Sie sind auf Seite 1von 463

The Global Innovation Index 2017

Innovation Feeding the World

The Global Innovation Index 2017
Innovation Feeding the World

Soumitra Dutta, Bruno Lanvin, and Sacha Wunsch-Vincent

The Global Innovation Index 2017: Innovation Feeding the World Cornell University, INSEAD, and the World Intellectual
is the result of a collaboration between Cornell University, Property Organization, 2017
INSEAD, and the World Intellectual Property Organization
This work is licensed under the Creative Commons Attribution
(WIPO) as co-publishers, and their Knowledge Partners.
Non-commercial No-Derivatives 3.0 IGO License. The user is
The report and any opinions expressed in this publication are allowed to reproduce, distribute, and publicly perform this
the sole responsibility of the authors. They do not purport to publication without explicit permission, provided that the
reflect the opinions or views of WIPO Member States or the content is accompanied by an acknowledgement that Cornell
WIPO Secretariat. University, INSEAD, and WIPO are the source. No part of this
publication can be used for commercial purposes or adapted/
The terms country, economy, and nation as used in this
translated/modified without the prior permission of WIPO.
report do not in all cases refer to a territorial entity that is a
Please write to treaties[dot]mail[at]wipo[dot]int to obtain
state as understood by international law and practice. The
terms cover well-defined, geographically self-contained
economic areas that may not be states but for which statistical To view a copy of the license, please visit http://
data are maintained on a separate and independent basis.
Any boundaries and names shown and the designations
When content, such as an image, graphic, data, trademark, or
used on any visual maps do not imply official endorsement
logo, is attributed to a third party, the user is solely responsible
or acceptance by any of the co-publishers. Chapters 211
for clearing the rights with the right holders.
contributions may deviate from UN terminology for countries
and regions. Suggested citation: Cornell University, INSEAD, and WIPO
(2017): The Global Innovation Index 2017: Innovation Feeding the
World, Ithaca, Fontainebleau, and Geneva.

ISSN 2263-3693
ISBN 979-10-95870-04-3

Printed and bound in Geneva, Switzerland, by the World

Intellectual Property Organization (WIPO), and in New Delhi,
India, by the Confederation of Indian Industry (CII).

Cover photo courtesy of



Preface: Releasing the Global Innovation Index 2017: Innovation Feeding the World v
By Soumitra Dutta, Cornell SC Johnson College of Business, Cornell University;
Francis Gurry, World Intellectual Property Organization; and Bruno Lanvin, INSEAD

Foreword: Innovation as the Key Driver of Sustainable Agriculture and Future vii
Food Security in the Developing World
By Chandrajit Banerjee, Director General, Confederation of Indian Industry

Foreword: Innovating to Feed the World ix

By Tim Ryan, US Chairman and Senior Partner, PwC

Foreword: Innovation in Food Production: Learning from the Past with an Open Mind for the Future xi
By Robson Braga de Andrade, President of CNI, Director of SESI, and President of SENAIs
National Council; and Guilherme Afif Domingos, President-Director of Sebrae

Contributors to the Report xiii

Advisory Board to the Global Innovation Index xv

Global Innovation Index 2017 Rankings xviii

Key Findings of the GII 2017 xxiii

Chapter 1: The Global Innovation Index 2017: Innovation Feeding the World 3
By Soumitra Dutta, Rafael Escalona Reynoso, and Jordan Litner, Cornell SC Johnson
College of Business, Cornell University; Bruno Lanvin, INSEAD; and Sacha Wunsch-
Vincent and Francesca Guadagno, WIPO

Annex 1: The Global Innovation Index (GII) Conceptual Framework 47

Annex 2: Adjustments to the Global Innovation Index Framework and

Year-on-Year Comparability of Results 55

Annex 3: Joint Research Centre Statistical Audit of the 2016 Global Innovation Index 59
By Michaela Saisana, Marcos Domnguez-Torreiro, and Daniel Vertesy,
European Commission, Joint Research Centre (JRC), Ispra, Italy

Annex 4: Measuring Innovation in Agriculture and Food Systems 73


Chapter 2: The Potential of a Global Diagnostic Tool for Agricultural Innovation Systems 81

By Christian Grovermann, Samy Gaiji, Karin Nichterlein, Abdoulaye Saley Moussa, Snia
Dias, Andrea Sonnino, and Delgermaa Chuluunbaatar, FAO

Chapter 3: The Role of Private-Sector R&D in Agricultural Innovation: Improving Yields, 89

Equipment Productivity, and Sustainability
By Barry Jaruzelski and Volker Staack, PwCs Strategy&; and Tom Johnson, PwC

Chapter 4: Innovation in Agriculture and Food Systems in the Digital Age 97

By Harold van Es and Joshua Woodard, Cornell University

Chapter 5: Digital Technologies Transforming Indian Agriculture 105

By Ankur Seth, formerly with the Confederation of Indian Industry; and Kavery Ganguly,
Confederation of Indian Industry

Chapter 6: Innovations in Food Distribution: Food Value Chain Transformations 113

in Developing Countries and their Implications for Nutrition
By Miguel I. Gmez and Katie D. Ricketts, Charles H. Dyson School of Applied Economics
and Management, Cornell SC Johnson College of Business, Cornell University

Chapter 7: Policies and Institutions Fostering Innovation and Agriculture Technologies in Brazil 121
By Robson Braga de Andrade, National Industry Confederation (CNI); and Guilherme Afif
Domingos, Brazilian Micro and Small Business Support Service (Sebrae)

Chapter 8: Mobilizing Science, Technology, and Innovation to Transform Japanese Agriculture 129
By Yuko Harayama, Council for Science, Technology and Innovation, Cabinet Office of

Chapter 9: Technological Future of the Agriculture and Food Sector in Russia 135
By Leonid Gokhberg and Ilya Kuzminov, National Research University Higher School of
Economics, Russia

Chapter 10: Innovation in the Agri-Food Sector in Latin America and the Caribbean 143
By Jos Luis Solleiro and Rosario Castan, National University of Mexico; Karla
Rodrguez, CamBioTec, A.C.; and Olivia Meja, National University of Mexico

Chapter 11: Enhancing Innovation in the Ugandan Agri-Food Sector: Progress, Constraints, 151
and Possibilities
By Travis Lybbert, Agricultural & Resource Economics, University of California Davis;
Kritika Saxena, Graduate Institute, Geneva; Julius Ecuru, Uganda National Council for
Science and Technology, Uganda; Dick Kawooya, University of South Carolina; and
Sacha Wunsch-Vincent, WIPO


Identifying and Ranking the Worlds Largest Clusters of Inventive Activity 161
By Kyle Bergquist, Carsten Fink, and Julio Raffo, WIPO

Appendix I: Country/Economy Profiles 179

Appendix II: Data Tables 313

Appendix III: Sources and Definitions 401


Appendix IV: Technical Notes 417

Appendix V: About the Authors 423


Releasing the Global Innovation

Index 2017: Innovation Feeding
the World

WIPO, 2017. Photo by Emmanuel Berrod.

We are pleased to present the Global Innovation Index years report analyses these demands and the opportuni-
(GII) 2017 on the theme Innovation Feeding the World. ties they present from different angles, including those
This year is a particularly noteworthy one for the GII, of data-driven strategies, the impact of biotechnological
as it marks the release of the 10th edition of the report. and digital technologies, effective policies, and strength-
The first edition was produced in 2007 by Soumitra Dutta ened networks, while at the same time suggesting new
at INSEAD with the goal of producing a comprehensive approaches for both developed and developing countries.
broad-based model of innovation that captured its com- Finally, the GII 2017 includes another innovation
plex nature in both developed and emerging economies. this year. The GII has long recognized that innovative
Over the last decade, the GII has gained international activity tends to be concentrated in geographic clusters.
recognition, establishing itself as both a leading reference However, no metrics have existed to measure innova-
on innovation and a tool for action for decision makers. tion performance at the cluster level on an internationally
Numerous countries have incorporated the GII into comparable basis. This years GII seeks to take a first step
their innovation agendas and metrics. Remarkably, in remedying this measurement gap. It presents a novel
in view of the GII, a large number of countries have approach towards identifying and ranking the worlds
increased their collection of innovation metrics that con- largest clusters of inventive activity, drawing on interna-
form to international standards; these countries also use tional patent filings. We hope that the cluster perspective
the metrics more. These changes are taking place with the offers a useful complement to the long-standing country-
cooperation of WIPO and other responsible international based rankings that will continue to form the core of
organizationsmost notably the UNESCO Institute for the GII.
Statisticswhich assist the country in question to resolve We thank our Knowledge Partners, the Confederation
issues and increase its data coverage. of Indian Industry (CII), PricewaterhouseCoopers
Innovation is not limited to the most advanced (PwC) and Strategy&, and the National Confederation
economies. Innovation is also not limited to the high- of Industry (CNI) and Servio Brasileiro de Apoio s
technology sectors. Innovation has today become a Micro e Pequenas Empresas (Sebrae) for their support of
global phenomenon, affecting all sectors of the economy, this years report.
including food and agriculturewhich are among the Likewise, we thank our prominent Advisory Board,
most ancient and basic sectors of activity. Feeding the which has been enriched by a new member this year:
world, while contributing to protecting the environment Chuan Poh Lim, Chairman, Agency for Science,
and providing quality and balanced nutrition to growing Technology and Research (A*STAR) of Singapore.
populations with different lifestyles and consumption pat- We hope that the collective efforts of innovation actors
terns, remains a complex challenge. Innovation has a key and decision makers who use the GII will continue to pave
role to play in addressing this challenge. the way for better innovation policies around the world.
The 2017 edition of the GII is dedicated to the theme of

innovation in agriculture and food systems. Agricultural

Dean, Cornell SC Johnson College of Business, Cornell University
and food-processing sectors continue to face an enor-
mous rise in global demand and increased competition FRANCIS GURRY
Director General, World Intellectual Property Organization (WIPO)
for limited natural resources. Innovation can be key to
maintaining the productivity growth required to meet BRUNO LANVIN
this rising demand in a sustainable fashion, and it can help Executive Director for Global Indices, INSEAD

enhance the networks that integrate food systems. This


Innovation as the Key Driver of

Sustainable Agriculture and Future
Food Security in the Developing World

The primary obligation of a nation is to protect its For last two years, in collaboration with the GII,
citizens from hunger and malnutrition by enabling CII has been engaged with the Indian government to
sustainable and equitable food production and distri- boost Indias ranking. I am delighted to report that this
bution channels. The developing world, characterized effort has improved Indias 2016 GII results. Another
by gross economic and social inequalities coupled with outcome of this sustained effort was the launch of the
inequitable access to safe, nutritional food and quality GII 2016 in India at a special event, jointly organized
healthcare, requires innovation to meet the ever-rising by the Department of Industrial Policy and Promotion
demand for food and to sustain its agricultural growth. (DIPP), the National Institution for Transforming India
Despite the fact that India is one of the worlds largest (NITI Aayog), and CII, in the presence of the Director-
producers of food grain, the largest producer of milk, General of the World Intellectual Property Organization
and its second largest producer of sugar, low-quality (WIPO). During the event, Indias Minister of State for
inputs such as low-grade seed, saline soil, inadequate Commerce and Industry instituted a high-level Task
irrigation, traditional farming methods (combined with Force on Innovation to suggest ways India can improve
small, scattered landholdings), restrictive access to for- its innovation eco-system.
mal credit, dependence on private moneylenders, and As a follow up to this launch, the first international
weak market linkages have long plagued its agriculture consultative exercise was organized in January 2017 in
sector. New Delhi to address existing data gaps in the GII.
To counter these challenges, a gradual infusion of International agencies such as UNESCO, among oth-
tech-based tools such as digital remote sensing, geo- ers, participated in the exercise where the first India
graphic and price information systems, crop and soil Innovation Index Portal was launched. These develop-
health monitoring, and farm management platforms has ments have created the desired momentum for states
taken place. These tools promise to rationalize processes to work on building their innovation ecosystems and
and enhance efficiency, productivity, distribution, and improving their innovation indicators.
access along the entire continuum of the food system In line with this years theme, Chapter 5 covers
from farm to fork and beyond. the current ecosystem of digital technologies in Indian
Public policy plays a pivotal role in making an envi- agriculturethe rise of agro-tech start-up ventures and
ronment conducive to this transition. The adoption of the advocacy initiatives that are the backbone needed to
innovation-led farm technologies has spurred public modernize Indian agriculture.
and private investments in R&D, helped technology CII has been a longstanding partner of GII. I would
transfer and uptake, as well as inter-sectoral cooperation. like to take this opportunity to congratulate the GII
Over the last two decades, this has enabled sustainable team once again for coming out with this important
agriculture to gradually gain momentum. edition, and for taking up a theme that resonates very
The Confederation of Indian Industry (CII) has been well in todays challenging times.

a strong proponent of this paradigm shift. The theme of

this years Global Innovation Index (GII), Innovation CHANDRAJIT BANERJEE
Director General
Feeding the World, thus resonates well with the agenda
Confederation of Indian Industry
and focus of CII in this sector, and like previous years
may prove benef icial for stimulating effective policy
dialogue within the government.

Innovating to Feed the World

We live in a world of finite resources but infinite passion support corporate ventures and to ensure that invest-
and creativity. At PwC, we are committed to building ments have a greater impact.
trust in society and solving important problems. But as In PwCs most recent CEO survey, we asked CEOs
problems become more global and complex, the solu- how the corporate community can help spread the
tions require a greater focus on innovation. The Global benefits of globalization more widely. The majority of
Innovation Index (GII) does just that by creating metrics them said the best way is to collaborate, particularly with
to evaluate innovation and by identifying new ways to government. As a GII Knowledge Partner, we hope to do
address the challenges that affect business and society. our part in helping to close the gap between innovation
At Strategy&, PwCs strategy consulting business, we and finding tangible solutions to important problems that
are proud to be part of the 2017 GII. This years theme affect communities around the world.
of innovation in food systems highlights one of the most
complex challenges humanity faces: managing the global TIM RYAN
US Chairman and Senior Partner
food supply. We know that without significantly expand- PwC
ing agricultural production over the next three decades,
the worlds population will increasingly face hunger,
malnutrition, and famine.
Resource scarcity is one of the key megatrends shap-
ing our world today and in the years to come, so meeting
the needs of the worlds people in a sustainable way will
require renewed focus on innovation in a variety of fields
and from a variety of stakeholders. In this case, addressing
global food insecurity involves technological innova-
tion, including leading-edge advances in data analytics;
global distribution and supply chain management; risk
assessment; economic f lexibility; a deeper understand-
ing of climate and weather conditions; and sustainability
practices. Its clear that no company, government, or any
other institution can solve the food crisis on its own. To
find a lasting solution, we have to work together.
In our research for the GII, we have identif ied
promising agricultural innovations being developed by
the private sector. Many of these are a result of more
corporate R&D investment in software and services,

and new technologies that are improving efficiency and

productivity. However, the public sectorwhich has
traditionally represented the majority of agricultural
R&D expenditurescontinues to play an important role
in spurring agricultural innovation. Theres a real oppor-
tunity for governments and businesses to collaborate to

Innovation in Food Production:

Learning from the Past with an
Open Mind for the Future

The National Confederation of Industry (CNI), the and improvements in cultivation techniques were the
Social Service of Industry (SESI), the National Service main drivers for the increase in agriculture productivity;
of Industrial Training (SENAI), and the Brazilian Micro today other challenges demand a new set of technologies
and Small Business Support Service (Sebrae) are more and policies.
and more concerned with innovation. We are convinced Agriculture and food production greatly impact the
that the only way to achieve sustainable development is environment. With the growing demand for agriculture
through innovation. Since 2008, CNI business leaders products, sustainable productivity growth in agriculture
have maintained the Entrepreneurial Mobilization for is a vital issue. This includes not only increasing crop
Innovation (MEI), putting innovation at the centre of productivity but also reducing inefficiencies in trans-
corporate strategy and enhancing the effectiveness of portation and food industrialization. Another significant
innovation policies in Brazil. issue relates to how best to adapt to climate change and
Innovation Feeding the World, this years theme the expected increased frequency of extreme weather
for the Global Innovation Index, is a central issue for events. New technologies could contribute a great deal
environmental sustainability and for the worlds social in this domain too.
and economic well-being. Innovations are spread across Fortunately, a vast array of new technologies prom-
different economic sectors, sustaining one another with ises to increase eff iciency in food production. New
new ideas and state-of-the-art technologies. Innovation equipment and devices are at the centre of such tech-
in agribusiness and food production now requires the nologies. Precision agriculture raises the possibility of
knowledge and technologies produced by other sectors. using knowledge and information technologies to adapt
Brazils role in grain production is not just a result of cultivation techniques to each specific location, with its
abundant natural resources and good climate conditions. own soil and climate characteristics. Crop sensors could
Historically, the country has developed a consistent and use agriculture inputs much more precisely by using
comprehensive system of research and development to the exact amount needed by a specific site. Drones and
support innovation and new agriculture technologies. robots have already automated several tasks in agricul-
This system benefits from the leadership of Embrapa ture production.
(Brazilian Agriculture Research Corporation), one of All these innovations are blurring the boundaries
the countrys most important public research enterprises, between industry, services, and agriculture. More and
which has provided Brazilian farmers with crucial tools more, industrial and service technologies are offering
needed for a modern and dynamic agroindustry. new possibilities in agriculture. These new possibilities
Inspired by Embrapa, in 2013 the government are also becoming more accessible to small innovative
launched the Brazilian Agency for Industrial Research businesses in all sectors. To seize the resulting oppor-
and Innovation (Embrapii), which manages non-refund- tunities, a new framework of policies and institutions is
able grants invested in projects carried out by compa- needed to take advantage of lessons learned from suc-

nies and research institutions and is acknowledged for cessful past experiences and envision new possibilities
its excellence, technological focus, and ability to meet for agriculture and food production. The theme of the
companies needs. Global Innovation Index this year could not be timelier.
The technology challenges for agro-industry are now
more complex than ever. In the past, soil fertilization, President, CNI; Director, SESI; President-Director, Sebrae
mechanization, plant breeding, genetic engineering, and President, SENAIs National Council

Contributors to the Report

The Global Innovation Index 2017: Innovation Feeding the World was developed under the general direction of Francis GURRY (Director General, World
Intellectual Property Organization), and the editors of the report, Soumitra DUTTA, Bruno LANVIN, and Sacha WUNSCH-VINCENT.

