes of that Directive, 2.9.2006, OJ L 241/26–58) regarding the
obligation to inform clients about costs.
ESA-DP (fn. 15), p. 21 et seq.
Sunstein 2014b, p. 728.
Sunstein 2014b, p. 729.
Walther 2015, p. 145.
“Traditional failures are often addressed by information provision or disclosure: to mitigate
asymmetric information, to reduce search costs and limit market power, and to remedy the under-
provision of a public good. But mandated information provision may be an ineffective remedy if
consumers either do not understand the information or believe that it is not relevant to their deci-
sionmaking. For example, if consumers mistakenly believe that they will pay their credit bill on
time every month, clear and transparent disclosure of late fees and interest rates may not change
behavior because consumers deem the information irrelevant at the time they make a purchase.”
See Campbell et al. 2011, p. 95.
238 R. Baisch

12.5.2 Classification of Customers

The EU-regulation distinguishes between professional and retail investors allowing

for opting in and out. This categorisation can be seen as a default rule; default rules
are another good example how nudging can ease choices. Further classifications
based on income, fortune and experience are conceivable. Part of the complexity of
effective retail market regulatory intervention is to identify the typical retail inves-
tor and to look at the approach to investment decisions.
As mentioned (above Sect. 12.2.3) the FCA has introduced additional investor
types, the sophisticated investor and the high net worth investor. Due to the fact that
self-certification is allowed, it is possible to opt out from the tighter investor protec-
tion regime for normal retail investors.100 The introduction of sub-divisions to fur-
ther classify retail investors and then to scrutinise financial products to find out
which product should be made available to which category might not seem to be
preferable because clients are too disparate to be pigeonholed.101 The question is
whether a national financial authority or for instance ESMA in Europe is capable
enough to act against hazardous financial products, which anyhow should not pass
the suitability test.102 Therefore, it could be argued that the regulator should focus
on conflict of interest issues and proper suitability-tests with simultaneous consid-
eration of ways to optimise disclosure and appropriate actions to avoid unfair adver-
tising practices.

12.5.3 Automatic Enrolments and Active Choice

With regard to pension schemes the question arises how much a state shall rely on
individual decisions to avoid poverty among the elderly. Many states know manda-
tory schemes but often such funds hardly cover even basic needs. A lot of research
has been done to find out what helps to motivate individuals to save for their retire-
ment.103 Reduced or downstream taxation is quite common to incentivise saving.
Various combinations of active or passive choices can be found. Subsidies for any
kind of retirement accounts rely on individuals to take action and save on a regu-
larly basis. What might happen is that this only induces individuals to shift assets
from taxable accounts to retirement accounts. Policies that raise retirement contri-
butions if individuals take no action such as automatic employer contributions to

Financial Services and Markets Act (Financial Promotion) Order 2005 Art. 48 and 50A.
In other areas than investment decisions the so-called choice architecture may help more. In the
U.S. the automatic enrolment in pension schemes is widely discussed. See with further references
Lusardi and Mitchell 2014, 34.
However, the grey capital market, closed-end funds and investment companies will remain a
See references in Benartzi et al. 2013.
12 Nudging: Information, Choice Architecture and Beyond 239

retirement accounts increase wealth accumulation substantially.104 In contrast, tax

incentives often are motivating especially those who would anyhow save. Traditional
policy incentives like tax deductions address rational actors maximising their utility.
The behavioural approach adds incentives which are nudging biased individuals.105
Therefore, to increase saving ratio from low and average earners often only
paternalistic rules will ensure the wanted result.106 While defaults may have a posi-
tive influence on consumer decisions, a practical alternative to defaults, requiring
individuals to make mandatory explicit choices might be better in many cases. To
make active decisions mandatory is optimal when consumers have a strong propen-
sity to procrastinate because savings preferences are highly heterogeneous.107 It fits
well in this picture that an analysis of the recognised Swedish pension system
observes that major changes in retirement savings behaviour require clear signals
and relatively intrusive instruments to affect behaviour and suggests strong penal-
ties for non-compliance rather than just nudges.108