The report was prepared and coordinated by a core team comprising: KNOWLEDGE PARTNERS
Confederation of Indian Industry
Anjan DAS, Executive Director
Soumitra DUTTA, Dean, Cornell SC Johnson College of Business,
Cornell University Jibak DASGUPTA, Director

Rafael ESCALONA REYNOSO, GII Lead Researcher, Cornell SC CNI/Sebrae

Johnson College of Business, Cornell University Gianna SAGAZIO, Innovation Director, Innovation Directory, National
Confederation of Industry (CNI)
Jordan LITNER, GII Project Manager, Cornell SC Johnson College of
Business, Cornell University Suely LIMA, Innovation Manager, Innovation Directory, National
Confederation of Industry (CNI)
Bruno LANVIN, Executive Director for Global Indices, INSEAD
Julieta Costa CUNHA, Project Manager, Innovation Directory, National
Francesca GUADAGNO, Economist and Project Manager, Innovation
Confederation of Industry (CNI)
Economics Section, WIPO
Idenilza MIRANDA, Industrial Development Specialist, Innovation
Sacha WUNSCH-VINCENT, Senior Economist, Economics and
Directory, National Confederation of Industry (CNI)
Statistics Division, WIPO
Fernanda DE NEGRI, Consultant, Innovation Directory, National
Confederation of Industry (CNI)
The following people, institutions, and sources have supported the
production of the GII: Guilherme Afif DOMINGOS, Chief Executive Officer, Brazilian Micro
and Small Business Support Service (Sebrae)
Cornell University Heloisa MENEZES, Technical Director, Brazilian Micro and Small
Sarah MAGNUS-SHARPE, Director, PR & Media Relations, Cornell SC Business Support Service (Sebrae)
Johnson College of Business, Cornell University Vinicius LAGES, Chief Management and Financial Officer, Brazilian
INSEAD Micro and Small Business Support Service (Sebrae)
Christine HIRZEL, Global Head, Boards & External Relations Kelly SANCHES, Industry Unit Manager, Technical Directory, Brazilian
Sophie BADRE, Director, Media Relations Europe & Asia Micro and Small Business Support Service (Sebrae)

Virginie BONGEOT-MINET, Centre Coordinator Analuiza LOPES, Industry Unit Substitute, Manager, Technical
Directory, Brazilian Micro and Small Business Support Service (Sebrae)
Chris HOWELLS, Managing Editor, INSEAD Knowledge
Hugo Lumazzini PAIVA, Project Manager, Industry Unit, Technical
Aileen HUANG, Associate Director, Media Relations, Asia
Directory, Brazilian Micro and Small Business Support Service (Sebrae)
World Intellectual Property Organization (WIPO)
Augusto TOGNI, Agribusiness Unit Manager, Technical Directory,
Carsten FINK, Chief Economist, Economics and Statistics Division
Brazilian Micro and Small Business Support Service (Sebrae)
Economics and Statistics Division
Andrea RESTREPO, Analyst, Agribusiness Unit, Technical Directory,
Communications Division and External Relations Division Brazilian Micro and Small Business Support Service (Sebrae)

WIPO Bureaus, External Offices, and WIPO Coordination Office in Celio CABRAL, Inovation Unit Manager, Technical Directory, Brazilian
New York Micro and Small Business Support Service (Sebrae)
Printing Plant Athos RIBEIRO, Analyst, Inovation Unit, Technical Directory, Brazilian
Micro and Small Business Support Service (Sebrae)

(Continued on next page)


PricewaterhouseCoopers/Strategy& Dong GUO, Statistician; Rita LANG, Senior Statistical Assistant;


Alessandro BORGOGNA, Partner, PwC Middle East Jrgen MUTH, Senior Statistical Assistant; and Valentin TODOROV,
Barry JARUZELSKI, Principal, PwC US Senior Information Management Officer, all from the Statistics Division,
Department of Policy, Research and Statistics, United Nations Industrial
Thomas JOHNSON, Principal, PwC US
Development Organization (UNIDO)
Marcus MORAWIETZ, Partner, PwC Germany
Hctor HERNANDEZ, Project Leader Scoreboard, Territorial
Tim RYAN, US Chairman and Senior Partner, PwC US Development Unit; Alexander TBKE, Team Leader Industrial
Research & Innovation and Technology Analysis, Territorial
Ivan DE SOUZA, Partner, PwC Brazil
Development Unit, both from the European Commission, Joint
Volker STAACK, Principal, PwC US Research Centre, Directorate for Growth and Innovation
Steven VELDHOEN, Partner, PwC Japan Richard LAMBERT, Manager, Global Government IP Sales, Clarivate
Kiran CHAUHAN, Senior Manager, PwC Canada Analytics

Laura W. GELLER, Senior Manager, PwC US Ben SOWTER, Head of Division, QS Intelligence Unit, QS Quacquarelli
Symonds Ltd
Spencer HERBST, Manager, PwC US
Petra STEINER, Key Account Manager, Bureau van Dijk Electronic
Antanina GARANASVILI, PhD Candidate in Economics, University of
Susan TELTSCHER, Head a.i.; Esperanza MAGPANTAY, Senior
Padova and Queen Mary, University of London
Statistician; and Nathalie DELMAS, Assistant, all at the ICT Data and
Michaela SAISANA, Head of the Competence Centre on Composite Statistics Division (IDS), Telecommunication Development Bureau
Indicators & Scoreboards (COIN), European Commission, Joint (BDT), International Telecommunication Union (ITU)
Research Centre (JRC); and Sven LANGEDIJK, Head of Unit,
Padmasai VARANASI, Junior Economist-Statistician, Research and
Modelling, Indicators and Impact Evaluation, European Commission,
Public Policy, World Federation of Exchanges
Joint Research Centre (JRC)
Sad Ould A. VOFFAL, Programme Specialist, Elise LEGAULT,
Hope STEELE, Editor, Steele Editorial Services
Programme Specialist, Chiao-Ling CHIEN, Assistant Programme
Neil WEINBERG, Managing Member, Neil Weinberg Design Group LLC Specialist, and Imededdine JERBI, Statistician, Education Indicators
and Data Analysis Section; Lydia DELOUMEAUX, Assistant
DATA COLLABORATORS Programme Specialist, and Lisa BARBOSA, Statistical Assistant,
We are also grateful to the following persons/institutions for their Culture Unit; Talal EL HOURANI, Statistician, Education Survey
collaboration with specific data requests: Section; Martin SCHAAPER, Head of Section, Science, Culture and
David BESCOND, Statistician; Steven KAPSOS, Head of Unit; Yves Communication; Luciana MARINS and Rohan PATHIRAGE, Assistant
PERARDEL, Senior Econometrician; and Marie-Claire SODERGREN, Programme Specialists, and Zahia SALMI and Ghania DJAFRI,
Senior Economist, all at the Data Production and Analysis Unit (DPAU), Statistical Assistants, Science, Technology and Innovation Unit, all from
Department of Statistics, International Labour Office (ILO) the United Nations Educational, Scientific and Cultural Organization
(UNESCO) Institute for Statistics (UIS)
Mohsen BONAKDARPOUR, Managing Director, IHS Markit; Karen
CAMPBELL, Senior Consultant, IHS Markit Clement WOLF, Public Policy Manager and Ethan GAUVIN, Public
Policy Analyst, both at Google
Barbara DANDREA, Senior Statistician, International Trade Statistics
Section, and Adelina MENDOZA, Senior Statistical Officer, Market Leila ZIA, Senior Research Scientist, Research Team and Dan
Access Intelligence Section, both from the Economic Research and ANDREESCU, Senior Software Engineer, Analytics Team, both at
Statistics Division, World Trade Organization (WTO) Wikimedia Foundation

Klaas DE VRIES, Associate Economist and Bart VAN ARK, Executive Matthew ZOOK, Professor at the University of Kentucky and
Vice President, Chief Economist & Chief Strategy Officer, The President, ZookNIC Inc.
Conference Board Energy Data Centre, headed by Duncan MILLARD, International
Piet DONSELAAR, Senior Policy Advisor, Ministry of Economic Energy Agency (IEA)
Affairs, Innovation & knowledge Department, Directorate-General for United Nations Commodity Trade Statistics Database,
Enterprise and Innovation, the Netherlands Department of Economic and Social Affairs/ Statistics Division,
Fred GAULT, Professorial Fellow; UNU-MERIT, Professor Extraordinaire,
Tshwane University of Technology (TUT) in South Africa, Member of PwC Global entertainment and media outlook 20162010,
TUT Institute for Economic Research on Innovation

Thierry GEIGER, Head of Analytics and Quantitative Research,

and Ciara BROWNE, Head of Partnerships, both from Global
Competitiveness and Risks, Word Economic Forum

Advisory Board to the

Advisory Board
Global Innovation Index

In 2011, an Advisory Board was set up to provide advice ADVISORY BOARD MEMBERS
on the research underlying the Global Innovation Index
President, The Information Technology and Innovation Foundation
(GII), generate synergies at its stages of development, and (ITIF), United States of America
assist with the dissemination of its messages and results.
The Advisory Board is a select group of leading interna-
Director General of the United Nations Educational, Scientific and
tional practitioners and experts with unique knowledge Cultural Organization (UNESCO)
and skills in the realm of innovation. Its members, while
Dongmin CHEN
coming from diverse geographical and institutional Professor/Dean, School of Innovation and Entrepreneurship, and
backgrounds (international organizations, the public Director, Office of Business Development for Science and Technology,
sector, non-governmental organizations, business, and Peking University, China
academia), participate in their personal capacity. We Fabiola GIANOTTI
are grateful for the time and support provided by the Director-General of the European Organization for Nuclear Research (CERN)
Advisory Board members. Leonid GOKHBERG
In 2017, we welcome a new member to the Advisory First Vice-Rector, Higher School of Economics (HSE), and Director,
Board: Chuan Poh Lim, Chairman, Agency for Science, HSE Institute for Statistical Studies and Economics of Knowledge,
Russian Federation
Technology and Research (A*STAR).
Executive Member, Council for Science, Technology and Innovation,
Cabinet Office, Government of Japan

Senior Researcher, UNU-MERIT (Maastricht University)

Beethika KHAN
Program Director, National Science Foundation (NSF),
United States of America

Chuan Poh LIM

Chairman, Agency for Science, Technology and Research (A*STAR)

Raghunath Anant MASHELKAR

Chairman, National Innovation Foundation and President,
Global Research Alliance

Professor, NSW Chief Scientist and Engineer, Australia

Sibusiso SIBISI
President and Chief Executive Officer, Council for Scientific and
Industrial Research (CSIR), South Africa

Member of the Board of Directors of Ultrapar Participaes S.A. and of
Embraer S.A.; Chairman of the Board of Directors of the Brazilian Enterprise
for Research and Innovation (EMBRAPII) and of the Brazilian Association of
Innovative Companies (ANPEI)

Houlin ZHAO
Secretary-General, International Telecommunication Union (ITU)
Global Innovation Index 2017 rankings

Country/Economy Score (0100) Rank Income Rank Region Rank Efficiency Ratio Rank Median: 0.62
Switzerland 67.69 1 HI 1 EUR 1 0.95 2
Sweden 63.82 2 HI 2 EUR 2 0.83 12
Netherlands 63.36 3 HI 3 EUR 3 0.93 4
United States of America 61.40 4 HI 4 NAC 1 0.78 21
United Kingdom 60.89 5 HI 5 EUR 4 0.78 20
Denmark 58.70 6 HI 6 EUR 5 0.71 34
Singapore 58.69 7 HI 7 SEAO 1 0.62 63
Finland 58.49 8 HI 8 EUR 6 0.70 37
Germany 58.39 9 HI 9 EUR 7 0.84 7
Ireland 58.13 10 HI 10 EUR 8 0.85 6
Korea, Rep. 57.70 11 HI 11 SEAO 2 0.82 14
Luxembourg 56.40 12 HI 12 EUR 9 0.97 1
Iceland 55.76 13 HI 13 EUR 10 0.86 5
Japan 54.72 14 HI 14 SEAO 3 0.67 49
France 54.18 15 HI 15 EUR 11 0.71 35
Hong Kong (China) 53.88 16 HI 16 SEAO 4 0.61 73
Israel 53.88 17 HI 17 NAWA 1 0.77 23
Canada 53.65 18 HI 18 NAC 2 0.64 59
Norway 53.14 19 HI 19 EUR 12 0.66 51
Austria 53.10 20 HI 20 EUR 13 0.69 41
New Zealand 52.87 21 HI 21 SEAO 5 0.65 56
China 52.54 22 UM 1 SEAO 6 0.94 3
Australia 51.83 23 HI 22 SEAO 7 0.60 76
Czech Republic 50.98 24 HI 23 EUR 14 0.83 13
Estonia 50.93 25 HI 24 EUR 15 0.79 19
Malta 50.60 26 HI 25 EUR 16 0.84 8
Belgium 49.85 27 HI 26 EUR 17 0.67 47
Spain 48.81 28 HI 27 EUR 18 0.70 36
Italy 46.96 29 HI 28 EUR 19 0.73 31
Cyprus 46.84 30 HI 29 NAWA 2 0.74 28
Portugal 46.05 31 HI 30 EUR 20 0.71 33
Slovenia 45.80 32 HI 31 EUR 21 0.68 44
Latvia 44.61 33 HI 32 EUR 22 0.74 26
Slovakia 43.43 34 HI 33 EUR 23 0.75 25
United Arab Emirates 43.24 35 HI 34 NAWA 3 0.49 104
Bulgaria 42.84 36 UM 2 EUR 24 0.80 15
Malaysia 42.72 37 UM 3 SEAO 8 0.68 46
Poland 41.99 38 HI 35 EUR 25 0.67 48
Hungary 41.74 39 HI 36 EUR 26 0.73 30
Lithuania 41.17 40 HI 37 EUR 27 0.59 84
Croatia 39.80 41 HI 38 EUR 28 0.66 52
Romania 39.16 42 UM 4 EUR 29 0.69 39
Turkey 38.90 43 UM 5 NAWA 4 0.84 9
Greece 38.85 44 HI 39 EUR 30 0.56 87
Russian Federation 38.76 45 UM 6 EUR 31 0.61 75
Chile 38.70 46 HI 40 LCN 1 0.60 77
Viet Nam 38.34 47 LM 1 SEAO 9 0.84 10
Montenegro 38.07 48 UM 7 EUR 32 0.63 62
Qatar 37.90 49 HI 41 NAWA 5 0.61 68
Ukraine 37.62 50 LM 2 EUR 33 0.83 11
Thailand 37.57 51 UM 8 SEAO 10 0.75 24
Mongolia 37.13 52 LM 3 SEAO 11 0.74 27
Costa Rica 37.09 53 UM 9 LCN 2 0.69 43
Moldova, Rep. 36.84 54 LM 4 EUR 34 0.78 22
Saudi Arabia 36.17 55 HI 42 NAWA 6 0.53 96

Kuwait 36.10 56 HI 43 NAWA 7 0.79 18

South Africa 35.80 57 UM 10 SSF 1 0.53 97
Mexico 35.79 58 UM 11 LCN 3 0.61 74
Armenia 35.65 59 LM 5 NAWA 8 0.80 17
India 35.47 60 LM 6 CSA 1 0.66 53
TFYR of Macedonia 35.43 61 UM 12 EUR 35 0.59 80
Serbia 35.34 62 UM 13 EUR 36 0.61 67
Panama 34.98 63 UM 14 LCN 4 0.69 38
Mauritius 34.82 64 UM 15 SSF 2 0.48 109
Global Innovation Index 2017 rankings (continued)

Country/Economy Score (0100) Rank Income Rank Region Rank Efficiency Ratio Rank Median: 0.62
Colombia 34.78 65 UM 16 LCN 5 0.52 100
Bahrain 34.67 66 HI 44 NAWA 9 0.56 88
Uruguay 34.53 67 HI 45 LCN 6 0.59 82
Georgia 34.39 68 UM 17 NAWA 10 0.63 60
Brazil 33.10 69 UM 18 LCN 7 0.52 99
Peru 32.90 70 UM 19 LCN 8 0.49 106
Brunei Darussalam 32.89 71 HI 46 SEAO 12 0.34 124
Morocco 32.72 72 LM 7 NAWA 11 0.61 71
Philippines 32.48 73 LM 8 SEAO 13 0.65 55
Tunisia 32.30 74 LM 9 NAWA 12 0.62 65
Iran, Islamic Rep. 32.09 75 UM 20 CSA 2 0.80 16
Argentina 32.00 76 UM 21 LCN 9 0.55 94
Oman 31.83 77 HI 47 NAWA 13 0.46 115
Kazakhstan 31.50 78 UM 22 CSA 3 0.46 116
Dominican Republic 31.17 79 UM 23 LCN 10 0.65 54
Kenya 30.95 80 LM 10 SSF 3 0.66 50
Lebanon 30.64 81 UM 24 NAWA 14 0.61 69
Azerbaijan 30.58 82 UM 25 NAWA 15 0.50 103
Jordan 30.52 83 UM 26 NAWA 16 0.65 57
Jamaica 30.36 84 UM 27 LCN 11 0.57 86
Paraguay 30.30 85 UM 28 LCN 12 0.61 72
Bosnia and Herzegovina 30.23 86 UM 29 EUR 37 0.47 112
Indonesia 30.10 87 LM 11 SEAO 14 0.69 42
Belarus 29.98 88 UM 30 EUR 38 0.39 120
Botswana 29.97 89 UM 31 SSF 4 0.38 121
Sri Lanka 29.85 90 LM 12 CSA 4 0.65 58
Trinidad and Tobago 29.75 91 HI 48 LCN 13 0.56 90
Ecuador 29.14 92 UM 32 LCN 14 0.62 66
Albania 28.86 93 UM 33 EUR 39 0.37 122
Tajikistan 28.16 94 LM 13 CSA 5 0.59 83
Kyrgyzstan 28.01 95 LM 14 CSA 6 0.47 114
Tanzania, United Rep. 27.97 96 LI 1 SSF 5 0.73 29
Namibia 27.94 97 UM 34 SSF 6 0.48 108
Guatemala 27.90 98 LM 15 LCN 15 0.56 91
Rwanda 27.36 99 LI 2 SSF 7 0.33 125
Senegal 27.11 100 LI 3 SSF 8 0.54 95
Cambodia 27.05 101 LM 16 SEAO 15 0.63 61
Uganda 26.97 102 LI 4 SSF 9 0.47 113
El Salvador 26.68 103 LM 17 LCN 16 0.48 107
Honduras 26.36 104 LM 18 LCN 17 0.52 101
Egypt 26.00 105 LM 19 NAWA 17 0.59 81
Bolivia, Plurinational St. 25.64 106 LM 20 LCN 18 0.57 85
Mozambique 24.55 107 LI 5 SSF 10 0.61 70
Algeria 24.34 108 UM 35 NAWA 18 0.47 111
Nepal 24.20 109 LI 6 CSA 7 0.49 105
Ethiopia 24.16 110 LI 7 SSF 11 0.72 32
Madagascar 24.15 111 LI 8 SSF 12 0.68 45
Cte d'Ivoire 23.96 112 LM 21 SSF 13 0.69 40
Pakistan 23.80 113 LM 22 CSA 8 0.62 64
Bangladesh 23.72 114 LM 23 CSA 9 0.55 93
Malawi 23.45 115 LI 9 SSF 14 0.53 98
Benin 23.04 116 LI 10 SSF 15 0.47 110
Cameroon 22.58 117 LM 24 SSF 16 0.56 92
Mali 22.48 118 LI 11 SSF 17 0.60 78
Nigeria 21.92 119 LM 25 SSF 18 0.52 102

Burkina Faso 21.86 120 LI 12 SSF 19 0.24 127

Zimbabwe 21.80 121 LI 13 SSF 20 0.56 89
Burundi 21.31 122 LI 14 SSF 21 0.41 117
Niger 21.18 123 LI 15 SSF 22 0.36 123
Zambia 20.83 124 LM 26 SSF 23 0.59 79
Togo 18.41 125 LI 16 SSF 24 0.28 126
Guinea 17.41 126 LI 17 SSF 25 0.40 118
Yemen 15.64 127 LM 27 NAWA 19 0.40 119
Note: World Bank Income Group Classification (July 2016): LI = low income; LM = lower-middle income; UM = upper-middle income; and HI = high income. Regions are based on the United Nations Classification: EUR = Europe;
NAC = Northern America; LCN = Latin America and the Caribbean; CSA = Central and Southern Asia; SEAO = South East Asia, East Asia, and Oceania; NAWA = Northern Africa and Western Asia; SSF = Sub-Saharan Africa.
Key Findings

Key Findings of the GII 2017

Key Findings of the GII 2017

From the Global Innovation Index the productivity crisis is more topi- growth in low- and middle-income
2017, devoted to measuring the inno- cal today than ever. The downturn economies and lagging agricultural
vation performance of 127 economies has amplified the phenomenon of R&D spending across all economies
and the theme Innovation Feeding lacklustre productivity gains in rich both need to be reversed. Second,
the World, six messages emerge. countries, in conjunction with weak- innovations need to disperse more
Many of these messages are con- ened technological innovation and effectively throughout the agricul-
cerned with innovation as a driver of diffusion. Emerging economies are tural and food sector, especially in
growth generally. One is concerned affected as well, with their catch-up developing countries.
specifically with the role of innova- to advanced-country productivity Helping to meet this need for
tion as a way to address the growing slowing. innovation in agricultural systems, a
need for advances in agriculture and Research and development wave of new agricultural technolo-
food value chains. (R&D) investments need to be gies and innovations is taking place
intensified. Although permanently that could help overcome lagging
subdued R&D growth was avoided productivity. The pace of agricultural
Finding 1: Crafting the foundations thanks to countercyclical innova- innovation has increased over the
for innovation-driven growth while tion policies and private innova- last few years, with innovations from
the global economy is at an important tion expenditures, R&D growth other sectors spilling over to agricul-
turning point is still lower today than it was in tural and food systems. Advances in
In a turn of events, growth is reaching 201113, and much lower than in areas such as genetics and nano- and
a novel and more sustained momen- 200508 (FigureA). Tighter gov- biotechnologies have proven their
tum as the GII goes to print this year. ernment R&D budgets in selected ability to be a source of higher yields
Laying the foundation for innovation- high-income countries and slower and better nutrient content, even
driven economic development is ever spending growth in emerging coun- though their full environmental and
more paramount. Related policies tries explain part of this slowdown. health impacts have yet to be fully
that will sustain innovation invest- Disconcertingly, and in addition to understood. Big data are reshap-
ments can help transform the cyclical f lattening public R&D, business ing the world of agriculture: digital
economic upswing into longer-term research expenditures seem to be agriculture has started to spread
growth. Such proactive innovation losing momentum. worldwide, helped by the develop-
policies are also a powerful antidote ment of innovations in information
to uncertainty because they boost the technology (IT)for example, sen-
confidence and thus also the invest- Finding 2: Smart, digital agricultural sors, drones and robotics, and virtual
ments of economic actors into the innovation and a better uptake of and augmented realityas well as
future. innovation in developing countries can data generation and analytics enabled

In spite of this new growth help overcome serious food challenges by remote sensing, and geographic
momentum, investment and produc- Today a fresh innovation drive is information systems.
tivity growth are still at historic lows. required to confront slow growth in Unfortunately, the new wave
China aside, investment growth in agricultural productivity and the bot- of technological advances is rolling
middle-income countries has now tlenecks in todays agricultural inno- out rather slowly in many parts of
fallen to levels similar to that of rich vation systems. First and foremost, the world, including in rich coun-
countries (Figure A). Furthermore, lagging agricultural productivity tries. And developing countries,

Figure A: Global investment and business R&D falling short chain. These are mostly obstacles
Key Findings of the GII 2017

concerned with liquidity constraints,

agricultural inputs of imperfect
Investment growth, 200515 quality, insufficient information and
awareness, and a lack of post-harvest
and distribution infrastructure.
15 Public authorities have critical
roles to play in helping stimulate
innovation in food and agricultural
5 value chains. For a start, the agricul-
ture and food sector should be part

0 and parcel of any national innova-

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
tion strategy. To this day, this is very
rarely the case.
World To overcome market failures,
Middle income
policy makers have a responsibility
High income
to provide funding mechanisms to
stimulate innovation in agriculture
and food production. Instruments
such as agricultural funds and
focused research institutes need to
work more efficiently. Developing
countries also need to engage more
R&D expenditures growth, 200515
in domestic R&D, for example,
10 while setting priorities in research
f ields appropriate to their spe-
GDP growth
cific resources and contexts. Local
8 Business R&D
Total R&D (sub-national) initiatives are also
important: grassroots innovations
6 are happening in farming that can
often be scaled up. In such contexts,

robust links between public research

4 institutions, firms, and the grassroots
level are key.
Efforts to enhance the efficiency
of the food and agriculture innova-
tion system should focus on reducing
0 lags between R&D efforts and the
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 widespread adoption of agricultural
Source: See Figure 1 from Chapter 1. innovations. Accelerating technol-
ogy transfers by establishing clear
rules of engagement in university-
industry interactions, including the
commercialization of intellectual

property derived from these, is a

particularly in Sub-Saharan Africa, occur along the agricultural value valuable option. Supporting the
have yet to benefit from earlier waves chain, be it in high-income or low- demand for innovation from farmers
of agricultural innovations. income economies, not only in novel and commercial farming operations
New technologies aside, the brunt technologies. In the case of develop- is equally important. Five recom-
of agricultural innovation is found in ing countries, there are many sig- mendations are:
improved processes and services that nificant bottlenecks along the value

First, provide adequate informa- Table A: Innovation achievers: Income group and years as an innovation achiever

Key Findings of the GII 2017

tion to farmers, ensure that key
Economy Income group Years as an innovation achiever (total)
workers along the value chain
Viet Nam Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7)
have suff icient relevant skills,
Kenya Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7)
and encourage the adoption of
Moldova, Rep. Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7)
new products and processes.
India Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7)
Second, empower farmers by Armenia Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012 (6)
providing access to digital tech- Ukraine Lower-middle income 2017, 2016, 2015, 2014, 2012 (5)
nologies and the new service Rwanda Low income 2017, 2016, 2015, 2014, 2012 (5)
platforms that have immense Uganda Low income 2017, 2016, 2015, 2014, 2013 (5)
potential to positively impact Mozambique Low income 2017, 2016, 2015, 2014, 2012 (5)
agriculture. Malawi Low income 2017, 2016, 2015, 2014, 2012 (5)

Third, recognize and help boost Senegal Low income 2017, 2015, 2014, 2013, 2012 (5)

entrepreneurship and venture Tajikistan Lower-middle income 2017, 2016, 2013 (3)

capital approaches within the Malta High income 2017, 2016, 2015 (3)

agricultural sector. Madagascar Low income 2017, 2016 (2)

Bulgaria Upper-middle income 2017, 2015 (2)
Fourth, both the private sec- Burundi Low income 2017 (1)
tor and government can help Tanzania, United Rep. Low income 2017 (1)
infuse excellence and innova- Source: See Table 5 from Chapter 1.
tive attitudes that are evident in
other vital sectorssuch as the
information and communication
technologies, or ICT, sector
into the agricultural sector.
Finally, improve national legal take most of the top 25 spots, with countries, with China being the only
and regulatory frameworks in middle-income countries growing middle-income country closing the
agriculture, and more gener- more distant to the top 25 this year, gap.
ally streamline regulations and rather than closing the gap. In terms of regions, the same
reduce bu reauc r ac y a rou nd The exception is still China. It patterns of innovation divides per-
far mer s, in par t icu lar when moves up by three spots in this edi- sist: Northern America; Europe;
striking a balance between tra- tion, becoming the 22nd most inno- and South East Asia, East Asia, and
ditional and advanced farming vative economy in the world after Oceania lead, followed at a great
technologies. having entered the top 25 in 2016 distance by Northern Africa and
as the first middle-income economy. Western Asia; Latin America and the
With the exceptions of Bulgaria Caribbean; Central and Southern
Finding 3: More innovation convergence and Malaysia, the gap between the Asia; and, finally, Sub-Saharan
is needed globally, with developing 1125 ranked economies and middle- Africa.
countries perfecting their innovation income economies remains large, Yet there are many positive devel-
systems especially in Institutions, Human opments too. For a start, in 2017 we
Innovation is becoming more global capital and research, Infrastructure, continue to see a number of countries
but divides remain; innovation lead- and Creative outputs. Outside these that perform significantly better on
ers are uncontested at the top but new countries, only a few upper-middle- innovation than their current level

players are emerging. income economiessuch as Turkey, of development would predict; it is

Switzerland leads the rankings the Russian Federation, and Viet hoped that this will trigger a virtu-
for the seventh consecutive year. Namare among the top 50 this ous cycle of development in the years
In the top 25, some economies year. Similarly, the innovation qual- to come. A total of 17 economies
such as the Netherlands, Denmark, ity ranking is led by the United States compose the cluster of innovation
Germany, Japan, France, Israel, and of America (USA), Japan, the United achievers this year. This group has
Chinamove up. Yet rich countries Kingdom, and other high-income grown this year relative to 2016.

Most of these economiesnine in the 1980s, the so-called Asian A potentially stronger pan-Asian
Key Findings of the GII 2017

in totalcome from the Sub- Tigers emerged, with Hong Kong innovation network is seeing the
Saharan Africa region, followed (China), Singapore, the Republic of light of day as China, Japan, and
by three economies in the Eastern Korea, and to some extent Malaysia the Republic of Korea increasingly
region of Europe. Table A shows developing their innovation agendas conduct some of their manufactur-
the list of innovation achievers; quite rapidly. In conjunction with ing activitiesincluding those in
particularly notable is the consis- Japan, these economies are the top technology-intensive sectorsin
tent progress in Sub-Sahara Africa, Asian countries in innovation in neighbouring Asian countries, lead-
with some new economies, such as the region. In the 1990s, the rise ing to regional production and inno-
Tanzania and Burundi, joining this of other South East Asian countries vation networks. However, these
group. Importantly, Kenya, Rwanda, such as Thailand was also forecast by intra-regional production activities
Senegal, Uganda, Mozambique, and economic and innovation experts still mostly concern low-skill and
Malawi stand out for being innova- complementing the large established low-wage assembly operations with
tion achievers at least five times in the players. The economic spurt of these Chinese, Japanese, or Korean firms
previous six years. Particular results- countries was temporarily stopped choosing to manufacture in, for
oriented activities in Viet Nam and short by the Asian financial crisis, example, Viet Nam, to benefit from
India leading to achievements on but has since continued unabated. In excellent framework conditions and
particular innovation components addition, thanks to its steadily perse- lower wages. Few collaborative R&D
are also especially notable. vering innovation agenda, China also projects exist between the Asian
Continuing with the trend iden- vigorously entered the picture while leading nations, their top innovation
tified in earlier editions of the GII, making strides in terms of innovation clusters, or these smaller newcom-
the average performance of the group activities and results. ers today, at either the firm or the
of low-income economies is getting Moving forward, a novel dynamic country level. The newly emerging
closer to the average performance of of innovation development is in place Asian economies, such as Malaysia,
the middle-income cluster. Both in today, potentially producing a new the Philippines, and Viet Nam, still
GII scores and also in their catch-up line-up of up-and-coming Asian experience low R&D and low resi-
on particular innovation variables, countries. New Asian Tigerssuch dent patenting levels. As a result, the
the innovation achievers mentioned as Indonesia, the Philippines, and Viet potential of intra-regional innovation
in TableA help close the gap. Namare emerging too, and they networks in Asia is far from fully
increasingly join not only Asian high- utilized.
tech value chains but also other activ- There is development in Central
Finding 4: Opportunities have emerged ities such as ICT offshoring. These and Southern Asia too, with inter-
to leverage the rise of new East Asia and other countries in Asia are also esting developments in countries
Innovation Tigers, fostering deeper active in improving their innovation such as the Islamic Republic of Iran,
regional innovation networks and performance. Although Singapore is Kazakhstan, and Bangladesh. But,
benefitting from the rise of India still uncontested as number 1 among first and foremost, Indias current
In terms of innovation and economic the smaller or emerging Asian econo- and imminent development, and its
development more broadly, Asia is mies, countries such as Viet Nam, the contribution to the region and the
definitely a more and more impor- Philippines, and Thailand are rapidly global innovation landscape, is vital
tant engine of innovation in the 21st catching up. Among them, Viet Nam these days. As demonstrated in the
century, complementing existing tops education expenditure in the GII for some years, India has con-
innovation efforts in high-income region and does very well in ICT sistently outperformed on innova-
economies, mostly in Northern use, gross capital formation, and FDI tion relative to its GDP per capita.
America and Europe. net inf lows. Malaysia has the best Recently it made important strides in

The different elements of a poten- cluster development and ICT use, innovation input and output perfor-
tially strong networked innovation the Philippines leads ICT services mance. India is now in the top half
powerhouse are coming together in exports, Thailand tops the quality of the GII rankings. The continual
Asia. For a start, and despite enduring of publications and trademarks, and improvement of India in terms of
economic setbacks, Japan has contin- Cambodia only recently engaged investment, tertiary education, the
ued to be a driving force of global on innovation activities but its FDI quality of its publications and uni-
innovation since the late 1970s. Later, inf lows are already high. versities, its ICT services exports,

Key Findings of the GII 2017

India ahead of lower-middle and upper-middle income
Figure B: India ahead of average lower-middle- and upper-middle-income economies


l India
80 l Upper-middle income
l Lower-middle income

GII 2017 score



Quality of High- & Research Graduates State Gross Global Growth ICT High-tech GERD Intellectual Patent
scientific medium-high talent in in science & of capital R&D rate of services exports performed property families
publications tech business engineering cluster formation companies PPP$ exports by receipts in 2+
manufactures enterprise development GDP/worker business offices

Source: See Figure 6 in Chapter 1.

and its innovation clusters deserves Finding 5: Preserving the innovation outperforming every year since
mention (Figure B). It is to be hoped momentum in Sub-Saharan Africa and 2011including in the 2017 edition.
that India will continue on this tra- tapping the innovation potential in Latin Noted improvements in Insti-
jectory, with innovation investments America are priorities tutions and Business sophistication
leading to more and more dynamic A recurrent finding of the last edi- have allowed the region as a whole
R&D-intensive firms that are active tions of the GII has been that the to catch up with Central and South-
in patenting, high-technology pro- innovation momentum in Sub- ern Asia in these factors. Boosted
duction, and exports. If India then Saharan Africa must be preserved, by economies such as South Africa,
increasingly connects its innovation while countries in Latin America and Mauritius, Botswana, Namibia,
system to the innovative countries in the Caribbean are working to meet Rwanda, and Burkina Faso, Sub-
the East mentioned above, as well as their innovation potential. Saharan Africa this year has its
to standing innovation powerhouses For several editions, the GII has highest scores in Institutions and
in the West, it will make a true differ- noted thatrelative to its level of Market sophistication. Larger econ-
ence in Asias regional role in innova- economic developmentthe Sub- omies such as South Africa, Kenya,
tion, and to global innovation more Saharan Africa region performs Botswana, and Namibia help foster
generally. comparatively well on innovation. the expansion in Infrastructure;
This is a promising prospect. Since 2012, Sub-Saharan Africa others such as Mauritius, Rwanda,
The emergence of innovative new has had more countries among the Senegal, and Zimbabwe are helping
Asian Tigers, an innovative India, group of innovation achievers than to do so in Human capital.

and better innovation networks in any other region. Kenya, Rwanda, This year, however, the drivers of
the region are likely to be among the Senegal, Uganda, Mozambique, and growth that have been active in the
most encouraging developments for Malawi stand out for being innova- region have seen a slowdown. Clearly,
worldwide innovation in the next tion achievers at least five times in in absolute terms the gap between
few decades. the past six years. Kenya is the chief these Sub-Saharan Africa economies
innovation achiever in the region, and some South East Asian innovation

Table B: Top cluster of countries or cross-border regions, within the top 25 level to assess innovation clusters
Key Findings of the GII 2017

more broadly.
Rank Cluster name Territory(ies)
However, measuring the ter-
1 TokyoYokohama Japan
ritorial dimension of innovation
2 ShenzhenHong Kong (China) China/Hong Kong (China)
remains challenging. Only a few GII
3 San JoseSan Francisco, CA United States
indicators are readily available at the
4 Seoul Korea, Rep.
regional or city level for a large set
10 Paris France
of countries. Besides, clusters often
12 FrankfurtMannheim Germany
do not stop at national borders. By
18 Eindhoven Netherlands/Belgium definition, the search for official and
21 London United Kingdom timely innovation data is challenging.
22 Tel Aviv Israel In an effort to contribute prelimi-
24 Stockholm Sweden nary solutions, a novel approach is
Source: Derived from Table 1 in Annex 2 in the Special Section on Clusters. presented in the GII 2017 that iden-
tifies the largest inventive clusters
as measured by Patent Cooperation
Treaty (PCT) patenting. Drawing
on advanced mapping techniques
and WIPO patenting data, TableB
leaders is also still large, in particular country is still facing a high degree
shows some of the leading innovation
when one considers that integration of uncertainty.
clusters that result from this analysis.
of global value chains and innovation To further support this economic
TokyoYokohama, ShenzhenHong
exports, participation in high-tech upswing and help the region prog-
Kong (China), and San JoseSan
production and exports, and patent- ress in terms of innovation, sustained
Francisco (the Silicon Valley area in
ing by Sub-Saharan economies are efforts in improved innovation
California) lead in terms of being the
still low. investments and more coordinated
largest inventive clusters, based on
Turning to Latin America and innovation systems are required. Also
this methodology.
the Caribbean, more must be done needed is broader regional R&D and
In the coming years, attempts to
to reach the regions full innovation innovation cooperation, which is still
foster data on local innovation clusters
potential. Chile, Mexico, and Brazil largely absent when compared with
should receive increased attention,
and some other countries in the region other regions identified by the GII as
and may possibly become a more
are undoubtedly important innova- being successful in innovation.
important component of the GII.
tion actors. Mexico is also an active
contributor to global value chains,
including in high-tech sectors. It is Finding 6: The largest sub-national
notable, however, that there is more clusters of inventive activity, as
potential for broad regional improve- measured by patenting, include Tokyo
ment on innovation, both in terms of Yokohama, ShenzhenHong Kong
overall innovation performance and (China), and San JoseSan Francisco, CA
also in terms of key innovation vari- This year the GII makes a first attempt
ables such as scientific publications, at assessing sub-national innovation
R&D, and patenting. For example, clusters. The importance of innova-
in recent years and also in 2017, no tion hubs at the sub-national and
economies from this region are iden- international levels has been at the
tified as innovation achieversnone forefront of GII discussions for the

outperform in innovation relative last 10 years for two main reasons.

to their level of development. The First, successful innovation clusters
region as such has faced important are essential for national innovation
economic challenges in the last year, performance. Second, one of the
with Brazil only slowly emerging most frequent questions from coun-
from an economic recession accord- tries has been whether the GII model
ing to current forecasts, although the can be applied at the sub-national

The Global Innovation Index 2017: Innovation Feeding the World

1: The Global Innovation Index 2017

SOUMITRA DUTTA, RAFAEL ESCALONA REYNOSO, and J ORDAN LITNER , Cornell SC Johnson College of Business, Cornell University

Since the release of the Global

Innovation Index (GII) last year, Key findings in brief
the world has seen reason to expect
recovery and indeed renewed eco- The six key findings of the GII 2017 are:
nomic growth. Although uncertainty 1. Creating new sources of innovation- 4. The prospect of regional Asian
remains high, the holding pattern of driven growth is now vital to innovation networks will also benefit
the global economy might well give transforming the current economic from the rise of new Asian Innovation
way to a more sustained upswing. It is upswing into the possibility of longer- Tigers and Indias high potential.
term growth.
still questionable, however, whether 5. Preserving the momentum of
the foundations for continued 2. Smart and digital agricultural innovation in Sub-Saharan Africa and
growth are in place; the probability of innovation and better diffusion to tapping the innovation potential in
a low-growth scenario is still high. developing countries are required to Latin America and the Caribbean must
help overcome serious food challenges. be priorities.
In this context, firms, institutions,
and policy makers can help sustain 3. More innovation convergence is 6. Regional clusters of inventive activity
the recovery and shape the future by needed globally, as low- and middle- are essential to national innovation
creating novel sources of innovation- income countries put more emphasis performance; improved innovation
on their innovation systems. metrics on this topic are required.
driven growth.

Nourishing the welcome economic

upswing while tackling low investment
and productivity
The global economy has been in a a strong fourth quarter in 2016.1 African economies (Kenya, Senegal,
holding pattern for several years; it Compared with previous years, and Uganda); and also in a handful
has never fully recovered from the these growth forecasts for the world of large advanced economiesthe
200708 crisis and has never returned economy have not been revised Republic of Korea (Korea), the
to a momentum of sustained growth. downwards but upwards in recent United States of America (USA),
In recent years, initial optimism and months.2 Business and consumer and Canada. Brazil and the Russian
hopes of recovery were rather quickly confidence are high.3 Federation (Russia) are expected
replaced with downward revisions to Projections also indicate that to experience growth again, with
economic growth. The growth rates growth across low-, middle- and the former emerging from a deep
experienced before the economic high-income economies will be recession.4
crisis remain elusive. broad-based and positive. Growth As a region, Latin America and

As the new edition of the Global in emerging economies continues the Caribbean face more positive
Innovation Index 2017 goes to print, to be the main driver of the eco- prospects, following the stabilization
however, a new, if modest, growth nomic upswing. Economic growth of commodity prices benefitting
momentum is in place. The worlds is predicted to be relatively strong low- and middle-income economies
leading economic institutions pre- in middle- and certain low-income worldwide. Africa will experience a
dict a pick-up of global economic economies such as China, India, modest pick-up, boosted also by new
activity in 2017 and 2018, following Indonesia, and Thailand; a few infrastructure projects.5

Figure 1: Global investment and business R&D falling short more perpetual low-growth scenario,
1: The Global Innovation Index 2017

in which growth cannot be sustained

and increased over time. This report
Figure 1a: Investment growth, 200515
is concerned with two related bottle-
necks in particular.
15 First, investment and productiv-
ity increases are still at historic lows.
10 And these low levels are at the ori-
gin of the lower growth than was
enjoyed in pre-crisis years.6 Despite

0 more positive recent developments,

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
investmentespecially in emerging
and developing countrieshas not
World yet recovered.7 In fact, investment
Middle income growth in middle-income countries
High income
15 has fallen to levels similar to that
of rich countries. China aside, the
growth of investment in middle-
income countries is even lower
than it is in high-income ones (see
Figure1a). Furthermore, the produc-
Figure 1b: R&D expenditures growth, 200515 tivity crisis is more topical today than
10 ever.8 Potential measurement issues
aside, global labour productivity in
GDP growth
2016 is as low as it was in 2015.9 The
Business R&D
8 downturn, in conjunction with forces
Total R&D
that weakened technological innova-
tion and diffusion, has amplified the
phenomenon of lower productivity

in rich countries.10 In the meantime,

4 emerging economies are affected as
well, with catch-up to advanced-
country productivity levels slowing.
Second, concerns around falter-
ing global economic integration are
mounting. Trade growth has been
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 historically weak since 2010hov-
ering around 2.5% between 2013
Source: 1a. World Bank World Development Indicators database, March 2017; 1b. Authors estimate based on the UNESCO Institute for Statistics (UIS) database
and the IMF World Economic Outlook database, March 2017. and 2015and was even weaker in
Note: Investment refers to real gross fixed capital formation. 2016, when it fell to 1.3%.11 Cross-
border foreign direct investments
(FDI) also fell further in 2016.12
Trade in particular is traditionally
That said, growth rates expe- Furthermore, a number of factors seen as both an important cause and

rienced before the economic crisis might derail the scenario of a durable an effect of global growth. A more
remain distant for close to all coun- upswing. neglected aspect of economic inte-
tries. Some large emerging econo- Many monetary, fiscal, and other gration, however, is that both trade
mies, such as China, are seeing their factors are at stake, as well as unprec- and FDI are key channels of the dif-
high growth rates reduced, and other edented levels of geopolitical and fusion of technology, know-how, and
advanced economies, such as Japan, economic uncertainty. The leading innovation more broadly.13 A reversal
see persistently low growth rates. economic institutions are wary of a of globalized economic activity, and

1: The Global Innovation Index 2017

Box 1: Benchmarking R&D expenditures across countries

Global expenditures on R&D (GERD) follow- which data are available). These include tradi- the crisis. Finally, in a number of countries
ing the 200809 financial crisis have varied tionally high R&D spending economies, such such as Spain, Portugal, and FinlandR&D
considerably (see Tables 1.1 and 1.2 on the as the United States of America, the United expenditures (both total and business) have
following page). Some countriessuch as Kingdom, Germany, and the Netherlands, as not recovered yet, with GERD and BERD still
China, India, Mexico, the Russian Federation, well as relatively newer actors such as Chile below pre-crisis levels in 2015.
and Polanddid not decrease their R&D and Slovakia.
efforts during the crisis and have intensified In yet other countries (e.g., Colombia and Note
them further after the crisis, with business Norway), GERD did not fall during the crisis, Thanks to Antanina Garanasvili, PhD Candidate in
expenditures on R&D (BERD) also following but BERD did. Governments pushed R&D Economics, University of Padova and Queen Mary,
University of London, and our colleagues from the
the same trend. Other countries saw declining investments to compensate for lower busi-
UNESCO Institute for Statistics (UIS) Martin Schaaper
GERD and BERD during the crisis, but above ness R&D during the crisis; their efforts were and Rohan Pathirage for help in producing Box 1.
pre-crisis levels in 2015 (the latest year for rewarded with higher GERD and BERD after