12.5.4 Nudging the Advisor

Against the background of independent advice, retrocessions and similar payments

from product produces towards investment advisors or portfolio managers were get-
ting more and more disreputable.109 Slack regulation was common in the problem-
atic field of inducements, which are causing conflicting interests; the effect of utility
maximisation – thereby motivating bankers to create and sell products which are not
in the best interest of their clients – was misjudged. Certainly the consumer needs
clear information about the cost structure of products to compare them. However,
there are other areas, where rational behaviour must be intensively kept in mind:
advisors will always be tempted to sell a product when they can maximise their
All kinds of inducements and incentives like trailer fees, retrocessions and other
commissions used to be common as means of sales promotion in the banking and
fund industry. Conflicts of interest are to be monitored to the extent such payments
may influence investment advisory and portfolio management services. Long-term
considerations should actually lead to recommendations triggered by the hope for
satisfied customers, while short-term temptations might result in preferring a good
salary and high bonuses. What is true for an individual advisor himself at the POS
is also true for banks and other financial intermediaries as a whole or for a certain
division. Especially MiFID II strengthens the independence of investment advice

Chetty et al. 2013.
Cartwright 2014, p. 519.
Bronchetti et al. 2013.
Carroll et al. 2009, p. 1669.
Weaver and Willén 2014.
Baisch and Weber 2015, p. 161 and pp. 171–172.
240 R. Baisch

and portfolio management services.110 Gone are the times when undisclosed fees
were helping to line one’s pockets. However, looking at international regulatory
activities, huge differences exist; some regulators aim to eliminate such kind of pay-
ment flows at all for independent advice, others just demand a full disclosure
approach. While for many investment products like shares or bonds an efficient
formation of prices – liquid markets presumed – is at least not impossible based on
Modern Finance Theory, in contrast, many of the structured products do not facili-
tate a clear price evaluation.111 Additionally, the advisor and the advised have both
to cope with human limitations. Shefrin argues112:
Kahneman (2011) might say that the judgments of portfolio managers and analysts reflect
subconscious System 1 (intuitive) thinking, even though the principles they articulate reflect
System 2 (conscious) thinking. Elements of the planning fallacy, also discussed by
Kahneman (2011), strike me as being part of this story. These elements involve under-
weighting or ignoring base rate information (…) when making predictions about future
returns, and overweighting singular (firm specific) information.

What kind of additional nudges at the point of sale are possible? Salience and
simplification embedded in of a Code of Conduct could lead to better investor pro-
tection to establish a fair treatment of customers. The approach towards POS-
regulation for financial intermediaries should be based on the enforcement of a
Code of Conduct, assuming that a biased behaviour is shown by retail investors.
Particularly the mandatory suitability or appropriateness tests113 prior to any
investment advice as well as an on-going assessment of the suitability of the recom-
mended financial instruments embracing also loss-aversion and attitude towards
risk features is an area where the outcomes of behavioural research can be applied.
The need to counteract conflicts of interest is also part of the sales process

12.6 Conclusion

One word of caution, however. Human behaviour

and nudges, in particular, are a notoriously complex equation.
Inputs and outputs do not always correlate.
Nor are they predictable.114

Art. 24 MiFID II.
According to Hens and Rieger 2008, 27 most of the common structured products do not “follow
rational guideline, but instead use behavioral factors, like loss-aversion or probability misjudge-
ment to be attractive in the eyes of potential investors”.
Shefrin 2014, p. 19.
While suitability contains a link to present as well as future aspects, appropriateness mainly
reflects the past. Based on detailed conduct of business rules for investment firms in combination
with the just discussed transparency requirements these tests are realising the general obligation to
act fairly, honestly and professionally and in accordance with the best interests of the client – previ-
ously known as fiduciary duty.
Speech Wheatley (fn. 90).
12 Nudging: Information, Choice Architecture and Beyond 241