(Continued on next page)

the associated networks of produc- more initiatives aimed at satisfying consistently been higher than GDP
tion and innovation, could have economists omnipresent calls for growth, also a ref lection of low gen-
adverse consequences for economic more infrastructure investment in eral GDP growth in that period. Still,
catch-up and technological leapfrog- economies across the board. about eight years after the crisis, the
ging, which have been historically To lay the foundation for future worst-case scenario of permanently
so critical for successful development growth, policy actions that foster reduced R&D growth has to date
cases such as China, Korea, and more human capital, research and develop- been avoided, thanks to these anticy-
recently Viet Nam.14 ment (R&D), and other innovation clical innovation policies and the role
Fortunately, trade, FDI, and pro- inputs and outputs, as captured by the of R&D champions such as China,
ductivity growth are also forecast to GII, are now required. Indeed, avail- Germany, and Korea, which have
be recovering in 2017 and further able economic evidence shows that consistently spent large and growing
increasing in 2018, in conjunc- an increase in R&D can effectively sums on R&D.
tion with output growth and the translate into an increase of GDP in Yet, although permanently sub-
cyclical recovery currently being the medium and longer term.17 dued R&D growth has been avoided,
experienced.15 Our study of global R&D data R&D growth is still inferior today
Policy initiatives to sustain invest- yields the following insights. Global than it was in 201113 immediately
ment, human capital, innovation, R&D growth fell in the aftermath of following the crisis, and much lower
and productivity growth could send the global financial crisis of 2009 (see than in 200508 when it averaged
a strong signal and be an important Figure1b and Box1).18 Governments around 6%. As governments have
antidote to uncertainty. stepped in to stimulate R&D effec- phased out some of their stimulus
tively. Business R&D investments programmes, and as spending cuts are
returned to faster growth in 2010. applied, tighter government R&D
Crafting the foundations for innovation- Encouragingly, by 2013 the share of budgets in selected high-income
driven growth as an antidote to business in total R&D had returned countries and slower spending
uncertainty to its pre-crisis levels. Broadly speak- growth in key emerging countries
Laying the foundations for innova- ing, our analysis indeed indicates explain part of this slowdown.20

tion-driven growth is paramount. that for the last four years, up until Disconcertingly, and in addition
Although not at levels seen after 2015 (when the most recent data are to f lattening public R&D, based on
the crisis, some government spend- available), global R&D intensity our estimates, business R&D growth
ing initiatives are underway again measured as global R&D expendi- seems to be losing momentum, with
in major economies; an uptick in tures relative to global GDPwas growth rates decreasing from about
investment will be felt in 2016 and at 1.7%, and thus at levels similar to 6% in 2013 to 5% in 2014 and about
2017.16 Still, there is room for even 200008.19 GERD growth has also 4.5% in 2015 (see Figure 1b).21 In
1: The Global Innovation Index 2017

Box 1: Benchmarking R&D expenditures across countries (continued)

Table 1.1: Gross domestic expenditure on R&D (GERD): Table 1.2: Business enterprise expenditure on R&D (BERD):
Crisis and recovery compared Crisis and recovery compared

Countries with no fall in GERD during the crisis that have expanded since Countries with no fall in BERD during the crisis that have expanded since

2008 2009 201012* 2013 2014 2015 2008 2009 201012* 2013 2014 2015
China 100 126 165 212 231 253 Poland 100 104 149 236 281 312
Poland 100 113 145 167 187 207 China 100 126 169 222 244 265
Costa Rica 100 134 140 166 179 n/a Costa Rica 100 114 102 174 216 n/a
Turkey 100 111 134 157 172 n/a Turkey 100 101 132 168 193 n/a
Colombia 100 100 118 174 167 166 Korea, Rep. 100 105 135 162 172 173
Korea, Rep. 100 106 133 155 166 168 Ireland 100 117 118 122 128 n/a
Mexico 100 105 113 117 127p 134p Mexico 100 112 111 107 115 122p
Norway 100 100 102 108 112 123 France 100 102 109 114 115 117p
Russian Fed. 100 111 107 114 118 118 Russian Fed. 100 110 102 109 112 111
India 100 106 120 n/a n/a n/a India 100 102 118 n/a n/a n/a

Countries with fall in GERD during the crisis but above pre-crisis levels in 2015 Countries with fall in BERD during the crisis but above pre-crisis levels in 2015

2008 2009 201012* 2013 2014 2015 2008 2009 201012* 2013 2014 2015
Slovakia 100 97 153 188 206 286 Colombia 100 73 106 139 172 179
Chile 100 93 103 126 125 130p Netherlands 100 93 119 129 135 138p
Israel 100d 96d 104d 115d 122d 124d Estonia 100 98 199 150 118 131p
Netherlands 100 99 111 116 121 124p Israel 100d 97d 105d 116d 124d 128d
Austria 100 97 108 117 121 123p Norway 100 97 100 107 114 125p
Brazil 100 99 112 124 121 n/a United Kingdom 100 97 101 107 113 118p
Germany 100 99 108 112 116 118p Germany 100 97 106 108 113 115
Singapore 100 82 95 101 114 n/a United States 100j 96j 96j 103j 107j 112j,p
United Kingdom 100 99 100 103 108 112p Chile 100 68 84 110 103 110p
United States 100j 99j 100j 104j 107j 111j,p Japan 100 88 93 99 104 103

GERD below crisis levels in 2015 BERD below crisis levels in 2015

2008 2009 201012* 2013 2014 2015 2008 2009 201012* 2013 2014 2015
Cuba 100 125 91 107 91 n/a Australia 100 96 97 98 n/a n/a
Romania 100 75 78 66 67 89 Sweden 100 90 88 92 87 97p
Iceland 100 98 90 68 79 89 Singapore 100 70 81 84 97 n/a
Spain 100 99 95 88 87 89 Canada 100g 99g 96g 90g 88g,p n/a
South Africa 100 93 86 89 n/a n/a Spain 100 93 90 85 84 85
Croatia 100 88 76 81 78 86 Portugal 100 100 92 80 77 78p
Portugal 100 106 97 85 83 83p South Africa 100 84 69 70 n/a n/a
Finland 100 97 97 88 84 77 Finland 100 93 91 81 77 69
Panama 100 70 80 45 n/a n/a Luxembourg 100 96 71 57 60 60
Uruguay 100 115 51 32 16 n/a

Source: OECD MSTI, February 2017; data used: Gross domestic expenditure on R&D (GERD) at constant 2010
PPP$, base year = 2008 (index 100). Source: OECD MSTI, February 2017; data used: Business enterprise expenditure on R&D (BERD) at constant
*Average values for the 2010 through 2012 period. Country data source is the UNESCO UIS database: 2010 PPP$, base year = 2008 (index 100).
UNESCO-UIS Science & Technology Data Center, update from March 2017. Data used: GERD in '000 PPP$ (in *Average values for the 2010 through 2012 period. Country data source is the UNESCO UIS database:
constant prices, 2005). UNESCO-UIS Science & Technology Data Center, update from March 2017. Data used: GERD, performed by
d = defence excluded (all or mostly); j = excludes most or all capital expenditure; p = provisional data. Business enterprise (in '000 PPP$, constant prices, 2005).
d = defence excluded (all or mostly); p = provisional data; g = excluding R&D in the social sciences and
humanities; j = excludes most or all capital expenditure.

several traditionally strong R&D Innovation feeding the world The situation is not improv-

1: The Global Innovation Index 2017

countries, including the USA, It is commonplace to equate innova- ing. Challenges such as rapidly
Germany, Japan, Korea, and China, tion with high-technology sectors. growing food demand, stagnating
business R&D growth is not rapid Yet the agriculture and food sec- farm incomes, diminishing natural
enough to offset the trends of zero tortraditionally considered low- resources, and climate change all
or negative growth elsewhere (see technologyis an important source aggravate the factors that contribute
Figure1b and Box1). of technological change, innovation, to issues of malnutrition worldwide.
The use of intellectual property and development. Today, more than Food security is more and more
(IP)a sign of continued innova- ever before, failure to perceive agri- affected because droughts, f loods,
tionhas intensified, albeit only in food systems as a source of innovation heat waves, and other extreme
selected middle- and high-income and to analyse their innovation input, weather events destroy agricultural
economies. The latest figures point outputs, linkages, and diffusion paths output. Risks of natural resource
to a 7.8% patent filing growth in accordingly would be a mistake. depletion and degradation call for
2015, much higher than it was in Agri-food systems face an unprec- intensified efforts towards greener,
the previous five years, yet that edented rise in global food demand more sustainable agricultural prac-
growth is mainly driven by China.22 while, at the same time, competition tices (see Chapters 3, 4, 5, and 9).
Turning to the future, as govern- for limited natural resources is at an Estimates indicate that global
ments prepare policies to sustain the all-time high. Feeding the world agricultural productivity and inno-
current growth momentum, a focus while simultaneously protecting the vation is not growing fast enough
on R&D and innovation should be a environment and providing balanced to meet future food demand, mostly
priority. Novel business practices or nutrition to growing populations because of the lagging total factor
new technologies could be potential remains a complex challenge. productivity growtha proxy for
triggers of much-needed productiv- innovationin low-income coun-
ity increases and engines of future Addressing the global food challenge tries (see Chapter 3).
economic growth. Historically, and The stakes of innovation in agricul- Innovation can help avert a global
to the present day, governments have ture and food are at least as high, if food crisis if policy makers and other
played an important role in building not higher, than in other fields. As actors change course on a global scale
human capital and driving research evidenced by the GII chapters this (see Box2).
as sponsors of basic or less applied year, progress in reducing malnutri-
R&D, as facilitators of private R&D tion is still too slow: Innovation in food and agriculture:
with tax reductions, or by exercising From moldboard plow to smart, digital
Global food demand in 2050 is
strong demand on innovation via agriculture
expected to increase by at least
government procurement or strate- The good news is that, historically,
60% above 2006 levels.25
gic initiatives.23 Governments might agricultural innovation has proven
need to boost their involvement to Around 795 million people in not only feasible but spectacularly
inspire business with the confidence the world, or about one in nine, successful, and has triggered key
to invest and innovate.24 suffer from hunger.26 structural and socioeconomic
As demonstrated by this years development.
About one in four people liv-
GII theme, these R&D and innova- Innovations in agriculture and
ing in Sub-Saharan Africa suf-
tion efforts are not and should not food production have been the start-
fers from chronic hunger, yet
be limited to sectors conventionally ing point of humanitys progress
the region with the largest num-
considered to be high-tech. For this towards organized social life. One can
ber of undernourished people is
reason, the 2017 GII edition on the think in particular of the moldboard
Southern Asia (281 million).27
theme of Innovation Feeding the plow and the cotton gin in the 18th

World focuses on innovation in One in three people in the world century; refrigeration in the 1850s;
agriculture and food systems and the is malnourished in one form or pasteurization in 1863; Mendels sci-
many scientific, technological, and another.28 entific plant breeding and the com-
other innovative advances made in bined harvester (early 20th century);
this field. and the green revolution in the 1950s,
which took millions out of hunger.29
1: The Global Innovation Index 2017

Box 2: Innovation, agriculture, and the UN 2030 Agenda for Sustainable Development

In September 2015, the Member States As this report shows, the achievement identifying resource needs and priorities. On
of the United Nations (UN) adopted the of Goal 2 (End hunger, achieve food security the basis of the GII, numerous workshops are
2030 Agenda for Sustainable Development, and improved nutrition and promote sustain- taking place in different countries to bring
incorporating 17 Sustainable Development able agriculture) will greatly benefit from innovation actors together with the aim of
Goals (SDGs) and 169 targets that are being innovation. The goal recognizes the role of improving data availability, boosting the coun-
implemented at the national level by the UN new technologies in boosting agricultural trys innovation performance, and design-
Member States to shape global development productivity and the need for public and ing strategic policy actions. Partnerships
in the period 201530. private investments in spurring technological are ongoing between the GII publishers and
The Agenda applies to all countries uni- change in this field. many UN partner organizationssuch as the
versally and aims at fostering social, environ- The SDGs and their associated targets International Telecommunication Union (ITU),
mental, and economic development. All the provide the basis for monitoring and review- the International Labour Organization (ILO),
SDGs rely to a greater or lesser extent upon ing countries progress in implementing sus- the United Nations Educational, Scientific
innovation for their means of implementation: tainable development at the global, regional, and Cultural Organization (UNESCO), and
Goal 9 (Build resilient infrastructure, promote and national levels. This process of review the United Nations Industrial Development
inclusive and sustainable industrialization and depends on a framework of statistical indica- Organization (UNIDO)as well as private data
foster innovation) makes explicit reference to tors being developed through an interna- providers to improve the required innovation
innovation and refers to several innovation tional consultative process led by the UN metrics.
factors referenced in the GII, such as infrastruc- Statistical Commission. In the process of implementing the 2030
ture, access to credit, access to information Disaggregated data are important for Agenda, the GII can provide countries with a
and communication technologies (ICTs) and monitoring and reviewing countries prog- data-based tool for policy making and con-
environmentally friendly technologies, sci- ress in implementing the SDGs as well as tribute to this shared endeavour of working
entific research, and technology capabilities. for assessing strengths and weaknesses and towards sustainable development globally.

As a result, agricultural pro- agricultural innovation systems (see irrigation systems, slowing down
ductivity has enjoyed periods of Chapters 7, 9, 10, and 11). their structural transformation and
extraordinary growth. From the First and foremost, lagging agri- development processes.31
1960s until the 1990s, the expansion cultural productivity growth in low- Indeed, in several developing
of land under cultivation and higher and middle-income countries and countries, productivity growth is
input useespecially in the form of lagging agricultural R&D spending still the result of expansions of cul-
fertilizers and high-yield varieties (public and/or private) across all tivated land and more intensive use
accounted for the bulk of agricultural economies (Chapter 3) need to be of inputs; technological change is
output growth (Chapter 3). Advances reversed. To reach that goal, both the having a much smaller impact in
in genetic engineering gave rise to a public and private sectors will need these countries (Chapter 3). Arable
wave of technological innovations to keep the R&D pipeline f lowing; land, however, cannot be expanded
and led the transition towards com- investments to ensure that innova- further because of growing urbaniza-
mercial agriculture in many regions. tive technologies and techniques are tion and environmental requirements
The green revolution enabled devel- brought to fruition are required. (Chapter 3). Concerns in these areas
oping economies to import cheaper Second, innovations need to are already materializing (see the cases
grains and grow crops with high- be better diffused throughout the of Russia and Uganda in Chapters 9
yield seed varieties, with tremendous agricultural and food sector, in and 11, respectively).
benefits for the economy and society particular in developing countries.
(see Chapters 5 and 10).30 Unfortunately, waves of technologi- A wave of smart agricultural

cal advances roll out rather slowly in innovations on the horizon

Stimulating investment in innovative many parts of the world. As a conse- Helping to meet this need for innova-
agriculture and food production quence, a number of developing coun- tion in agricultural systems, a wave
In the same spirit, today a new inno- tries, most notably in Sub-Saharan of new agricultural technologies and
vation drive is needed to confront Africa, have yet to benefit from ear- innovations is taking place that could
declining agricultural productiv- lier waves of agricultural innovations, help overcome lagging productivity.
ity and the bottlenecks of todays such as high-yield varieties and drip The pace of agricultural innovation

has increased over the last 10 years, or low-income economies (see to spur the consolidation of small

1: The Global Innovation Index 2017

with innovations from other sectors Table1 in Chapter 10 and Figure2 farms into large commercial farms
spilling over to agricultural and food in Chapter 11), and not only in also require innovation that makes
systems (see Chapters 3, 4, 5, and novel technologies. Activities along farm management more efficient, for
8). In the next decades, advances in the agri-food value chain range example (see Chapter 8).
biotechnology, autonomous vehicles, from supplying inputs such as seeds, Hence a mix of technological
and a broader shift of agricultural wholesalers, and retailer agro-dealers and non-technological innovation is
innovation to data, services, and to farming activities such as planting, required in agri-food value chains.
software could enable vital progress. farming, and harvesting and to post- Some technologies will need to
Rapid progress is underway in harvest activities such as bulking and diffuse and be adapted from rich
radically new technologies and new processing of raw output, branding countries to developing economies,
processes as applied to agricultural and marketing of value-added agri- while the latter are still adopting the
and food production. Advances in food products. Effective linkages technologies of the previous agricul-
areas such as genetics and nano- and and improved service delivery along tural innovation wave (genetically
biotechnologies have proven their this chain are just as critical, if not modified crops, drip irrigation, and
ability to be a source of higher yields more, than new technologies that can so on). At the same time, developing
and better nutrient content, even maximize the innovation potential in countries increasingly need to fur-
though their full environmental and agriculture. ther engage in their own domestic
health impacts have yet to be fully In the case of developing coun- R&Dfor example, they need to
understood. Chapter 9 mentions tries, there are many significant bot- pursue domestic seed varieties and
exciting examples of new-generation tlenecks along the value chain. These set research priorities fitting for their
sequencing, bioreactor-based syn- are mostly obstacles concerned with specific contexts, such as R&D in
thetic food production, total recy- liquidity constraints, agricultural aquaculture (see Chapter 9).
cling, bio-controlled and artificial inputs of imperfect quality, insuffi-
agroecosystems, and vertical farm- cient information and awareness, and Incentivizing agricultural innovation with
ing, to name a few such innovations a lack of post-harvest and distribution good institutions, stronger linkages, and
(see Table1 in Chapter 9 and also infrastructure (see Chapter 11). out-of-the-box thinking
Chapters 3, 4, 5 and 8). For example, most developing Public authorities have critical roles
An unprecedented convergence countries suffer from important to play in helping stimulate innova-
of biology, agronomy, plant and ani- weaknesses when it comes to benefit- tion in food and agriculture. For a
mal science, digitization, and robotics ting from inputs appropriate to their start, the agricultural and food sec-
is transforming the agri-food value particular circumstances, such as suit- tor should be part and parcel of any
chain. Big data are reshaping the world able seeds and services geared towards national innovation strategy (see
of agriculture: digital agriculture has the countrys context, such as finance Chapter 8 for Japans approach to
started to spread worldwide, helped and distribution (see, for example, the creating the project Technologies
by the development of innovations case of Uganda in Chapter 11). The for Creating Next-Generation
in information technology (IT)for financial sector provides an example: Agriculture, Forestry and Fisheries).
example, sensors, drones and robot- small rural farmers often face signifi- To this day, this is very rarely the
ics, and virtual and augmented real- cant barriers in accessing credits and case because innovation policies
ityas well as data generation and insurance. This reduces investment often focus on new sectors while
analytics enabled by remote sensing, while increasing households vulner- neglecting strengths in traditional or
and geographic information systems. ability (see also Chapter 3). resource-based industries.32
Organizational innovations are On this basis, the promotion of
Fostering innovation along the agricultural also as important as product or pro- specific activities that have the power

value chain, including in services and cess innovations. Digitization of retail to convince local players that prog-
processes and logistics, equipment-sharing, ress is feasible and desirable should be
New technologies aside, the brunt and life-long learning are examples undertaken.
of agricultural innovation is found of ways organizational innovations More traditionally, policy mak-
in improved processes and services can increase agricultural productivity ers have a responsibility to provide
that occur along the agricultural (Chapter 9). Complex organizational funding mechanisms to stimulate
value chain, be it in high-income changessuch as changes intended innovation in agriculture and food

production. The mechanisms can be between feeding the poor with mod- India (see Chapter 5). A f lurry of new
1: The Global Innovation Index 2017

in several forms: ern intensive agriculture and creating start-ups is on the rise, on par with
ground-breaking new crop varieties other high-technology sectors, and
For example, as seen in Brazil (see Chapter 8), while also looking at with ideas that can have an immedi-
(see Chapter 7), policy makers environmental issues and health. ate impact on societal well-being.
can create sectoral agricultural Cooperation and consultation Fourth, both the private sector
funds to foster technologies in remain a key ingredients needed to and government can also help infuse
areas such as agronomy, veteri- get popular support for the resulting excellence and innovative attitudes in
nary medicine, biotechnology, policies and to leave room for out-of- other vital sectors into the agriculture
econom ics, and ag r icu lt ura l the-box thinking. sector. In India, such an approach has
sociology; and to promote tech- helped enhance the impact of infor-
nological updates in the agri- Fostering skills and inspiring mation technology (IT) in unlocking
culture industry and stimulate agricultural entrepreneurship value for the grassroots level in areas
the expansion of investments in One of the key obstacles to the rapid such as mobile payments or health (see
tropical agricultural biotechnol- adoption of innovative approaches in Chapter 5). Over the last five years,
ogy and in the diffusion of new agriculture and food production still the Indian agriculture sector has also
technologies. is to be found in inadequate infor- attracted leading IT companies and
mation, a lack of skills, and, some- investors; available technology and
The creation of focused research
times, the lack of acceptance of new digital solutions are expanding at an
institutes (e.g., the Institute of
products or ways to produce them. impressive pace.
Innovation in Biotechnology in
Experiences from various parts of
Sao Paolo) is also a possibility
the world in this years GII chapters Scaling up local initiatives and
(see Chapter 7 on Brazil).
indicate how priorities need to be ensuring technology diffusion
Providing tax relief to enhance pursued in this area. Local (sub-national) initiatives are
far mer s incomes and of fer- First, agricultural extension also important: grassroots innova-
ing preferential access to land efforts to disseminate knowledge tions that can often be scaled up
and market support for promis- about new technologies and tech- are happening in low- and middle-
ing agricultural techniques and niques, and to demonstrate their busi- income economies farming. In
technologies is also a good way ness case, are required. These services such contexts, links between public
forward. include training in technology and research institutions, firms, and the
managerial skills and in the diffusion grassroots level are key.
Crafting balanced legal frameworks of information such as metrological Efforts to enhance the efficiency
Improving national legal and regu- data. This would provide adequate of the innovation system should focus
latory frameworks in and around information to farmers, ensure that on reducing lags between successful
agriculturefor example, by pro- key workers along the value chain R&D efforts and the widespread
moting the uptake of patents and have sufficient relevant skills, and adoption of agricultural innovations.
plant varieties; promoting the use encourage the adoption of new prod- In a number of countries (see Chapters
of trademarks, which can support ucts and processes. 9, 10, and 11), several factors
innovation; adopting public safety Second, farmers need to be including the lack of complementary
laws on biodiversity and genetically empowered by providing access to investments and capacityhamper
modified varieties; and more gen- digital technology and the new ser- spillovers from public research to
erally streamlining regulations and vice platforms that have immense enterprises. Accelerating technol-
reducing bureaucracy around farm- potential to positively impact agri- ogy transfers through the establish-
ersall contribute to a more condu- culture (see Chapters 3 and 5). ment of clear rules of engagement

cive environment (see Chapter 10). Third, entrepreneurship within in university-industry interactions,
Governments and policy makers the agriculture sector needs to be including the commercialization
also have the delicate task of providing recognized and inspired to a much of IP derived from these, is a good
a proper balance between inefficient more significant extent. In India, for option.33 Supporting the demand for
agriculture in need of more technol- example, venture capital has started innovation with farmers and com-
ogy, better fertilizers, and so on and f lowing to agricultural projects mercial farming operations is equally
advanced bio-farming, as well as through programmes such as Startup important.