Regulatory requirements and industry standards establishing an environment

where customers are treated according to their individually assessed risk attitude
constitute the best precaution against negative outcomes of human flaws when mak-
ing financial decisions. However, there is a potential barrier to obtaining acceptance
with this approach to suitability because the widespread rejection of paternalistic
regulation which still has the model of a reasonable investor in mind. Therefore, any
kind of provision should be avoided which too paternalistically overrides people’s
ability to make choices. Recognising human limitations because of the complexity
of financial decision making should not lead to a reduction of choices and invest-
ment opportunities for those customers who can handle the complexity or just like
to gamble.115 Despite the fact that suitability requirements are problematic in terms
of supervision and enforcement,116 it should be possible to protect those deserving
protection without harming the others. To comply with sound governance structures
reputable market participants will behave accurately and in line with the require-
ments like it is stated in the respective IOSCO principles:
Whenever an intermediary recommends the purchase of a particular complex financial
product, including where the intermediary advises or otherwise exercises investment man-
agement discretion, the intermediary should be required to take reasonable steps to ensure
that recommendations, advice or decisions to trade on behalf of such customer are based
upon a reasonable assessment that the structure and risk-reward profile of the financial
product is consistent with such customer’s experience, knowledge, investment objectives,
risk appetite and capacity for loss.117

Given that many consumers struggle to maximize their own welfare, the ques-
tion remains whether regulators can do better. When regulating code of conduct
requirements, the presumed preferences of the majority of consumers should be
reflected. The challenging question is: what would they do were they fully informed
and well advised?118 This is the challenging task for the law. Ulen119 phrased it this
way: “Law – by taking due account of these predictable, routine deviations from
rationality – can better influence behaviour to realize both social goals and to help
individuals better enhance their well-being.”120 (Fig. 12.5)
This is not only true for any kind of choice architecture but also for more pater-
nalistic forms of regulation. Focusing on areas where there is strong evidence of
self-harm (including financial losses) while limiting the influence of political beliefs
and encouraging active and lively discussions will help to improve financial market

See Jolls 2006, pp. 39–40.
Moloney 2010, p. 243, advocates mystery shopping. For a practical example see FSA, Assessing
the quality of investment advice in the retail banking sector: A mystery shopping review, February
2013, <>.
IOSCO, Suitability Requirements With Respect To the Distribution of Complex Financial
Products Final Report, January 2013, Principle 5.
Campbell et al. 2011, p. 95.
Ulen 2014, p. 120.
See also Mitchell 2014, pp. 182–184, regarding the interaction of behavioural law and econom-
ics with legal norms.
242 R. Baisch

Fig. 12.5 Geared to

consumer’s needs

regulation.121 Additionally, the potential trade-off should be kept in mind which

might occur undesirably if defaults reduce the incentive to gather information.122
Kahneman’s analysis of human thinking can give the impression that human naivety
and simple-mindedness is overemphasised.123 Therefore, Gigerenzers’s approach to
value the positive outcome of simple rules of thumb or similar mental shortcuts
might complement to an optimised cognition and perception of human behaviour as
it should be imputed to individuals when designing regulation.124 Pushing it a bit,
according to Kahneman individuals are desperate in thrall to all kinds of fallacies,
which fosters the call for paternalism. Nudge theory should not weaken the motiva-
tion of individuals to actively inform them. It must be said that education could, at
least partially, compensate negative outcomes of human flaws. Gigerenzer proposes
that with the teaching of critical thinking about statistical probability and heuristics
individuals can get more risk savvy.125 Based on such an enhanced level of financial
literacy consumer then are better able (i) to comprehend investment risks, (ii) to
accept that a certain level of it is unavoidable in combination with given return
expectations and finally (iii) tolerate potential losses.
Enhanced prevention based on smarter information requirements will give con-
sumers a better chance to avoid human flaws. Combined with a strict Code of
Conduct at the point of sale, a Suitability Test including the real risk perceptions of