Figure 2: Framework of the Global Innovation Index 2017

1: The Global Innovation Index 2017

Global Innovation Index

Innovation Efficiency Ratio


Innovation Input Innovation Output

Sub-Index Sub-Index

Human Knowledge and

capital and Market Business technology Creative
Institutions research Infrastructure sophistication sophistication outputs outputs

Political Knowledge Knowledge Intangible

environment Education ICTs Credit workers creation assets

Regulatory Tertiary General Innovation Knowledge Creative goods

environment education infrastructure Investment linkages impact and services

Business Research & Ecological competition, Knowledge Knowledge Online
environment development sustainability & market scale absorption diffusion creativity

More accurately measuring agricultural and which countries lead and lag in 92.5% of the worlds population and
innovation to simulate progress agricultural innovation. 97.6% of the worlds GDP (in current
Agriculture today is radically dif- A transition towards sustainable US dollars).
ferent from agriculture a couple of growth is paramount if the world Four measures are calculated: the
decades ago: more digital, smarter, is to cope successfully with the overall GII, the Input and Output
and more integrated. A better under- global challenges it is facing today. Sub-Indices, and the Innovation
standing of agricultural innovation Agriculture and food systems can Efficiency Ratio (Figure2).
in general, but these new forms of play a tremendous role in this, but
innovation in particular, is now cru- a concerted effort towards more The overall GII score is the
cial (Chapter 2 and Annex 4). Data granular agriculture-specif ic data simple average of the Input and
are needed to better inform decision collection is needed to understand Output Sub-Index scores.
makers about gaps and opportuni- what works and what does not, and
The Innovation Input Sub-
ties in agricultural capacity, and how governments and public poli-
Index is comprised of five input
to monitor and evaluate require- cies can help promote innovation in
pillars that capture elements of
ments and progress, recognizing agriculture and food.
the national economy that enable
the broader agricultural innovation
innovative activities: (1) Insti-
systemincluding informal actors,

tutions, (2) Human capital and

households, extension services, The GII 2017 conceptual framework
research, (3) Infrastructure, (4)
rural advisory services and farmer The GII helps to create an environ-
Market sophistication, and (5)
organizations, and the quantitative ment in which innovation factors are
Business sophistication.
and qualitative dimension of their continually evaluated. It provides a
interactions.34 Annex 4 describes key tool of detailed metrics for 127
available and missing data sources, economies this year, representing

The Innovation Output Sub- are not particularly sensitive to the 1. Switzerland
1: The Global Innovation Index 2017

I ndex provides infor mation missing values. As noted by the audit, 2. Sweden
about outputs that are the results this more stringent threshold has 3. Netherlands
of innovative activities within notably improved the confidence in 4. United States of America
the economy. There are two the country ranks for the GII and the 5. United Kingdom
output pillars: (6) Knowledge two sub-indices, and thus the reliabil- 6. Denmark
and technology outputs and (7) ity of the GII rankings (see Annex 3). 7. Singapore
Creative outputs. The rules on missing data and mini- 8. Finland
mum coverage per sub-pillar will be 9. Germany
The Innovation Eff iciency
progressively tightened, leading to 10. Ireland
Ratio is the ratio of the Output
the exclusion of countries that fail to
Sub-Index score over the Input
meet the desired minimum coverage Furthermore, stability remains
Sub-Index score. It shows how
in any sub-pillar (see Annex 2 for across the top 25 economies with
much innovation output a given
more details). only a few exceptions. China moves
country is getting for its inputs.
up by three places, becoming the
Each pillar is divided into three 22nd most innovative economy in the
sub-pillars and each sub-pillar is The Global Innovation Index 2017 results world after entering the top 25 in the
composed of individual indicators, The GII 2017 results have shown GII 2016. Israel gains four positions
for a total of 81 indicators this year. consistency in areas such as top this year, ranking 17th and swap-
Further details on the GII frame- rankings and the innovation divide. ping spots with New Zealand (21st).
work and the indicators used are However, there also have been some Other economies move up by two
provided in Annex 1. It is important new high-level developments as or more places: Japan (14th), France
to note that each year the variables described below. (15th), and Norway (19th). Australia
included in the GII computation are moves down four spots, ranking 23rd
reviewed and updated to provide Stability at the top, led by Switzerland, this year. Hong Kong (China) and
the best and most current assessment Sweden, and the Netherlands Canada each lose two or more posi-
of global innovation. Other meth- In 2017, the GII remains relatively tions, ranking 16th and 18th respec-
odological issuessuch as missing stable at the top. Switzerland leads the tively. The Czech Republic regains
data, revised scaling factors, and new rankings for the seventh consecutive its place in the top 25, gaining three
countries added to the samplealso year, while Sweden maintains its 2nd positions from last year and moving
impact year-on-year comparability of place. The Netherlands ranks 3rd, to 24th. Belgium leaves the top 25
the rankings (details of these changes although most of this improvement is this year, ranking 27th.
to the framework and factors impact- the result of methodological changes Box3 discusses the measure of
ing year-on-year comparability are and improved data availability. The innovation quality among GII 2017
provided in Annex 2). USA remains stable at the 4th spot, economies. Box 4 delves into the
Most notably, a more stringent while the UK moves down two innovation divide between the top
criterion for the inclusion of coun- positions to take 5th place. Denmark 25 ranked economies (24 of which
tries in the GII was adopted in 2016, improves another two positions are high-income) and the group of
following the Joint Research Centre this year, ranking 6th. Singapore, middle- and low-income economies.
( JRC) recommendation of past GII Finland, and Ireland move down,
audits (see Annex 3 in this report occupying the 7th, 8th, and 10th
and in previous years). Economies spots, respectively. Germany, which 2017 results: The worlds top innovators
and countries were included in the entered the top 10 in 2016, contin- The following section describes and
GII 2017 only if 66% of data were ues its advancement, moving up one analyses the prominent features of the

available within each of the two position from last year and occupy- GII 2017 results for the global lead-
sub-indices and if at least two of sub- ing the 9th spot. Hence, despite some ers in each component of the GII and
pillars in each pillar could be com- movement, the top 10 does not see the best performers in light of their
puted. This more stringent criterion any new entrant this year. income level.35 A short discussion
for inclusion in the GII ensures that Figure3 shows movement in the of the rankings at the regional level
country scores for the GII and for the top 10 ranked economies over the last follows.36
two Input and Output Sub-Indices four years:

Figure 3: Movement in the top 10 of the GII

1: The Global Innovation Index 2017

2014 2015 2016 2017
1 Switzerland Switzerland Switzerland Switzerland

2 United Kingdom United Kingdom Sweden Sweden

3 Sweden Sweden United Kingdom Netherlands

4 Finland Netherlands USA USA

5 Netherlands USA Finland United Kingdom

6 USA Finland Singapore Denmark

7 Singapore Singapore Ireland Singapore

8 Denmark Ireland Denmark Finland

9 Luxembourg Luxembourg Netherlands Germany

10 Hong Kong (China) Denmark Germany Ireland

Note: Year-on-year GII rank changes are influenced by performance and methodological considerations; see Annex 2.

Tables 1 through 3 on pages Education (28th), and Information (11th). It improves in the Innovation
1419 present the rankings of and communication technologies Input Sub-Index (2nd), with gains in
all economies included in the GII (ICTs, 30th). Switzerland ranks 1st all pillars but Market sophistication
2017 for the GII and the Input and in Knowledge creation and in a num- (10th). Among the largest improve-
Output Sub-Indices. ber of important indicators, including ments, Sweden gains 11 positions in
patent families in 2 or more offices, Innovation linkages (6th), 10 posi-
The top 10 in the Global Innovation Index PCT patent applications, and high- tions in Knowledge impact (10th),
Switzerland has earned the num- and medium-high-tech manufac- 7 positions in ICTs (13th), and 6
ber 1 position in the GII for the tures. With its favourable business positions in Knowledge absorption
seventh consecutive year. It has environment and solid innovation (7th). Its largest drops are in Tertiary
maintained this top spot since 2011, capabilities, Switzerland remains education (28th), Ecological sus-
as well as its number 1 position in highly successful in transforming its tainability (20th), Trade, competi-
the Innovation Output Sub-Index resources into more numerous, and tion, and market scale (28th), and
and in the Knowledge and technol- more varied, innovation outputs. Creative goods and services (18th).
ogy outputs pillar since 2012. Its lead Despite this strong performance, At the indicator level, Sweden keeps
seems largely uncontested. For the Switzerland presents a few areas of its 1st position in PCT patent appli-
first time it ranks among the top 10 weakness, especially on the input cations, while achieving a big leap
in all pillars and is the 3rd economy in side. These include ease of starting in labour productivity growth. It
the world in innovation quality (see a business, graduates in science and improves the most in governments
Box3). Thanks to its improvements engineering, gross capital formation, online service, e-participation, and

in Institutions (8th), Infrastructure ease of getting credit, and growth rate JV-strategic alliance deals, while
(6th), and Creative outputs (3rd), of GDP per worker. benef iting from the new measure
its Innovation Efficiency Ratio has Sweden holds the second high- averaging FDI net in-f lows (see
improved from 5th to 2nd. As in previ- est position in the GII, remaining the Annex 2). Areas of weakness include
ous years, it ranks among the top 25 in top Nordic economy and ranking pupil-teacher ratio, GDP per unit of
all sub-pillars, with only three excep- among the top 10 in all pillars with energy use, ease of getting credit,
tions: Business environment (33rd), the exception of Creative outputs FDI net inf lows, trademarks by
Table 1: Global Innovation Index rankings
1: The Global Innovation Index 2017

Country/Economy Score (0100) Rank Income Rank Region Rank Efficiency Ratio Rank Median: 0.62
Switzerland 67.69 1 HI 1 EUR 1 0.95 2
Sweden 63.82 2 HI 2 EUR 2 0.83 12
Netherlands 63.36 3 HI 3 EUR 3 0.93 4
United States of America 61.40 4 HI 4 NAC 1 0.78 21
United Kingdom 60.89 5 HI 5 EUR 4 0.78 20
Denmark 58.70 6 HI 6 EUR 5 0.71 34
Singapore 58.69 7 HI 7 SEAO 1 0.62 63
Finland 58.49 8 HI 8 EUR 6 0.70 37
Germany 58.39 9 HI 9 EUR 7 0.84 7
Ireland 58.13 10 HI 10 EUR 8 0.85 6
Korea, Rep. 57.70 11 HI 11 SEAO 2 0.82 14
Luxembourg 56.40 12 HI 12 EUR 9 0.97 1
Iceland 55.76 13 HI 13 EUR 10 0.86 5
Japan 54.72 14 HI 14 SEAO 3 0.67 49
France 54.18 15 HI 15 EUR 11 0.71 35
Hong Kong (China) 53.88 16 HI 16 SEAO 4 0.61 73
Israel 53.88 17 HI 17 NAWA 1 0.77 23
Canada 53.65 18 HI 18 NAC 2 0.64 59
Norway 53.14 19 HI 19 EUR 12 0.66 51
Austria 53.10 20 HI 20 EUR 13 0.69 41
New Zealand 52.87 21 HI 21 SEAO 5 0.65 56
China 52.54 22 UM 1 SEAO 6 0.94 3
Australia 51.83 23 HI 22 SEAO 7 0.60 76
Czech Republic 50.98 24 HI 23 EUR 14 0.83 13
Estonia 50.93 25 HI 24 EUR 15 0.79 19
Malta 50.60 26 HI 25 EUR 16 0.84 8
Belgium 49.85 27 HI 26 EUR 17 0.67 47
Spain 48.81 28 HI 27 EUR 18 0.70 36
Italy 46.96 29 HI 28 EUR 19 0.73 31
Cyprus 46.84 30 HI 29 NAWA 2 0.74 28
Portugal 46.05 31 HI 30 EUR 20 0.71 33
Slovenia 45.80 32 HI 31 EUR 21 0.68 44
Latvia 44.61 33 HI 32 EUR 22 0.74 26
Slovakia 43.43 34 HI 33 EUR 23 0.75 25
United Arab Emirates 43.24 35 HI 34 NAWA 3 0.49 104
Bulgaria 42.84 36 UM 2 EUR 24 0.80 15
Malaysia 42.72 37 UM 3 SEAO 8 0.68 46
Poland 41.99 38 HI 35 EUR 25 0.67 48
Hungary 41.74 39 HI 36 EUR 26 0.73 30
Lithuania 41.17 40 HI 37 EUR 27 0.59 84
Croatia 39.80 41 HI 38 EUR 28 0.66 52
Romania 39.16 42 UM 4 EUR 29 0.69 39
Turkey 38.90 43 UM 5 NAWA 4 0.84 9
Greece 38.85 44 HI 39 EUR 30 0.56 87
Russian Federation 38.76 45 UM 6 EUR 31 0.61 75
Chile 38.70 46 HI 40 LCN 1 0.60 77
Viet Nam 38.34 47 LM 1 SEAO 9 0.84 10
Montenegro 38.07 48 UM 7 EUR 32 0.63 62
Qatar 37.90 49 HI 41 NAWA 5 0.61 68
Ukraine 37.62 50 LM 2 EUR 33 0.83 11
Thailand 37.57 51 UM 8 SEAO 10 0.75 24
Mongolia 37.13 52 LM 3 SEAO 11 0.74 27
Costa Rica 37.09 53 UM 9 LCN 2 0.69 43
Moldova, Rep. 36.84 54 LM 4 EUR 34 0.78 22

Saudi Arabia 36.17 55 HI 42 NAWA 6 0.53 96

Kuwait 36.10 56 HI 43 NAWA 7 0.79 18
South Africa 35.80 57 UM 10 SSF 1 0.53 97
Mexico 35.79 58 UM 11 LCN 3 0.61 74
Armenia 35.65 59 LM 5 NAWA 8 0.80 17
India 35.47 60 LM 6 CSA 1 0.66 53
TFYR of Macedonia 35.43 61 UM 12 EUR 35 0.59 80
Serbia 35.34 62 UM 13 EUR 36 0.61 67
Panama 34.98 63 UM 14 LCN 4 0.69 38
Mauritius 34.82 64 UM 15 SSF 2 0.48 109
(Continued on next page)
Table 1: Global Innovation Index rankings (continued)

1: The Global Innovation Index 2017

Country/Economy Score (0100) Rank Income Rank Region Rank Efficiency Ratio Rank Median: 0.62
Colombia 34.78 65 UM 16 LCN 5 0.52 100
Bahrain 34.67 66 HI 44 NAWA 9 0.56 88
Uruguay 34.53 67 HI 45 LCN 6 0.59 82
Georgia 34.39 68 UM 17 NAWA 10 0.63 60
Brazil 33.10 69 UM 18 LCN 7 0.52 99
Peru 32.90 70 UM 19 LCN 8 0.49 106
Brunei Darussalam 32.89 71 HI 46 SEAO 12 0.34 124
Morocco 32.72 72 LM 7 NAWA 11 0.61 71
Philippines 32.48 73 LM 8 SEAO 13 0.65 55
Tunisia 32.30 74 LM 9 NAWA 12 0.62 65
Iran, Islamic Rep. 32.09 75 UM 20 CSA 2 0.80 16
Argentina 32.00 76 UM 21 LCN 9 0.55 94
Oman 31.83 77 HI 47 NAWA 13 0.46 115
Kazakhstan 31.50 78 UM 22 CSA 3 0.46 116
Dominican Republic 31.17 79 UM 23 LCN 10 0.65 54
Kenya 30.95 80 LM 10 SSF 3 0.66 50
Lebanon 30.64 81 UM 24 NAWA 14 0.61 69
Azerbaijan 30.58 82 UM 25 NAWA 15 0.50 103
Jordan 30.52 83 UM 26 NAWA 16 0.65 57
Jamaica 30.36 84 UM 27 LCN 11 0.57 86
Paraguay 30.30 85 UM 28 LCN 12 0.61 72
Bosnia and Herzegovina 30.23 86 UM 29 EUR 37 0.47 112
Indonesia 30.10 87 LM 11 SEAO 14 0.69 42
Belarus 29.98 88 UM 30 EUR 38 0.39 120
Botswana 29.97 89 UM 31 SSF 4 0.38 121
Sri Lanka 29.85 90 LM 12 CSA 4 0.65 58
Trinidad and Tobago 29.75 91 HI 48 LCN 13 0.56 90
Ecuador 29.14 92 UM 32 LCN 14 0.62 66
Albania 28.86 93 UM 33 EUR 39 0.37 122
Tajikistan 28.16 94 LM 13 CSA 5 0.59 83
Kyrgyzstan 28.01 95 LM 14 CSA 6 0.47 114
Tanzania, United Rep. 27.97 96 LI 1 SSF 5 0.73 29
Namibia 27.94 97 UM 34 SSF 6 0.48 108
Guatemala 27.90 98 LM 15 LCN 15 0.56 91
Rwanda 27.36 99 LI 2 SSF 7 0.33 125
Senegal 27.11 100 LI 3 SSF 8 0.54 95
Cambodia 27.05 101 LM 16 SEAO 15 0.63 61
Uganda 26.97 102 LI 4 SSF 9 0.47 113
El Salvador 26.68 103 LM 17 LCN 16 0.48 107
Honduras 26.36 104 LM 18 LCN 17 0.52 101
Egypt 26.00 105 LM 19 NAWA 17 0.59 81
Bolivia, Plurinational St. 25.64 106 LM 20 LCN 18 0.57 85
Mozambique 24.55 107 LI 5 SSF 10 0.61 70
Algeria 24.34 108 UM 35 NAWA 18 0.47 111
Nepal 24.20 109 LI 6 CSA 7 0.49 105
Ethiopia 24.16 110 LI 7 SSF 11 0.72 32
Madagascar 24.15 111 LI 8 SSF 12 0.68 45
Cte d'Ivoire 23.96 112 LM 21 SSF 13 0.69 40
Pakistan 23.80 113 LM 22 CSA 8 0.62 64
Bangladesh 23.72 114 LM 23 CSA 9 0.55 93
Malawi 23.45 115 LI 9 SSF 14 0.53 98
Benin 23.04 116 LI 10 SSF 15 0.47 110
Cameroon 22.58 117 LM 24 SSF 16 0.56 92
Mali 22.48 118 LI 11 SSF 17 0.60 78
Nigeria 21.92 119 LM 25 SSF 18 0.52 102

Burkina Faso 21.86 120 LI 12 SSF 19 0.24 127

Zimbabwe 21.80 121 LI 13 SSF 20 0.56 89
Burundi 21.31 122 LI 14 SSF 21 0.41 117
Niger 21.18 123 LI 15 SSF 22 0.36 123
Zambia 20.83 124 LM 26 SSF 23 0.59 79
Togo 18.41 125 LI 16 SSF 24 0.28 126
Guinea 17.41 126 LI 17 SSF 25 0.40 118
Yemen 15.64 127 LM 27 NAWA 19 0.40 119
Note: World Bank Income Group Classification (July 2016): LI = low income; LM = lower-middle income; UM = upper-middle income; and HI = high income. Regions are based on the United Nations Classification: EUR = Europe;
NAC = Northern America; LCN = Latin America and the Caribbean; CSA = Central and Southern Asia; SEAO = South East Asia, East Asia, and Oceania; NAWA = Northern Africa and Western Asia; SSF = Sub-Saharan Africa.
Table 2: Innovation Input Sub-Index rankings
1: The Global Innovation Index 2017