Glaeser 2006, p. 156.
Carlin et al. 2013.
Kahneman 2011. Neuroscientist Gazzaniga 2011 tackles the question, why individuals feel so
in control despite the range of automatic and deliberative factors that underlies their choices.
Gigerenzer 2014 (2013), p. 30; see also his comment in chapter 2 fn. 14.
System 1 (fast, instinctive and emotional), as Kahneman describes it, might lead in some cases
even to better solutions then System 2 (deliberative and logical) because System 2 is susceptible to
the misjudgement of probabilities due to limited working memory and brain power and a wide-
spread simple lack of statistical knowledge and mathematical ignorance.
12 Nudging: Information, Choice Architecture and Beyond 243

the consumer leads to an enhanced level of investor protection. Nudging in combi-

nation with some more paternalistic elements can de-bias detrimental human short-
comings while distending paternalism should be avoided.126 In addition, adequate
enforcement strategies are necessary; hereby, the question arises, whether financial
intermediaries need the carrot or the stick to implement processes to cope with the
limitations of biased investors. Regarding enforcement the carrot (nudges) does not
have the required disciplinary effect; therefore, the paternalistic stick (penances) is

Acknowledgement This contribution was written within the University Research Priority
Program Regulation of Financial Markets from the University of Zurich.


Akerlof, George A., and Robert J. Shiller. 2015. Phishing for phools. Princeton: Princeton
University Press.
Baisch, Rainer, and Rolf H. Weber. 2015. Investment suitability requirements in the light of behav-
ioural findings − Challenges for a legal framework coping with ambiguous risk perception. In
Behavioural law and economic, ed. Klaus Mathis, 159–192. Dordrecht/Heidelberg/New York/
London: Springer.
Benartzi, Shlomo, Ehud Pelejj, and Richard H. Thaler. 2013. Choice architecture and retirement
saving plans. In The behavioral foundations of public policy, ed. Eldar Shafir, 245–263.
Princeton: Princeton University Press.
Black, Barbara. 2013. Behavioral economics and investor protection: Reasonable investors, effi-
cient markets. Loyola University Chicago Law Journal 44: 1493–1508.
Blumenthal-Barby, Jennifer S. 2013. Choice architecture: A mechanism for improving decisions
while preserving liberty. In Paternalism – Theory and practice, ed. Christian Coons and
Michael Weber, 179–196. Cambridge: Cambridge University Press.
Boaz, David. 1997. Libertarianism: A primer. New York: The Free Press.
Bovens, Luc. 2009. The ethics of nudge. In Preference change, approaches from philosophy, eco-
nomics and psychology, ed. Till Grüne-Yanoff and Sven O. Hansson, 207–220. Dordrecht:
Bronchetti, Erin T., Thomas S. Dee, David B. Huffman, and Ellen Magenheim. 2013. When a
nudge isn’t enough: Defaults and saving among low-income tax filers. National Tax Journal
66: 609–634.
Camerer, Colin F., Samuel Issacharoff, George F. Loewenstein, Ted O’Donoghue, and Matthew
Rabin. 2003. Regulation for conservatives: Behavioral economics and the case for ‘Asymmetric
Paternalism’. University of Pennsylvania Law Review 151: 1211–1254.
Campbell, John, Howell Jackson, Brigitte Madrian, and Peter Tufano. 2011. Consumer financial
protection. Journal of Economic Perspectives 25: 91–114.
Carlin, Bruce, Simon Gervais, and Gustavo Manso. 2013. Libertarian paternalism, information
production, and financial decision-making. Review of Financial Studies 26: 2204–2228.
Carroll, Gabriel D., James J. Choi, David Laibson, Brigitte Madrian, and Andrew Metrick. 2009.
Optimal defaults and active decisions. The Quarterly Journal of Economics 124: 1639–1674.

Gigerenzer 2014 (2013), chapter 1 fn. 14, argues that “poor cognitive abilities are not written
into our genes, but are largely a consequence of the lack of an intellectually stimulating
244 R. Baisch

Cartwright, Edward. 2014. Behavioral economics, 2nd ed. London: Routledge.

Chetty, Raj, Friedman, John N., Soren, Leth-Petersen, Nielsen, Torben H., and Tore, Olsen. 2013.
Active versus passive decisions and crowd-out in retirement savings accounts: Evidence from
Denmark, NBER working paper No. 18565,
Conly, Sarah. 2013. Against autonomy, justifying coercive paternalism. Cambridge: Cambridge
University Press.
Cunningham, Lawrence A. 2002. Behavioral finance and investor governance. Washington and
Lee Law Review 59: 767–837.
Davidoff, Steven M., and Claire A. Hill. 2013. Limits of disclosure. Seattle University Law Review
36: 599–637.
Dworkin, Gerald. 2013. Defining paternalism. In Paternalism – Theory and practice, ed. Christian
Coons and Michael Weber, 25–38. Cambridge: Cambridge University Press.
Gabaix, Xavier, and David Laibson. 2006. Shrounded attributes, consumer myopia, and informa-
tion suppression in competitive markets. Quarterly Journal of Economics 121: 505–540.
Gal, David. 2006. A psychological law of inertia and the illusion of loss aversion. Judgment and
Decision Making 1: 23–32.
Gazzaniga, Michael S. 2011. Who’s in charge? Free will and the science of the brain. New York:
Gigerenzer, Gerd. 2014 (in German 2013). Risk savvy: How to make good decisions. London:
Glaeser, Edward. 2006. Paternalism and psychology. University of Chicago Law Review 73:
Goodin, Robert. 1991. Permissible paternalism: In defence of the Nanny State. The Responsive
Community 1: 42–51.
Hansen, Pelle G., and Andreas M. Jespersen. 2013. Nudge and the manipulation of choice: A
framework for the responsible use of the nudge approach to behaviour change in public policy.
European Journal of Risk Regulation 1: 3–28.
Hausman, Daniel M., and Brynn Welch. 2010. Debate: To nudge or not to nudge. The Journal of
Political Philosophy 18: 123–136.
Helleringer, Geneviève. 2015. In Behavioural law and economics, ed. Klaus Mathis, 193–210.
Dordrecht/Heidelberg/New York/London: Springer.
Hens, Thorsten, and Marc, Oliver Rieger. 2008. The dark side of the moon: Structured products
from the customer’s perspective. Working Paper No. 459.
Johnson, Eric J., Suzanne B. Shu, Benedict G.C. Dellaert, Craig Fox, Daniel G. Goldstein, Gerald
Haeubl, Richard P. Larrick, John W. Payne, Ellen Peters, David Schkade, Brian Wansink, and
Elke U. Weber. 2012. Beyond nudges: Tools of a choice architecture. Marketing Letters 23:
Jolls, Christine. 2006. Behavioral law and economics. Yale Law School, Public Law Working
Paper No. 13.
Kahneman, Daniel. 2011. Thinking, fast and slow. New York: Farrar, Straus and Giroux.
Kahneman, Daniel. 2013. Behavioral economics and investor protection: Keynote address. Loyola
University Chicago Law Journal 44: 1333–1340.
Kahneman, Daniel, and Amos Tversky (eds.). 2000. Choices, values, and frames. New York:
Russell Sage Foundation.
Kahneman, Daniel, Paul Slovic, and Amos Tversky. 1982. Judgment under uncertainty: Heuristics
and biases. Cambridge: Cambridge University Press.
Keynes, John M. 1936. The general theory of employment, interest, and money. London:
Kingsford Smith, Dimity. 2010. Financial services regulation and the investor as consumer. In
Handbook of research on international consumer law, ed. Geraint Howells, Iain Ramsay, and
Thomas Wilhelmsson, 431–481. Cheltenham: Edward Elgar.
Korobkin, Russell B., and Thomas S. Ulen. 2000. Law and behavioral science: Removing the
rationality assumption from law and economics. California Law Review 88: 1051–1144.
12 Nudging: Information, Choice Architecture and Beyond 245