Country/Economy Score (0100) Rank Income Rank Region Rank Median: 43.15
Singapore 72.25 1 HI 1 SEAO 1
Sweden 69.72 2 HI 2 EUR 1
Switzerland 69.60 3 HI 3 EUR 2
Finland 68.93 4 HI 4 EUR 3
United States of America 68.87 5 HI 5 NAC 1
Denmark 68.68 6 HI 6 EUR 4
United Kingdom 68.25 7 HI 7 EUR 5
Hong Kong (China) 66.95 8 HI 8 SEAO 2
Netherlands 65.79 9 HI 9 EUR 6
Canada 65.57 10 HI 10 NAC 2
Japan 65.45 11 HI 11 SEAO 3
Australia 64.61 12 HI 12 SEAO 4
New Zealand 64.14 13 HI 13 SEAO 5
Norway 63.99 14 HI 14 EUR 7
France 63.41 15 HI 15 EUR 8
Korea, Rep. 63.34 16 HI 16 SEAO 6
Germany 63.33 17 HI 17 EUR 9
Austria 62.92 18 HI 18 EUR 10
Ireland 62.86 19 HI 19 EUR 11
Israel 61.01 20 HI 20 NAWA 1
Iceland 60.10 21 HI 21 EUR 12
Belgium 59.53 22 HI 22 EUR 13
United Arab Emirates 57.96 23 HI 23 NAWA 2
Luxembourg 57.36 24 HI 24 EUR 14
Spain 57.28 25 HI 25 EUR 15
Estonia 56.99 26 HI 26 EUR 16
Czech Republic 55.72 27 HI 27 EUR 17
Malta 54.91 28 HI 28 EUR 18
Italy 54.43 29 HI 29 EUR 19
Slovenia 54.40 30 HI 30 EUR 20
China 54.22 31 UM 1 SEAO 7
Cyprus 53.92 32 HI 31 NAWA 3
Portugal 53.80 33 HI 32 EUR 21
Lithuania 51.92 34 HI 33 EUR 22
Latvia 51.25 35 HI 34 EUR 23
Malaysia 50.94 36 UM 2 SEAO 8
Poland 50.20 37 HI 35 EUR 24
Greece 49.73 38 HI 36 EUR 25
Slovakia 49.66 39 HI 37 EUR 26
Brunei Darussalam 49.27 40 HI 38 SEAO 9
Hungary 48.36 41 HI 39 EUR 27
Chile 48.31 42 HI 40 LCN 1
Russian Federation 48.21 43 UM 3 EUR 28
Croatia 47.96 44 HI 41 EUR 29
Bulgaria 47.61 45 UM 4 EUR 30
Saudi Arabia 47.33 46 HI 42 NAWA 4
Mauritius 47.13 47 UM 5 SSF 1
Qatar 46.96 48 HI 43 NAWA 5
South Africa 46.85 49 UM 6 SSF 2
Montenegro 46.83 50 UM 7 EUR 31
Romania 46.36 51 UM 8 EUR 32
Colombia 45.75 52 UM 9 LCN 2
TFYR of Macedonia 44.53 53 UM 10 EUR 33
Mexico 44.52 54 UM 11 LCN 3

Bahrain 44.41 55 HI 44 NAWA 6

Peru 44.21 56 UM 12 LCN 4
Costa Rica 43.97 57 UM 13 LCN 5
Serbia 43.79 58 UM 14 EUR 34
Botswana 43.58 59 UM 15 SSF 3
Brazil 43.47 60 UM 16 LCN 6
Uruguay 43.47 61 HI 45 LCN 7
Oman 43.46 62 HI 46 NAWA 7
Belarus 43.24 63 UM 17 EUR 35
Kazakhstan 43.15 64 UM 18 CSA 1
(Continued on next page)
Table 2: Innovation Input Sub-Index rankings (continued)

1: The Global Innovation Index 2017

Country/Economy Score (0100) Rank Income Rank Region Rank Median: 43.15
Thailand 42.92 65 UM 19 SEAO 10
India 42.84 66 LM 1 CSA 2
Mongolia 42.71 67 LM 2 SEAO 11
Turkey 42.32 68 UM 20 NAWA 8
Georgia 42.16 69 UM 21 NAWA 9
Albania 42.03 70 UM 22 EUR 36
Viet Nam 41.75 71 LM 3 SEAO 12
Argentina 41.38 72 UM 23 LCN 8
Moldova, Rep. 41.35 73 LM 4 EUR 37
Panama 41.28 74 UM 24 LCN 9
Bosnia and Herzegovina 41.14 75 UM 25 EUR 38
Rwanda 41.07 76 LI 1 SSF 4
Ukraine 41.05 77 LM 5 EUR 39
Azerbaijan 40.70 78 UM 26 NAWA 10
Morocco 40.59 79 LM 6 NAWA 11
Kuwait 40.30 80 HI 47 NAWA 12
Tunisia 39.99 81 LM 7 NAWA 13
Armenia 39.71 82 LM 8 NAWA 14
Philippines 39.40 83 LM 9 SEAO 13
Jamaica 38.69 84 UM 27 LCN 10
Trinidad and Tobago 38.22 85 HI 48 LCN 11
Kyrgyzstan 38.16 86 LM 10 CSA 3
Lebanon 37.99 87 UM 28 NAWA 15
Dominican Republic 37.80 88 UM 29 LCN 12
Namibia 37.76 89 UM 30 SSF 5
Paraguay 37.62 90 UM 31 LCN 13
Kenya 37.19 91 LM 11 SSF 6
Jordan 37.07 92 UM 32 NAWA 16
Uganda 36.71 93 LI 2 SSF 7
Sri Lanka 36.28 94 LM 12 CSA 4
Ecuador 36.07 95 UM 33 LCN 14
El Salvador 36.06 96 LM 13 LCN 15
Guatemala 35.86 97 LM 14 LCN 16
Iran, Islamic Rep. 35.71 98 UM 34 CSA 5
Indonesia 35.68 99 LM 15 SEAO 14
Tajikistan 35.50 100 LM 16 CSA 6
Burkina Faso 35.28 101 LI 3 SSF 8
Senegal 35.23 102 LI 4 SSF 9
Honduras 34.77 103 LM 17 LCN 17
Cambodia 33.19 104 LM 18 SEAO 15
Algeria 33.12 105 UM 35 NAWA 17
Egypt 32.69 106 LM 19 NAWA 18
Bolivia, Plurinational St. 32.62 107 LM 20 LCN 18
Nepal 32.51 108 LI 5 CSA 7
Tanzania, United Rep. 32.31 109 LI 6 SSF 10
Benin 31.30 110 LI 7 SSF 11
Niger 31.18 111 LI 8 SSF 12
Malawi 30.75 112 LI 9 SSF 13
Bangladesh 30.64 113 LM 21 CSA 8
Mozambique 30.45 114 LI 10 SSF 14
Burundi 30.21 115 LI 11 SSF 15
Pakistan 29.43 116 LM 22 CSA 9
Cameroon 29.03 117 LM 23 SSF 16
Nigeria 28.94 118 LM 24 SSF 17
Togo 28.81 119 LI 12 SSF 18

Madagascar 28.78 120 LI 13 SSF 19

Cte d'Ivoire 28.39 121 LM 25 SSF 20
Ethiopia 28.16 122 LI 14 SSF 21
Mali 28.14 123 LI 15 SSF 22
Zimbabwe 27.98 124 LI 16 SSF 23
Zambia 26.14 125 LM 26 SSF 24
Guinea 24.86 126 LI 17 SSF 25
Yemen 22.38 127 LM 27 NAWA 19
Note: World Bank Income Group Classification (July 2016): LI = low income; LM = lower-middle income; UM = upper-middle income; and HI = high income. Regions are based on the United Nations Classification: EUR = Europe;
NAC = Northern America; LCN = Latin America and the Caribbean; CSA = Central and Southern Asia; SEAO = South East Asia, East Asia, and Oceania; NAWA = Northern Africa and Western Asia; SSF = Sub-Saharan Africa.
Table 3: Innovation Output Sub-Index rankings
1: The Global Innovation Index 2017

Country/Economy Score (0100) Rank Income Rank Region Rank Median: 25.60
Switzerland 65.78 1 HI 1 EUR 1
Netherlands 60.92 2 HI 2 EUR 2
Sweden 57.92 3 HI 3 EUR 3
Luxembourg 55.43 4 HI 4 EUR 4
United States of America 53.93 5 HI 5 NAC 1
United Kingdom 53.52 6 HI 6 EUR 5
Germany 53.46 7 HI 7 EUR 6
Ireland 53.41 8 HI 8 EUR 7
Korea, Rep. 52.06 9 HI 9 SEAO 1
Iceland 51.42 10 HI 10 EUR 8
China 50.87 11 UM 1 SEAO 2
Denmark 48.71 12 HI 11 EUR 9
Finland 48.06 13 HI 12 EUR 10
Israel 46.75 14 HI 13 NAWA 1
Malta 46.29 15 HI 14 EUR 11
Czech Republic 46.24 16 HI 15 EUR 12
Singapore 45.14 17 HI 16 SEAO 3
France 44.94 18 HI 17 EUR 13
Estonia 44.87 19 HI 18 EUR 14
Japan 43.99 20 HI 19 SEAO 4
Austria 43.27 21 HI 20 EUR 15
Norway 42.29 22 HI 21 EUR 16
Canada 41.73 23 HI 22 NAC 2
New Zealand 41.59 24 HI 23 SEAO 5
Hong Kong (China) 40.81 25 HI 24 SEAO 6
Spain 40.34 26 HI 25 EUR 17
Belgium 40.17 27 HI 26 EUR 18
Cyprus 39.75 28 HI 27 NAWA 2
Italy 39.50 29 HI 28 EUR 19
Australia 39.06 30 HI 29 SEAO 7
Portugal 38.30 31 HI 30 EUR 20
Bulgaria 38.08 32 UM 2 EUR 21
Latvia 37.97 33 HI 31 EUR 22
Slovenia 37.21 34 HI 32 EUR 23
Slovakia 37.20 35 HI 33 EUR 24
Turkey 35.48 36 UM 3 NAWA 3
Hungary 35.13 37 HI 34 EUR 25
Viet Nam 34.92 38 LM 1 SEAO 8
Malaysia 34.49 39 UM 4 SEAO 9
Ukraine 34.19 40 LM 2 EUR 26
Poland 33.78 41 HI 35 EUR 27
Moldova, Rep. 32.33 42 LM 3 EUR 28
Thailand 32.22 43 UM 5 SEAO 10
Romania 31.95 44 UM 6 EUR 29
Kuwait 31.91 45 HI 36 NAWA 4
Croatia 31.63 46 HI 37 EUR 30
Armenia 31.60 47 LM 4 NAWA 5
Mongolia 31.55 48 LM 5 SEAO 11
Lithuania 30.42 49 HI 38 EUR 31
Costa Rica 30.20 50 UM 7 LCN 1
Russian Federation 29.31 51 UM 8 EUR 32
Montenegro 29.30 52 UM 9 EUR 33
Chile 29.09 53 HI 39 LCN 2
Qatar 28.84 54 HI 40 NAWA 6

Panama 28.67 55 UM 10 LCN 3

United Arab Emirates 28.52 56 HI 41 NAWA 7
Iran, Islamic Rep. 28.47 57 UM 11 CSA 1
India 28.11 58 LM 6 CSA 2
Greece 27.96 59 HI 42 EUR 34
Mexico 27.07 60 UM 12 LCN 4
Serbia 26.90 61 UM 13 EUR 35
Georgia 26.61 62 UM 14 NAWA 8
TFYR of Macedonia 26.32 63 UM 15 EUR 36
Uruguay 25.60 64 HI 43 LCN 5
(Continued on next page)
Table 3: Innovation Output Sub-Index rankings (continued)

1: The Global Innovation Index 2017

Country/Economy Score (0100) Rank Income Rank Region Rank Median: 25.60
Philippines 25.57 65 LM 7 SEAO 12
Saudi Arabia 25.00 66 HI 44 NAWA 9
Bahrain 24.92 67 HI 45 NAWA 10
Morocco 24.85 68 LM 8 NAWA 11
South Africa 24.74 69 UM 16 SSF 1
Kenya 24.71 70 LM 9 SSF 2
Tunisia 24.62 71 LM 10 NAWA 12
Dominican Republic 24.54 72 UM 17 LCN 6
Indonesia 24.52 73 LM 11 SEAO 13
Jordan 23.96 74 UM 18 NAWA 13
Colombia 23.82 75 UM 19 LCN 7
Tanzania, United Rep. 23.63 76 LI 1 SSF 3
Sri Lanka 23.42 77 LM 12 CSA 3
Lebanon 23.28 78 UM 20 NAWA 14
Paraguay 22.99 79 UM 21 LCN 8
Brazil 22.72 80 UM 22 LCN 9
Argentina 22.62 81 UM 23 LCN 10
Mauritius 22.51 82 UM 24 SSF 4
Ecuador 22.20 83 UM 25 LCN 11
Jamaica 22.03 84 UM 26 LCN 12
Peru 21.60 85 UM 27 LCN 13
Trinidad and Tobago 21.27 86 HI 46 LCN 14
Cambodia 20.91 87 LM 13 SEAO 14
Tajikistan 20.81 88 LM 14 CSA 4
Azerbaijan 20.46 89 UM 28 NAWA 15
Oman 20.19 90 HI 47 NAWA 16
Ethiopia 20.16 91 LI 2 SSF 5
Guatemala 19.93 92 LM 15 LCN 15
Kazakhstan 19.85 93 UM 29 CSA 5
Cte d'Ivoire 19.53 94 LM 16 SSF 6
Madagascar 19.53 95 LI 3 SSF 7
Bosnia and Herzegovina 19.32 96 UM 30 EUR 37
Egypt 19.31 97 LM 17 NAWA 17
Senegal 18.98 98 LI 4 SSF 8
Bolivia, Plurinational St. 18.66 99 LM 18 LCN 16
Mozambique 18.64 100 LI 5 SSF 9
Pakistan 18.16 101 LM 19 CSA 6
Namibia 18.11 102 UM 31 SSF 10
Honduras 17.96 103 LM 20 LCN 17
Kyrgyzstan 17.86 104 LM 21 CSA 7
El Salvador 17.31 105 LM 22 LCN 18
Uganda 17.23 106 LI 6 SSF 11
Mali 16.82 107 LI 7 SSF 12
Bangladesh 16.80 108 LM 23 CSA 8
Belarus 16.72 109 UM 32 EUR 38
Brunei Darussalam 16.51 110 HI 48 SEAO 15
Botswana 16.36 111 UM 33 SSF 13
Malawi 16.15 112 LI 8 SSF 14
Cameroon 16.12 113 LM 24 SSF 15
Nepal 15.90 114 LI 9 CSA 9
Albania 15.69 115 UM 34 EUR 39
Zimbabwe 15.61 116 LI 10 SSF 16
Algeria 15.56 117 UM 35 NAWA 18
Zambia 15.52 118 LM 25 SSF 17
Nigeria 14.90 119 LM 26 SSF 18

Benin 14.78 120 LI 11 SSF 19

Rwanda 13.66 121 LI 12 SSF 20
Burundi 12.40 122 LI 13 SSF 21
Niger 11.18 123 LI 14 SSF 22
Guinea 9.97 124 LI 15 SSF 23
Yemen 8.90 125 LM 27 NAWA 19
Burkina Faso 8.45 126 LI 16 SSF 24
Togo 8.02 127 LI 17 SSF 25
Note: World Bank Income Group Classification (July 2016): LI = low income; LM = lower-middle income; UM = upper-middle income; and HI = high income. Regions are based on the United Nations Classification: EUR = Europe;
NAC = Northern America; LCN = Latin America and the Caribbean; CSA = Central and Southern Asia; SEAO = South East Asia, East Asia, and Oceania; NAWA = Northern Africa and Western Asia; SSF = Sub-Saharan Africa.
1: The Global Innovation Index 2016

Box 3: Innovation quality: The USA, Japan, the UK, China, and India at the top of their income groups

Measuring the quality of innovation-related average score of top 3 universities); (2) inter- receive abroad (indicator 6.1.5, citable docu-
input and output indicators is as essential as nationalization of local inventions (indicator ments H index). Figure 3.1 shows how the
tracking their magnitude. To this end, three 5.2.5, patent families filed in three offices, scores on these three indicators add up,
additional indicators were introduced into changed to patent families filed in two and captures the top 10 highest performing
the GII in 2013: (1) quality of local universi- offices in the GII 2016); and (3) the number high- and middle-income economies.
ties (indicator 2.3.4, QS university ranking of citations that local research documents

Figure 3.1: Metrics for quality of innovation: Top 10 high- and top 10 middle-income economies

1 United States of America

2 Japan
High-income economies

3 Switzerland
4 Germany
5 United Kingdom
6 Sweden
7 Korea, Rep.
8 Netherlands
9 Denmark
10 Finland
Average (48 economies)

16 China
27 India
Middle-income economies

28 Russian Federation
29 Brazil
30 Argentina
32 South Africa
34 Mexico n 2.3.4 QS university ranking average score of top 3 universities
38 Malaysia
n 5.2.5 Patent families filed in at least 2 offices
41 Turkey
n 6.1.5 Citable documents H index
43 Thailand
Average (62 economies)

0 50 100 150 200 250 300

Sum of scores

Source: GII 2017 data.

Notes: Numbers to the left of the economy name are the innovation quality rank. Economies are classified by income according to the World Bank Income Group Classification (July 2016). Upper- and lower-middle income categories
are grouped together as middle-income economies.

(Continued on next page)

origin, and printing and publishing points (see page 26 in the GII 2016), and technology outputs (2nd). The
manufactures. which are now better accounted for.37 Netherlands has improved its rank-

The Netherlands reaches the As a result, this year the Netherlands ings in a number of other areas as
3rd position this year, ranking 2nd ranks 6th in FDI net inf lows and well, including Education (18th),
in the Innovation Output Sub-Index 1st in outf lows. As discussed in Innovation linkages (7th), and
and 4th in the Innovation Efficiency more detail in Box4, newly avail- Knowledge impact (17th), in part
Ratio. Indeed, the Netherlands had able data positively affect two pil- because of gains in GERD financed
lost five positions last year as a result lars of the NetherlandsBusiness by abroad and expenditure in edu-
of large f luctuations in selected data sophistication (1st) and Knowledge cation. Areas of weakness include

1: The Global Innovation Index 2017

Box 3: Innovation quality: The USA, Japan, the UK, China, and India at the top of their income groups (continued)

Top 10 high-income economies: The Korea keeps the top spot in patent families, India is 2nd in innovation quality for
USA, Japan, Switzerland, and Germany a reduction in its scores for university rank- the second consecutive year. Indias positive
in the lead ings, combined with a significant improve- performance is the result of maintaining its
Among the high-income group, five econo- ment in patent families for Sweden, can 2nd position in both university rankings and
miesthe United States of America (USA), explain this switch. The Netherlands (8th, citable documents among middle-income
Japan, Switzerland, Germany, and the United up by two) scores better in patent families, economies. The country shows a small
Kingdom (UK)have remained among the compensating for a fall in university rank- reduction in the score of patent families,
top five in innovation quality since the incep- ings. Denmark and Finland enter the top which, however, does not affect its quality
tion of this metric. This year the USA moves 10 this year, replacing France and Canada. of innovation ranking.
to the 1st position, taking the place of Japan. While the latter two show high scores in With slight reductions in all three indi-
The USA achieves this ranking as a result of both university rankings and citable papers, cators, the Russian Federation moves to
continuous top scores in particular quality improved scores for patents filed from both the 3rd position among the upper-middle-
indicators and an improvement in its score Denmark and Finland is the main reason for income economies and 28th overall, posi-
in patent families. The USA takes the top this change. tioned between India and Brazil. Brazils
position in citable papers, sharing this spot performance also shows slight reductions
with the UK for the fifth consecutive year. In Top 10 middle-income economies: in scores for all three indicators, resulting
2017 the USA also remains the world leader China and India lead; the Russian in a ranking of 29th among middle-income
in the quality of its universities, outranking Federation and Argentina re-join the economies.
the UK for the second consecutive year. Also group Argentina, 5th among middle-income
contributing to the USAs improvement, A large gap remains between high-income economies and 30th overall, shows reduced
Japan shows a reduction in the scores for and middle-income economies. Without scores in university rankings and patent
both university rankings and citable docu- China, the difference in average scores families and a marginal improvement in
ments this year. between these two groups in both the citable documents, yet its overall score puts
This year, for the first time, Switzerland university rankings (1.13) and citable docu- it ahead of South Africa (6th among middle-
ranks 3rd in the quality of innovation met- ments (0.64) is expanding, while in patents income and 32nd overall) and Mexico (7th
ric. Although showing a slightly weaker filed the distance is narrowing (0.14). and 34th).
performance than last year in the quality of China moves up one spot to 16th posi- The inclusion of the Russian Federation
universities and a constant one in citable tion in innovation quality, retaining for the and Argentina in the middle-income group
documents, the country enjoys a top score fifth consecutive year its position as the top led to the downward movement of Mexico,
in patent families, helping it to achieve an middle-income economy and getting closer Malaysia, Turkey, and Thailandeconomies
overall quality score above those of both to high-income economies. This movement that have been in the middle-income top
the UK and Germany. These two countries, can be attributed to higher scores in univer- 10 since the innovation quality metric was
on the other hand, show stable scores in sity rankings (4th) and citable documents introduced. In addition, this inclusion also
citable documents this year, but a reduction (14th). Although other middle-income moved Colombia and Ukraine out of this
in those for patent families and university economies still depend greatly on their uni- list, although the performance of these
rankings, respectively. versity rankings to move ahead in the quality economies has diverged greatly from that
Sweden improves its rankings, moving of innovation, Chinaand to some extent of previous years.
up two positions to replace the Republic of South Africadisplay a balance between
Korea (Korea) at the 6th position. Although the three components of the quality index.