Langevoort, Donald C. 1992. Theories, assumptions, and securities regulation: Market efficiency
revisited. University of Pennsylvania Law Review 140: 851–920.
Lusardi, Annamaria, and Olivia Mitchell. 2014. The economic importance of financial literacy:
Theory and evidence. Journal of Economic Literature 52: 5–44.
Madrian, Brigitte C., and Dennis F. Shea. 2001. The power of suggestion: Inertia in 401(k), partici-
pation and savings behavior. Quarterly Journal of Economics 116: 1149–1187.
Mitchell, Gregory. 2014. Alternative behavioural law and economics. In Oxford handbook of
behavioral economics and the law, ed. Eyal Zamir and Doron Teichman, 167–191. Oxford:
Oxford University Press.
Moloney, Niamh. 2008. EC securities regulation, 2nd ed. Oxford: Oxford University Press.
Moloney, Niamh. 2010. How to protect investors: Lessons from the EC and the UK. Cambridge:
Cambridge University Press.
Moloney, Niamh. 2012. The investor model underlying the EU’s investor protection regime:
Consumers or investors? European Business Organization Law Review 13: 169–193.
Moloney, Niamh. 2014. EC securities regulation and financial markets regulation, 3rd ed. Oxford:
Oxford University Press.
Mullane, Maggie, and Steven, Sheffrin. 2012. Regulatory nudges in practice, white paper prepared
for the conference on responsible regulation, February 2012.
Paredes, Troy A. 2003. Blinded by the light: Information overload and its consequences for securi-
ties regulation. Washington University Law Review 81: 417–485.
Rapp, Robert N. 1998. Rethinking risky investments for that little old lady: A realistic role for
modern portfolio theory in assessing suitability obligations of stockbrokers. Ohio Northern
University Law Review 24: 189–279.
Rebonato, Ricardo. 2012. Taking liberties, a critical examination of libertarian paternalism.
New York: Palgrave Macmillan.
Reiss, Julian. 2013. The philosophy of economics: A contemporary introduction. New York:
Ryan, Andrea, Gunnar Trumbull, and Peter Tufano. 2011. A brief post-war history of U.S. con-
sumer finance. Business History Review 85: 461–498.
Sarra, Janis P. 2013. Embedding fairness as a fundamental norm in financial markets. In An explo-
ration of fairness, ed. Janis P. Sarra, 193–240. Toronto: Carswell.
Shefrin, Hersh. 2014. Investors’ judgments, asset pricing factors, and sentiment.
Sunstein, Cass R. 2005. Laws of fear, beyond the precautionary principle. Cambridge: Cambridge
University Press.
Sunstein, Cass R. 2011. Empirically informed regulation. University of Chicago Law Review 78:
Sunstein, Cass R. 2013. Behavioral economics, consumption, and environmental protection.
Regulatory Policy Program Working Paper RPP-2013-19. Cambridge, MA: Mossavar-
Rahmani Center for Business and Government, Harvard Kennedy School, Harvard University.
Sunstein, Cass R. 2014a. Why nudge, the politics of libertarian paternalism. New Haven: Yale
University Press.
Sunstein, Cass R. 2014b. Nudges. gov: Behaviorally informed regulation. In Oxford handbook of
behavioral economics and the law, ed. Eyal Zamir and Doron Teichman, 719–747. Oxford:
Oxford University Press.
Sunstein, Cass R. 2014c. The ethics of nudging (November 20, 2014). Available at SSRN: http://
Sunstein, Cass R. 2015. Behavioral economics, consumption, and environmental protection. In
Handbook on research in sustainable consumption, ed. Lucia Reisch and John Thøgersen,
313–327. Cheltenham: Edward Elgar.
Sunstein, Cass R., and Richard H. Thaler. 2003. Libertarian paternalism is not an oxymoron.
University of Chicago Law Review 70: 1159–1202.
246 R. Baisch