Tertiary education (49th), General other pillars. It improves its position Tertiary education (54th), Ecological
infrastructure (30th), Ecological sus- in Human capital and research (13th), sustainability (61st), and Intangible

tainability (39th), Credit (35th), and Business sophistication (8th), and assets (38th). In the latter, the country
Investment (26th). Creative outputs (10th), while losing improves by seven positions this year,
The United States of America eight positions in Infrastructure (21st) a welcome improvement as this is the
(USA) maintains its 4th position this and three in Knowledge and technol- only output sub-pillar where the USA
year. The USA keeps its top ranking ogy outputs (7th). At the sub-pillar does not rank in the top 25. The USA
in pillar 4Market sophistication level, the USA ranks in the top 25 with holds the top rank in many indicators,
and ranks among the top 25 in all just four exceptions: Education (41st), including QS university ranking,
1: The Global Innovation Index 2017

Box 4: The global innovation divide

The top 25 GII ranks are occupied by a ranking. In addition, top marks for intellectual outputs. Conversely, this distance is larger
stable set of high-income countries that property payments, ICT services imports, and this year in both Market and Business sophis-
consistently lead in innovation. One major country-code TLDs help explain this rise. tication. This change is in addition to China
change took place last year: China, as the Some changes occur this year in the scoring higher in Business sophistication and
only middle-income economy included in composition of the top 25 group. For one, Knowledge and technology outputs than its
this group of innovation leaders, took up the Belgium drops out of the top 25 this year peers in the 1125 group.
25th position in 2016. China remains in this while the Czech Republic moves back by
top group and keeps moving ahead (22nd relying on a better performance in high- and Middle-income economies: China,
this year). Chinas innovation ranking in 2017 medium-high-tech manufactures, as well the only middle-income economy
reflects scores in Business sophistication and as improved scores for domestic credit to among the top 25 group; Bulgaria and
Knowledge and technology outputs that are private sector and FDI net outflows. Malaysia still at great distance
above the average of the rest of the 1125 The distance between the top 25 and Aside from China, which has been among
group. In particular, top scores in some the groups that follow is still apparent. Figure the top 25 since 2016, this year Bulgaria
indicatorsdomestic market scale, firms 4.1 shows the average scores for six groups: and Malaysia are the two middle-income
offering formal training, patents by origin, (1) the top 10, composed of all high-income economies nearest to that group, with
utility models by origin, high-tech exports economies; (2) ranks 11 through 25, which Malaysia slipping back to 37th and Bulgaria
less re-exports, industrial designs by origin, are also all high-income economies with overtaking it. Bulgaria (36th) is now the clos-
and creative goods exportsare all factors the sole exception of upper-middle-income est upper-middle-income economy to the
contributing to this improved ranking. Over China; (3) other high-income economies; top 25. In particular, Bulgaria performs better
the past two years, in both absolute and rela- (4) upper-middle-income economies; (5) this year in Information and communication
tive terms in relation to other countries, China lower-middle-income economies; and (6) technologies (ICTs) with an improved perfor-
has shown the strongest improvement in low-income economies. mance in governments online service and
patent applications by origin, university rank- e-participation as well as in variables in other
ings, citable documents H index, utility model The difference between the top 10 pillars, including research talent in business
applications by origin, gross expenditure on innovation leaders and others in the enterprise and growth rate of PPP$ GDP per
R&D, and PCT international applications by top 25 worker. Malaysia, on the other hand, main-
origin. In addition, China this year displays a Overall, the top 10 perform better than the tains strengths in graduates in science and
strong performance in three indicators intro- 1125 group in all pillars. The gap between engineering, high-tech imports and exports,
duced in the GII 2016: global R&D companies, these two groups is larger this year in both and creative goods exports, among other
domestic market scale, and research talent in of the output-side pillars of the index. This indicators. Both of these economies con-
business enterprise. contrast shows also that variations in perfor- tinue to operate close to those high-income
Stability is a feature among the top 10 mance are narrower in two of the input-side economies outside of the top 10, which is
economies this year, with Switzerland at pillars, Institutions and Market sophistication. especially evident in Business sophistication,
number 1 for the seventh consecutive year. In contrast, these gaps have expanded in Knowledge and technology outputs, and
Although some variations in rankings are Human capital and research, Infrastructure, Creative outputs.
noticed, such as the Netherlands regaining and Business sophistication. With the exception of these two
3rd place (thanks in part to methodological A number of high-income economies countries, the gap between the group of
reasons explained in the country description in the 1125 rangeHong Kong (China) 1125 ranked economies (as well as high-
on page 20), no economy moves in or (16th), Canada (18th), Norway (19th), and income economies) and the upper-middle-
out of this group in 2017. The Netherlands New Zealand (21st)perform above the top income group remains wide, especially in
noteworthy upward movement relies mostly 10 average in various pillars (i.e., Institutions, Institutions, Human capital and technology,
on its consistently high performance in areas Infrastructure, and Market sophistication). and Infrastructure; the gap is less wide in
such as Business sophistication, Creative out- This year, for the first time, China displays Creative outputs. With respect to last year,
puts, and Knowledge and technology out- a score higher than the top 10 average partially influenced by methodological con-
puts. Within Knowledge diffusion, available in Knowledge and technology outputs. siderations, the divide between these groups
data for intellectual property receipts and ICT Furthermore, China shows that the gaps are increases in Institutions and, to a lesser extent,
services exports rank the Netherlands in the narrower between the top 10 average scores in Market sophistication. Yet the gap seems
top 10. FDI net outflows is also a strength and and its scores in Institutions, Human capital to be lessening in Infrastructure and Human
partly responsible for this improvement in and research, Infrastructure, and Creative capital and research.

(Continued on next page)


1: The Global Innovation Index 2017

Box 4: The global innovation divide (continued)

Only a few upper-middle-income econ- sophistication pillars, areas in which this 19 economies) and Latin America and the
omiesRomania (42nd), Turkey (43rd), the group also continues to perform above the Caribbean (31.7; 18 economies) have similar
Russian Federation (45th), Viet Nam (47th), average of the lower-middle-income group. scores while the difference in average scores
Montenegro (48th), and Ukraine (50th)are This suggests that efforts to strengthen insti- between Central and Southern Asia (28.5; 9
among the top 50 this year. tutions and enable the necessary factors to economies) and Sub-Saharan Africa (24.8; 25
promote stronger business environments economies) is expanding. When contrasted
Low-income economies moving closer continue to expand among these countries. with the 2016 results, these averages show
to middle-income ones Latin America and the Caribbean to be the
Continuing with the trend identified in earlier The persistence of regional innovation region with the widest average improve-
editions, the group of low-income economies divides: Regional scores ment, followed by Central and Southern
keeps closing the gap that separates them The regional rankings based on the GII scores Asia, Northern Africa and Western Asia, and
from the middle-income group. However, shows that the Northern America region Europe. Conversely, Sub-Saharan Africa
this gap remains significant in Infrastructure, consisting of the USA and Canadais still shows the largest average score reduction,
Market sophistication, Creative outputs, and at the top (57.5; 2 economies), followed by followed by South East Asia, East Asia, and
Knowledge and technology outputs. This Europe (47.1; 39 economies) and South East Oceania and Northern America.
year there is no difference between these Asia, East Asia, and Oceania (44.0; 15 econo-
groups in the Institutions and Business mies). Northern Africa and Western Asia (34.3;

Figure 4.1: Innovation divide: China rising among the top 25

Average scores
Top 10 (high income)
Creative Human capital 1125 (high income plus China)
outputs and research Other high income

Upper-middle income
Lower-middle income
25 Low income

Knowledge and Infrastructure

technology outputs

Business sophistication Market sophistication

Source: GII 2017 data.

Note: Countries/economies are classified according to the World Bank Income Group Classification (July 2016).

venture capital deals, citable docu- 2016). The country improves in all Investment. It improves substantially
1: The Global Innovation Index 2017

ments, computer software spending, pillars except for Market sophistica- also in Education and Creative goods
and IP receipts; it also gains the 1st tion, where it retains the 6th spot, and and services, moving up by nine
position in global R&D companies, Knowledge and technology outputs positions in both sub-pillars. Despite
state of cluster development (see also (16th), where it loses two positions. these improvements, Singapore
the Special Section on Clusters, which At the sub-pillar level, Demark shows a relatively weak position in
shows that the USA has largest num- improves the most in Education (4th), Education, where it ranks 76th. In
ber of clusters in the world), ICTs and ICTs (14th), Ecological sustainability this sub-pillar, Singapore is weak
organizational model creation, and (11th), Innovation linkages (17th), in all indicators except PISA results.
cultural and creative services exports. Knowledge diffusion (17th), and Room for improvement also exists in
This year the country also ranks 1st in Intangible assets (25th). Denmark growth rate of GDP per worker, ICT
the quality of innovation aggregate, ranks in the top 3 in a number of services exports, and trademarks and
overtaking Japan (see Box3). indicators, including expenditures on industrial designs by origin. Apart
The United Kingdom (UK) education, researchers, ICT use, and from these areas of opportunity,
moves to 5th place this year. The scientific and technical articles. It also Singapore maintains its 1st place in
UK improves its position in a number improves its position in many areas FDI net outf lows, while losing it in
of input pillars, namely Institutions such as government expenditure per high- and medium-high-tech manu-
(9th), Human capital and research pupil, PISA scales, GDP per unit factures, high-tech exports, market
(6th), and Business sophistication of energy use, university/industry capitalization, and FDI net inf lows.
(13th).38 At the sub-pillar levels, the research collaboration, JV-strategic Singapore ranks 1st also in other eight
UKs largest gains are in Political alliance deals, ICT services exports, indicators: government effectiveness,
environment (18th), Education and ICTs and organizational model regulatory quality, cost of redun-
(22nd), and Knowledge absorption creation. Opportunities for further dancy dismissal, PISA scales, tertiary
(28th). The country loses ground improvement still exist, notably in inbound mobility, ease of protecting
in both output pillarsKnowledge Tertiary education (19th), General minority investors, applied tariff rate,
and technology outputs (13th) goes infrastructure (44th), Trade, compe- and IP payments.
down by four, with the largest drop tition, and market scale (37th), and Finland moves down to the
in Knowledge diffusion (38th); and Knowledge impact (34th). Relatively 8th position this year from 5th in
Creative outputs (4th) by one. At weak indicators include graduates in 2016. Finland keeps its 4th place in
the indicator level, expenditure on science and engineering, gross capital the Input Sub-Index, but loses three
education, government expenditure formation, utility models by origin, positions in the Output Sub-Index
by pupil, IP payments, ICT services growth rate of GDP per worker, and (13th). It maintains its 1st rank in
imports and exports, growth rate of trademarks by origin. Human capital and research, while
GDP per worker, and national feature Singapore still holds the top improving in Infrastructure (8th). In
films see some of the largest improve- rank in the South East Asia, East Asia, all other pillars, however, Finland
ments. By contrast, items such as PISA and Oceania region while dropping loses between one and four posi-
results, ICT use, and patent families by one position (see Box6). It keeps tions. At the sub-pillar level, 12 out
lose most positions (see also Box3). its top spot in the Innovation Input of 21 sub-pillars move down. The
The UK maintains its 1st spot in Sub-Index and gains three positions largest drops are in Creative goods
citable documents, and gains the 1st in the Innovation Output Sub-Index and services (40th), Political environ-
rank in governments online services, (17th). Singapore ranks in the top 5 in ment (8th), and Knowledge diffusion
e-participation, and ICT and business all input pillars and 1st in Institutions. (14th). The largest gains are in ICTs
model creation. In terms of innovation outputs, (9th) and Knowledge impact (32nd).
Denmark ranks 6th in this Singapore loses one position in Finland also loses positions in a num-

years GII, improving in both the Knowledge and technology outputs ber of indicators, including venture
Innovation Input and Output Sub- (11th) while gaining one in Creative capital deals, GERD performed by
Indices, where it ranks respectively outputs (32nd). At the sub-pillar business, IP receipts and payments,
6th and 12th. Denmark has the most level, Singapore holds its 1st spot in ICTs and business model creation,
notable forward shift in the top 10 Political environment, Regulatory ICTs and organizational model cre-
(progressing continuously, from 10th environment, and Tertiary educa- ation, cultural and creative services
overall in the GII 2015 and 8th in tion, and gains the top rank in exports, and national feature films.

Indeed, as this list shows, Finlands important sub-pillars: Knowledge environment (2nd), Business envi-

1: The Global Innovation Index 2017

downward movement this year is the impact (2nd) and Knowledge dif- ronment (2nd), Ecological sus-
result of a drop in a variety of indica- fusion (1st). Opportunities lie in tainability (1st), Credit (3rd), and
tors. Apart from Human capital and General infrastructure (34th), Credit Knowledge absorption (3rd). Hong
research and the sub-pillar Business (40th), Investment (29th), Knowledge Kong (China), however, drops
environment, Finland ranks 1st in creation (38th), and Creative goods significantly in Education (73rd),
several indicators: rule of law, ease of and services (33rd). Ireland shows which is a weak sub-pillar this year,
resolving insolvency, environmental weakness in a number of particular and R&D (33rd). This is partly the
performance, and patent families. indicators, including domestic credit result of a new missing value (school
Germany continues its climb up to private sector, market capitaliza- life expectancy) and a drop in global
the GII rankings, gaining a position tion, intensity of local competition, R&D companies (43rd). Other weak
from last year when it entered the top industrial designs by origin, and cul- indicators include GERD financed
10 for the first time. Germany is 1st tural and creative services exports. by abroad, IP payments, and ICT
in logistics performance and patents Ireland holds the top position in IP services imports and exports. Despite
by origin. It is 2nd in global R&D payments, ICT services exports, and these downward movements, Hong
companies expenditures, down from FDI net outf lows, and shows a better Kong (China) preserves its top spot
1st place in 2016, and 3rd in state of ranking than in 2016 in a number of in JV-strategic alliance deals, high-
cluster development and citable docu- important indicators, including PISA tech imports, and FDI net inf lows
mentsthe same as last year. On the results, researchers, global R&D and improves its rank in PISA results,
pillar level, Germany safeguards all its companies, gross capital formation, patents by origin, and utility models
respectable positions while improv- and GDP per unit of energy use. by origin.
ing in Infrastructure (20th). It ranks Canada remains in the 10th
in the top 25 economies across all pil- The top 10 in the Innovation Input Sub- position in the Innovation Input
lars, and in the top 10 economies for Index Sub-Index, while ranking 18th
output pillars. Areas of opportunity The Innovation Input Sub-Index overall, down three positions from
include Education (29th), Ecological considers the elements of an economy 2016. Canadas strengths on the input
sustainability (36th), Credit (28th), that enable innovative activity across side are a result of having top 25
Investment (41st), and Creative goods five pillars. The top 10 economies rankings in six of the seven pillars.
and services (28th). At the indicator in the Innovation Input Sub-Index Canada shows particular strengths in
level, Germany improves in govern- are Singapore, Sweden, Switzerland, Institutions (7th) and Market sophis-
ment expenditure by pupil (up by Finland, the USA, Denmark, the UK, tication (3rd), while improving in
5 spots), tertiary enrolment (up by Hong Kong (China), the Netherlands, Human capital and research (20th).
11), governments online service (up and Canada. Hong Kong (China) and This year, however, Canada loses
by 13), market capitalization (up by Canada are the only economies in seven positions in Infrastructure (18th)
6), FDI net inf lows (up by 19), and this group that are not also in the GII and four in Business sophistication
ICTs and business model creation top 10. The Netherlands entered the (24th). In Infrastructure, it loses posi-
(up by 6). Germany has opportunity top 10 in 2017, while Japan, ranked tions in all sub-pillarsin particular
for improvement in ease of starting 9th on the input side last year, exited in Ecological sustainability, where
a business, gross capital formation, the top 10 this year. it loses 19 positions in ISO 14001
females employed with advanced Hong Kong (China) drops environmental certificates (73rd). In
degrees, IP payments, growth rate of from 2nd to 8th in the Innovation Business sophistication, Canada drops
GDP per worker, and new businesses. Input Sub-Index this year and ranks most in innovation linkages, driven
Ireland is ranked 10th this year, 16th overall, down from 14th in by a decline in ranking in university/
down three positions from last year. 2016. It retains its good positions in industry research collaboration. Top

Ireland ranks in the top 25 across all Institutions (3rd) and Market sophis- 10 sub-pillar rankings for Canada this
pillars, but loses positions in Market tication (2nd), but falls in three out year are Political environment (6th,
sophistication (25th), Business sophis- of five input pillars, with the largest a strength), Regulatory environ-
tication (10th), Knowledge and tech- drop in Human capital and research ment (10th), Business environment
nology outputs (5th), and Creative (28th). In 9 of the 15 input sub-pillars, (7th), General infrastructure (7th),
outputs (13th). At the sub-pillar level, Hong Kong (China) ranks in the top Credit (8th), and Investment (2nd,
Ireland places in the top 2 in two 10, holding top spots in Regulatory also a strength). Canada improves in

Education in 2017, in part because of industrial designs by origin and ICT advances its ranking in Knowledge
1: The Global Innovation Index 2017

stronger rankings in expenditure on and organizational model creation. creation (13th) and Knowledge dif-
education, government expenditure Luxembourg also keeps the top posi- fusion (21st), ranking 1st in scientific
by pupil, and PISA results. tion in the Innovation Efficiency and technical articles and improving
Ratio rankings. in PCT patent applications, growth
The top 10 in the Innovation Output Sub- The Republic of Korea rate of GDP per worker, ISO 9001
Index (Korea) attains the 9th position in quality certificates, IP receipts,
The Innovation Output Sub-Index the Innovation Output Sub-Index ICT services exports, and FDI net
variables provide information on ele- this year, up by two positions. Korea outf lows.
ments that are the result of innovation gains six positions in Creative
within an economy. Although scores outputs, ranking 15th this year. It Top performers by income group
on the Input and Output Sub-Indices improves in Creative goods and ser- Viewing economies among their
might differ substantially, leading to vices (35th) and maintains the top income-group peers can illustrate
important shifts in rankings from one spot in industrial designs by origin. important relative competitive
sub-index to the other for particular Although the country drops one spot advantages and help decision makers
countries, the data confirm that efforts in Knowledge and technology outputs glean important lessons for improved
made to improve enabling environ- (6th), it improves in one of its areas of performance that are applicable on
ments are rewarded with better greatest strengthKnowledge cre- the ground. The GII also assesses
innovation outputs. The top 10 econ- ation (2nd)where it maintains its results relative to the development
omies in the Innovation Output Sub- top rankings in patents by origin and stages of countries.
Index this year are Switzerland, the PCT patent applications and advances Table4 shows the 10 best-ranked
Netherlands, Sweden, Luxembourg, to the top spot in utility models by economies in each index by income
the USA, the UK, Germany, Ireland, origin. Korea also improves its rank group. Switzerland, Sweden, and the
Korea, and Iceland. in Human capital and research (2nd), Netherlands are among the high-
The 10 economies leading the where it holds its 1st place in R&D. income top 10 on the three main indi-
Innovation Output Sub-Index Although its gross R&D expendi- ces, and the top 3 in the Innovation
remain broadly consistent with ture goes down by one position, Output Sub-Index. Compared to last
their rankings in 2016, with several Korea manages to retain its 2nd and year, Hungary and Estonia leave the
shifts and one substitution: three 3rd positions in GERD performed group, making space for the Czech
economies move upward within the by business and GERD financed by Republic and Korea.
top 10 (the Netherlands, the USA, business, respectively. The countrys Among the 10 highest-ranked
and Germany), while five econo- areas of relative weakness include upper-middle-income economies,
mies move downward (Sweden, ICT services exports and printing nine remain from 2016 (see also
Luxembourg, the UK, Ireland, and and publishing manufactures on the Box 4): China (22nd this year),
Iceland). Korea enters the top 10 on side of outputs; and tertiary inbound Bulgaria (36th), Malaysia (37th),
the Output side, while Finland exits mobility, GDP per unit of energy use, Romania (42nd), Turkey (43rd),
the top 10 in 2017. Seven of these knowledge-intensive employment, Montenegro (48th), Thailand (51st),
economies are ranked in the GII top and FDI inf lows on the inputs side. Costa Rica (53rd), and South Africa
10; the profiles of the other three Iceland ranks 10th in the (57th). The newcomer to this group
economies are discussed below. Innovation Output-Sub Index in of the 10 best upper-middle-income
Luxembourg ranks 4th in the 2017. This year, Iceland gains four performers is the Russian Federation
Innovation Output Sub-Index in 2017 positions in Knowledge and technol- (45th), which displaces Mauritius
and 12th in the overall GII. On the ogy outputs (18th) and reaches 2nd (64th). China, Malaysia, Bulgaria,
output side, Luxembourg loses four place in Creative outputs. Iceland and Romania are among the 10

positions in Knowledge and technol- maintains the top spot in Creative best-ranked upper-middle-income
ogy outputs (15th), while gaining 1st goods and services and Online cre- economies across all three main indi-
place in Creative outputs. In this pillar, ativity, ranking 1st in three of the ces and in the Innovation Efficiency
it maintains its strengths in cultural indicators across these sub-pillars: Ratio.
and creative services exports, national national feature films, printing and The same analysis for lower-
feature films, and generic top-level publishing manufactures, and generic middle-income countries shows
domains (TLDs) and improves in top-level domains (TLDs). Iceland that eight of the top 10 countries

Table 4: Ten best-ranked economies by income group (rank)

1: The Global Innovation Index 2017

Global Innovation Index Innovation Input Sub-Index Innovation Output Sub-Index Innovation Efficiency Ratio

High-income economies (48 in total)