Thaler, Richard H. 1980. Toward a positive theory of consumer choice. Journal of Economic
Behavior and Organization 1: 39–60.
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge: Improving decisions about health, wealth
and happiness. New Haven: Yale University Press.
Thaler, Richard H., Cass R. Sunstein, and John P. Balz. 2013. Choice architecture. In The behav-
ioral foundations of public policy, ed. Eldar Shafir, 428–439. Princeton: Princeton University
Trout, J.D. 2005. Paternalism and cognitive bias. Law and Philosophy 24: 393–434.
Ulen, Thomas S. 2013. A behavioral view of investor protection. Loyola University Chicago Law
Journal 44: 1357–1376.
Ulen, Thomas. 2014. The importance of behavioural law. In Oxford handbook of behavioral eco-
nomics and the law, ed. Eyal Zamir and Doron Teichman, 93–124. Oxford: Oxford University
van Aaken, Anne. 2016. Constitutional limits to nudging: A proportionality assessment. In Choice
architecture in democracies: Exploring the legitimacy of nudging, ed. Möllers Christoph et al.
Oxford: Hart Publishing.
Walther, Torsten. 2015. Key investor documents and their consequences on investor behaviour.
Journal of Business Economics 2: 129–156.
Weaver, Kent, and Alexander Willén. 2014. The Swedish pension system after twenty years: Mid-
course corrections and lessons. OECD Journal on Budgeting 13(3): 1–26.
White, Mark D. 2013. The manipulation of choice, ethics and libertarian paternalism. New York:
Palgrave Macmillan.
Whitman, Douglas G., and Mario J. Rizzo. 2007. Paternalist slopes. NYU Journal of Law & Liberty
2: 411–443.
Chapter 13
Nudging and Environmental Law
Perspectives and Examples

Felix Ekardt and Jutta Wieding

Abstract This article analyses nudging in constitutional and environmental law.

On the level of constitutional law, nudging is criticized for being paternalistic. On
second thought, many forms of nudging are in contrast not paternalistic, because
they do not protect an individual from themselves, but others from negative effects
of the actions of an individual. Especially in environmental policies, nudging there-
fore proves to be legitimate. However, nudging might not be as effective as policy
instrument for environmental issues, as many hope. Even though this article shows
interesting possible examples of nudging, it also argues why the effect of nudging
as main policy instrument in solving environmental problems will be limited.

13.1 Problem: Nudging and the Environment

Despite much talk on environmental issues, the ecological situation in Europe and
worldwide gives reason for concern. Climate change as a consequence of using fos-
sil fuels for example will entail dire and existential consequences in the course of
the twenty-first century. This includes food and water scarcity, leading to billions of
deaths due to supply shortages, as well as wars and civil wars over decreasing
resources, migration flows, massive natural disasters, exploding petrol and gasoline
prices – and last, but not least, economic damage, which will exceed costs of effec-
tive climate policies by far. Considering per capita emissions and alleged reduction
measures, Europe and countries like Germany are far from being forerunners. Alone
the shifting of production out of the EU between 1990 and 2008 accounts for more
emissions than the total of 10 % which has been reduced within the EU according
to statistics.1

Ekardt et al. 2015.
F. Ekardt (*) • J. Wieding
Forschungsstelle Nachhaltigkeit und Klimapolitik, Leipzig, Germany

© Springer International Publishing Switzerland 2016 247

K. Mathis, A. Tor (eds.), Nudging - Possibilities, Limitations and Applications
in European Law and Economics, Economic Analysis of Law in European
Legal Scholarship 3, DOI 10.1007/978-3-319-29562-6_13
248 F. Ekardt and J. Wieding

Originally developed to foster general health and well-being, the concept of

nudging is an attempt to find new instrumental answers to those and other environ-
mental issues. Unlike environmental policies, nudging does not aim at dictating citi-
zens a certain behaviour or putting a cost incentive on them. The aim is rather to
give people a ‘nudge’ into the right direction. Proponents of this concept call it a
gentle form of paternalism, which is not intruding with individual freedom: A ‘lib-
ertarian paternalism’.2 It is not only – and maybe not even foremost – applicable to
protecting the environment, yet, this paper will assess its effectiveness regarding
environmental policy. This is interesting, because looking at an example of applica-
tion in a field outside the centre of debate will broaden the perspective for evaluat-
ing the theory. It might also assess a new approach to changing people’s behaviour,
such as consumption patterns and lifestyle choices, which is an essential element in
solving may environmental issues.

13.2 Sustainability: The Objective of Environmental

Protection and Environmental Law

Protection of the environment and its policy instruments can be measured by the
goal of sustainability which has been shaping international politics for a while.3 It
addresses the question, whether politics has the obligation to foster lifestyle and
economic behaviour that will work over time on a global scale. This concept of
sustainability implies by definition the call for more intertemporal and global jus-
tice. Especially regarding climate change and use of resources, this goal is not even
close to being achieved. Per capita greenhouse gas emissions in Germany e.g.
remain at a multiple of what is permanently and globally feasible. A look at relative
emission reductions since 1990 reveals trick calculations like the shifting of produc-
tion, and through that emissions abroad, while the respective products are still con-
sumed in Germany.
Another view of sustainability however is often a kind of title of judgement on
everything which is desirable in human societies. This would include notions of
justice and public welfare, or could be even broader. In contrast, a three-pillar model
is often cited, which reduces sustainability to the necessary balance between eco-
logic, economic and social issues. This is diverting the main idea of the debate on
sustainability since 1987, which has been a remarkable change in paradigm: More
generational and global justice through making the rich protect resources and pro-
viding opportunities for development of the global poor. Firstly, the talk of ‘three
pillars’ trivializes sustainability by suggesting that political decisions are supposed
to consider the different areas equally, especially if intertemporal and global rele-
vance is left on the side-lines, if seen at all. A second problem is the separation of

Thaler and Sunstein 2008.
on this chapter, see Ott and Döring 2008; Ekardt 2015.
13 Nudging and Environmental Law 249

economic, social and economic aspects, which is close to impossible in relevant

areas. And what exactly is implied by the thoroughly vague and divers notion of
‘social’? If it were to include everything to do with mankind, sustainability would
be completely trivialized. Thirdly, the three-pillar model could be interpreted to the
effect that protection of livelihood strongly depends on economic growth. This
however is highly controversial.4
These and other aspects are clearly reflected in many parts of the Rio Declaration
of 1992, which is the key international origin of the modern discourse on sustain-
ability. Disregarding the definition of sustainability however, a question remains
regarding constitutional and basic rights, nationally, in Europe and internationally.
It concerns whether politics on all levels has an obligation to act to secure, e.g. in
the field of climate globally stable emission levels, disregarding how the problem is
called. Likewise, the question remains, which might be the right instrument to
achieve that. Exploring answers to both questions brings nudging into play.

13.3 Is Nudging Paternalistic: And Would It Then

Be Despicable in the Context of Liberal Democratic
Constitutional Law?

13.3.1 Nudging and Paternalism

Proponents of nudging call it ‘libertarian paternalism’, stressing that this is not an

oxymoron, because it is questionable, whether a paternalistic policy instrument is in
accordance with constitutions in liberal democracies. If an individual gets nudged
to stop smoking against their original conviction, and this were to happen for the
sole reason of protecting them from themselves, this could be considered paternal-
istic. What are we therefore to think of nudging, and is it generally possible to
imagine paternalistic forms of action?
Even disregarding nudging, one could argue against a global environmental turn,
e.g. in climate matters, on the basis of human-rights guaranteed freedom. Is it gov-
ernment’s concern how I choose to live? And is this is not exactly what is intro-
duced with regulations for a real climate turn? The answer has to be differentiated.5
Everything that concerns the freedom of several individuals and therefore the con-
flict between development opportunities of several citizens is a question of justice.
This might then not only be my personal problem. The next chapter will analyze
how to deal with those conflicts. This is also a question of a good life. If criterion
for legal an ethical justice is freedom, or self-determination, including the various
implications which derive from it, it is possible to limit a freed