1 Switzerland (1) Singapore (1) Switzerland (1) Luxembourg (1)
2 Sweden (2) Sweden (2) Netherlands (2) Switzerland (2)
3 Netherlands (3) Switzerland (3) Sweden (3) Netherlands (4)
4 United States of America (4) Finland (4) Luxembourg (4) Iceland (5)
5 United Kingdom (5) United States of America (5) United States of America (5) Ireland (6)
6 Denmark (6) Denmark (6) United Kingdom (6) Germany (7)
7 Singapore (7) United Kingdom (7) Germany (7) Malta (8)
8 Finland (8) Hong Kong (China) (8) Ireland (8) Sweden (12)
9 Germany (9) Netherlands (9) Korea, Rep. (9) Czech Republic (13)
10 Ireland (10) Canada (10) Iceland (10) Korea, Rep. (14)

Upper-middle-income economies (35 in total)

1 China (22) China (31) China (11) China (3)
2 Bulgaria (36) Malaysia (36) Bulgaria (32) Turkey (9)
3 Malaysia (37) Russian Federation (43) Turkey (36) Bulgaria (15)
4 Romania (42) Bulgaria (45) Malaysia (39) Iran, Islamic Rep. (16)
5 Turkey (43) Mauritius (47) Thailand (43) Thailand (24)
6 Russian Federation (45) South Africa (49) Romania (44) Panama (38)
7 Montenegro (48) Montenegro (50) Costa Rica (50) Romania (39)
8 Thailand (51) Romania (51) Russian Federation (51) Costa Rica (43)
9 Costa Rica (53) Colombia (52) Montenegro (52) Malaysia (46)
10 South Africa (57) TFYR of Macedonia (53) Panama (55) Dominican Republic (54)

Lower-middle-income economies (27 in total)

1 Viet Nam (47) India (66) Viet Nam (38) Viet Nam (10)
2 Ukraine (50) Mongolia (67) Ukraine (40) Ukraine (11)
3 Mongolia (52) Viet Nam (71) Moldova, Rep. (42) Armenia (17)
4 Moldova, Rep. (54) Moldova, Rep. (73) Armenia (47) Moldova, Rep. (22)
5 Armenia (59) Ukraine (77) Mongolia (48) Mongolia (27)
6 India (60) Morocco (79) India (58) Cte d'Ivoire (40)
7 Morocco (72) Tunisia (81) Philippines (65) Indonesia (42)
8 Philippines (73) Armenia (82) Morocco (68) Kenya (50)
9 Tunisia (74) Philippines (83) Kenya (70) India (53)
10 Kenya (80) Kyrgyzstan (86) Tunisia (71) Philippines (55)

Low-income economies (17 in total)

1 Tanzania, United Rep. (96) Rwanda (76) Tanzania, United Rep. (76) Tanzania, United Rep. (29)
2 Rwanda (99) Uganda (93) Ethiopia (91) Ethiopia (32)
3 Senegal (100) Burkina Faso (101) Madagascar (95) Madagascar (45)
4 Uganda (102) Senegal (102) Senegal (98) Mozambique (70)
5 Mozambique (107) Nepal (108) Mozambique (100) Mali (78)
6 Nepal (109) Tanzania, United Rep. (109) Uganda (106) Zimbabwe (89)

7 Ethiopia (110) Benin (110) Mali (107) Senegal (95)

8 Madagascar (111) Niger (111) Malawi (112) Malawi (98)
9 Malawi (115) Malawi (112) Nepal (114) Nepal (105)
10 Benin (116) Mozambique (114) Zimbabwe (116) Benin (110)

Note: Economies with top 10 positions in the GII, the Input Sub-Index, the Output Sub-Index and the Innovation Efficiency Ratio within their income group are highlighted in bold.

from 2016 remain in the top 10 Sub-Index score. It assesses the effec- Clustering innovation leaders, innovation
1: The Global Innovation Index 2017

this year. These include Viet Nam tiveness of innovation systems and achievers, and innovation performers at and
(47th), Ukraine (50th), the Republic policies. It must be noted, however, below development relative to GDP: The GII
of Moldova (54th), Armenia (59th), that economies might also reach a bubble chart
India (60th), Morocco (72nd), relatively high Innovation Efficiency The GII helps also identify countries
the Philippines (73rd), and Kenya Ratio as a result of particularly low performance in innovation relative to
(80th). New this year to the top 10 input scores. Because of this, effi- their level of GDP. Figure4 on pages
lower-middle-income countries are ciency ratios must be analysed jointly 3031 presents the GII scores
Mongolia (52nd) and Tunisia (74th), with GII, Input, and Output scores, plotted against GDP per capita in
which displace Georgia (68th) and and with the development stages of PPP$ (in natural logs). The econo-
Tajikistan (94th). Seven of the top 10 the economies in mind. mies that appear close to the trend
lower-middle-income countries have The 10 countries with the high- line show results that are in accor-
rankings in the top 10 for each of est Innovation Efficiency Ratios are dance with what is expected based
the three indices and the Innovation countries that combine certain lev- on their level of development. The
Efficiency Ratio, with the exceptions els of innovation inputs with more further up and above the trend line a
of Morocco, Tunisia, and Kenya. robust output results (see Table1): country appears, the better its inno-
There has also been a strong Luxembourg, Switzerland, China, vation performance is when com-
consistency among low-income the Netherlands, Iceland, Ireland, pared with that of its peers at the same
countries, with eight out of 10 Germany, Malta, Turkey, and Viet stage of development. Red-coloured
economies remaining in the top 10. Nam. Compared to previous years, bubbles in the figure correspond to
The United Republic of Tanzania is new middle-income economies the efficient innovators (a major-
the top-ranked low-income country joined the top 10 most efficient ity of them are situated above the
(96th), having moved up nine spots economies: China, which entered the trend line), while the blue-coloured
in the overall GII since 2016, and top 10 last year, is accompanied this bubbles represent those countries
with improvements in the Innovation year by Turkey and a lower-middle- in the lower half of the Innovation
Input (109th) and Output (76th) income economy, Viet Nam, which Efficiency Ratio.
Sub-Indices (see Box5). Following makes the most spectacular progress In the group of innovation leaders
in the ranking of low-income coun- this year (see Box6). we find the same top 25 economies
tries are Rwanda (99th), Senegal Economies from Europe; South as in 2016, with two exceptions: the
(100th), which displaces the now- East Asia, East Asia, and Oceania; Czech Republic is moving back into
lower-middle-income economy and Northern Africa and West Asia this group while Belgium is moving
Cambodia (101st), Uganda (102nd), take up the first 20 positions in this out. All of these are high-income
Mozambique (107th), Nepal (109th), ratio ranking. Among high-income economies, with the sole exception of
Ethiopia (110th), Madagascar (111th), economies, Sweden, the Czech China, which belongs to the upper-
Malawi (115th), and Benin (116th), Republic, Korea, Kuwait, Estonia, middle-income group. These econo-
which displaces Mali (118th). and the UK are in the group of mies are located in four regions, with
Ranking well across all main indices the 20 most efficient economies in the majority in South East Asia, East
of the GII, the United Republic of innovation. Among upper-middle- Asia, and Oceania and in Europe, and
Tanzania, Senegal, Mozambique, income economies, Bulgaria and the the rest in Northern America and in
Nepal, and Malawi are among the Islamic Republic of Iran are in the Northern Africa and Western Asia.
top 10 low-income countries. All top 20 in terms of efficiency. From All of the economies in this group
economies in the low-income top the lower-middle-income group, have a GII score above 50. These
10, except Rwanda and Uganda, the top 20 most efficient economies economies show mature innovation
are in the low-income top 10 in the include Ukraine and Armenia. No systems with solid institutions and

Innovation Efficiency Ratio. low-income economies are in the top high levels of market and business
20 this year in innovation efficiency sophistication, allowing investment
Maximizing innovation resources and rankings. in human capital and infrastructure
synergies: The Innovation Efficiency Ratio to translate into quality innovation
The Innovation Efficiency Ratio is outputs.
calculated as the ratio of the Output Economies that perform at least
Sub-Index score over the Input 10% above their peers for their level

of GDP are called innovation achiev- Table 5: Innovation achievers: Income group and years as an innovation achiever

1: The Global Innovation Index 2017

ers. These are shown in Table5 listed
Economy Income group Years as an innovation achiever (total)
by income group and years as an
Viet Nam Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7)
innovation achiever. These econo-
Kenya Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7)
mies show better results in inno-
Moldova, Rep. Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7)
vation because they continuously
India Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7)
improve their innovation systems,
Armenia Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012 (6)
have more structured institutional
Ukraine Lower-middle income 2017, 2016, 2015, 2014, 2012 (5)
frameworks, develop linkages that
Rwanda Low income 2017, 2016, 2015, 2014, 2012 (5)
allow knowledge absorption and the
Uganda Low income 2017, 2016, 2015, 2014, 2013 (5)
f low of highly skilled human capital,
Mozambique Low income 2017, 2016, 2015, 2014, 2012 (5)
and foster a higher integration with
international markets. Although these Malawi Low income 2017, 2016, 2015, 2014, 2012 (5)

traits translate into proper resource Senegal Low income 2017, 2015, 2014, 2013, 2012 (5)

allocation for education, higher levels Tajikistan Lower-middle income 2017, 2016, 2013 (3)

of economic growth, and income for Malta High income 2017, 2016, 2015 (3)

workers, they are not homogenous Madagascar Low income 2017, 2016 (2)

among these economies. Bulgaria Upper-middle income 2017, 2015 (2)

A total of 17 economies compose Burundi Low income 2017 (1)

the group of innovation achievers. Tanzania, United Rep. Low income 2017 (1)

This group has grown since the 2016 Note: World Bank Income Group Classification (July 2016): LI = low income; LM = lower-middle income; UM = upper-middle income; and HI = high income.

edition of the GII. Most of these

economiesnine in totalcome
from the Sub-Saharan Africa region, United Republic of Tanzania only group, and 2 are low-income
followed by three economies in the in 2017. With the exception of economies.
Eastern region of Europe. A stronger Senegal, Bulgaria, and the latter two
performance in innovation outputs economies, all have been identified as
this year allows the Czech Republic innovation achievers in the two most Regional rankings
to leave the achiever group and move recent years. Kenya, the chief innova- This section discusses regional and
into the group of leader economies. tion achiever in the region, has been sub-regional trends, with snapshots
Portugal moves also out of this group considered as such every year since for some of the economies leading in
and into the group of economies per- 2011. Most of these economies per- the rankings.
forming on par with their develop- form above their peers in Innovation Table 6 on page 32 presents
ment for their level of GDP, partially linkages, particularly in GERD a heatmap with the scores for the
as a result of a weaker performance financed by abroad and FDI net top 10, along with average scores by
in general infrastructure and knowl- inf lows. These economies also share income and regional group. To put
edge absorption. Two new econo- strengths in government expenditure the discussion of rankings further
mies join this group: Burundi and the on education per pupil, gross capital into perspective, Figure5 on page
United Republic of Tanzania from formation, and the growth rate of 33 presents, for each region, bars
Sub-Saharan Africa, while Armenia GDP per worker. representing the median pillar scores
from Northern Africa and Western This analysis also allows for iden- (second quartile) as well as the range
Asia and Bulgaria from the Eastern tifying a group of economies that of scores determined by the first and
Europe region appear in this list for perform at least 10% below their peers second quartile; regions are presented
the second year in a row. for their level of GDP. This cluster in decreasing order of their average

Importantly, Kenya, Rwanda, includes 39 countries from different GII rankings (except for the EU,
Senegal, Uganda, Mozambique, and regions and income groups: 9 are which is placed at the end).
Malawi stand out for being innova- from the high-income group (6 of
tion achievers at least five times in these are from the Northern Africa Northern America (2 economies)
the previous six years. Madagascar and Western Asia region), 17 are from Northern America, the UN-defined
has done so in the two most recent the upper-middle-income group, 11 region that includes both the USA
years and both Burundi and the are from the lower-middle-income and Canada, holds two of the top

Figure 4: GII scores and GDP per capita in PPP$ (bubbles sized by population)
1: The Global Innovation Index 2017



leaders SE


50 BE

GII score

40 HR
Innovation MD
achievers ZA MX SA KW
MW MG PK Innovation underperformers
BF ML CM BD relative to GDP

Efficient innovators
Inefficient innovators

400 1,600 6,400 25,600 102,400

GDP per capita in PPP$ (logarithmic scale)

Note: Efficient innovators are countries/economies with Innovation Efficiency ratios 0.66; Inefficient innovators have ratios < 0.66; the trend line is a polynomial of degree three with intercept (R = 0.6431).

Figure 4: GII scores and GDP per capita in PPP$ (bubbles sized by population): ISO-2 Country Codes

1: The Global Innovation Index 2017

Country/ Economy Code Country/ Economy Code Country/ Economy Code

Albania................................................................. AL Guatemala ........................................................... GT Oman ..................................................................OM

Algeria ................................................................. DZ Guinea..................................................................GN Pakistan ............................................................... PK
Argentina .............................................................AR Honduras .............................................................HN Panama ................................................................ PA
Armenia ..............................................................AM Hong Kong (China) ...............................................HK Paraguay .............................................................. PY
Australia...............................................................AU Hungary ...............................................................HU Peru...................................................................... PE
Austria ..................................................................AT Iceland .................................................................. IS Philippines ...........................................................PH
Azerbaijan ............................................................ AZ India......................................................................IN Poland ...................................................................PL
Bahrain ................................................................BH Indonesia ..............................................................ID Portugal ............................................................... PT
Bangladesh ..........................................................BD Iran, Islamic Rep.................................................... IR Qatar ....................................................................QA
Belarus ................................................................. BY Ireland................................................................... IE Romania...............................................................RO
Belgium ............................................................... BE Israel ......................................................................IL Russian Federation ...............................................RU
Benin ....................................................................BJ Italy....................................................................... IT Rwanda ...............................................................RW
Bolivia, Plurinational St........................................BO Jamaica ................................................................JM Saudi Arabia......................................................... SA
Bosnia and Herzegovina.......................................BA Japan ....................................................................JP Senegal ................................................................SN
Botswana ............................................................BW Jordan ...................................................................JO Serbia................................................................... RS
Brazil ....................................................................BR Kazakhstan........................................................... KZ Singapore............................................................. SG
Brunei Darussalam ...............................................BN Kenya ................................................................... KE Slovakia................................................................ SK
Bulgaria ...............................................................BG Korea, Rep. ........................................................... KR Slovenia ................................................................ SI
Burkina Faso......................................................... BF Kuwait.................................................................KW South Africa ......................................................... ZA
Burundi ................................................................. BI Kyrgyzstan ........................................................... KG Spain .....................................................................ES
Cambodia .............................................................KH Latvia ....................................................................LV Sri Lanka .............................................................. LK
Cameroon............................................................ CM Lebanon ............................................................... LB Sweden .................................................................SE
Canada ................................................................. CA Lithuania...............................................................LT Switzerland ..........................................................CH
Chile ......................................................................CL Luxembourg ......................................................... LU Tajikistan ............................................................... TJ
China....................................................................CN Madagascar.........................................................MG Tanzania, United Rep. ...........................................TZ
Colombia .............................................................. CO Malawi ...............................................................MW Thailand ............................................................... TH
Costa Rica............................................................. CR Malaysia .............................................................. MY TFYR of Macedonia.............................................. MK
Cte dIvoire .......................................................... CI Mali ..................................................................... ML Togo ..................................................................... TG
Croatia..................................................................HR Malta ...................................................................MT Trinidad and Tobago ..............................................TT
Cyprus .................................................................. CY Mauritius.............................................................MU Tunisia.................................................................. TN
Czech Republic ......................................................CZ Mexico................................................................. MX Turkey .................................................................. TR
Denmark ..............................................................DK Moldova, Rep. .....................................................MD Uganda ................................................................UG
Dominican Republic .............................................DO Mongolia.............................................................MN Ukraine ................................................................UA
Ecuador .................................................................EC Montenegro ........................................................ ME United Arab Emirates ........................................... AE
Egypt.................................................................... EG Morocco ..............................................................MA United Kingdom ...................................................GB
El Salvador ........................................................... SV Mozambique ....................................................... MZ United States of America...................................... US
Estonia ..................................................................EE Namibia ...............................................................NA Uruguay ...............................................................UY
Ethiopia.................................................................ET Nepal ...................................................................NP Viet Nam ..............................................................VN
Finland ...................................................................FI Netherlands ......................................................... NL Yemen .................................................................. YE
France .................................................................. FR New Zealand ........................................................ NZ Zambia ................................................................ ZM

Georgia ................................................................ GE Niger .................................................................... NE Zimbabwe .......................................................... ZW

Germany .............................................................. DE Nigeria .................................................................NG
Greece ..................................................................GR Norway ................................................................NO

Table 6: Heatmap for GII top 10 economies and regional and income group averages (1100)
1: The Global Innovation Index 2017

Business sophistication
Market sophistication

technology outputs

Creative outputs
Knowldege and
Human capital

and research


Switzerland 67.69 89.47 63.29 65.10 67.51 62.61 69.60 69.06 62.50 65.78 0.95
Sweden 63.82 88.31 63.71 69.13 64.87 62.58 69.72 62.51 53.33 57.92 0.83
United Kingdom 63.36 88.24 54.70 63.32 59.02 63.69 65.79 62.88 58.97 60.92 0.93
United States of America 61.40 86.25 57.21 61.04 83.45 56.41 68.87 54.38 53.48 53.93 0.78
Finland 60.89 88.44 63.32 67.14 70.19 52.18 68.25 46.49 60.54 53.52 0.78
Singapore 58.70 91.43 66.13 63.19 70.17 52.50 68.68 43.93 53.48 48.71 0.71
Ireland 58.69 94.36 63.67 69.15 71.20 62.88 72.25 47.33 42.94 45.14 0.62
Denmark 58.49 92.18 66.41 64.35 61.59 60.12 68.93 48.79 47.32 48.06 0.70
Netherlands 58.39 83.53 60.13 61.55 60.00 51.44 63.33 51.06 55.85 53.46 0.84
Germany 58.13 87.62 55.07 62.06 55.05 54.51 62.86 55.88 50.94 53.41 0.85

Average 37.12 63.05 34.03 46.19 47.23 34.97 45.10 25.77 32.53 29.15 0.63


Northern America 57.53 88.62 55.26 61.54 78.56 52.13 67.22 46.52 49.14 47.83 0.71
Europe 47.10 75.57 46.41 56.10 51.72 42.93 54.54 35.24 44.05 39.65 0.72
South East Asia, East Asia, and Oceania 44.03 69.62 41.40 52.80 57.37 41.08 52.46 33.73 37.50 35.61 0.68
Northern Africa and Western Asia 34.33 59.33 32.43 46.35 44.87 28.62 42.32 22.80 29.89 26.34 0.61
Latin America and the Caribbean 31.73 54.51 26.84 43.56 45.11 31.11 40.23 17.35 29.13 23.24 0.58
Central and Southern Asia 28.53 47.28 24.25 37.52 43.78 27.29 36.02 20.57 21.51 21.04 0.59
Sub-Saharan Africa 24.88 52.19 18.53 30.45 36.21 27.88 33.05 14.77 18.64 16.71 0.51

Income level

High income 48.85 79.28 48.34 58.64 55.46 44.41 57.23 36.65 44.30 40.47 0.70
Upper-middle income 34.13 59.47 31.50 45.74 45.69 31.05 42.69 21.14 30.00 25.57 0.60
Lower-middle income 28.80 47.61 22.34 35.91 43.48 27.02 35.27 19.75 24.92 22.34 0.62
Low income 23.38 49.11 17.44 28.32 33.13 28.99 31.40 14.17 16.55 15.36 0.49

Worst Average Best

Source: GII 2017 data.

Note: Darker shadings indicate better performances. Countries/economies are classified according to the World Bank Income Group (July 2016; see; and special
classification based on the online version of the United Nations publication Standard Country or Area Codes for Statistical Use, originally published as Series M, No. 49, and now commonly referred to as the M49 standard (April 2017; see

Figure 5: Median scores by regional group and by pillar

1: The Global Innovation Index 2017


Human capital and research


Market sophistication

Business sophistication

n Northern America
n Europe
Knowledge and technology outputs n South East Asia, East Asia, and Oceania
n Northern Africa and Western Asia
n Latin America and the Caribbean
n Central and Southern Asia
n Sub-Saharan Africa
n European Union

Creative outputs

0 20 40 60 80 100


Source: GII 2017 data.

Note: The bars show the median scores (second quartiles); the lines show the range for scores between the first and third quartiles. Countries/economies are classified according to the United Nations geographical classification. The European
Union overlaps (it includes 27 European countries, and Cyprus in Western Asia).

25 economies in this years GII. in the top 10 economies in both the Sub-Saharan Africa (25 economies)
1: The Global Innovation Index 2017

Both the USA and Canada are high- Innovation Input Sub-Index (5th) For several editions, the GII has
income economies and rank in the and the Innovation Output Sub- noted that the Sub-Saharan Africa
top 10 economies in terms of GDP. Index (5th). Canada is 18th overall region performs relatively well
The USA ranks 4th overall this and is in the top 25 economies in the on innovation (see Box 5). Since
year, unchanged from 2016, and is Innovation Input Sub-Index (10th) 2012, Sub-Saharan Africa has had
and the Innovation Output Sub- more countries among the group of
Index (23rd), unchanged from last innovation achievers than any other
year. region. It will be important for Africa

Box 5: Sub-Saharan Africa: The innovation momentum in the most promising region continues

Since 2012 and to this day, the number of Business sophistication, efforts to improve measured by GDP per capita). Countries also
Sub-Saharan Africa countries in the group of Human capital and research as well as have the opportunity to be pillar outperform-
innovation achievers has been the highest Infrastructure have also translated into higher ers if they outperform their peers on more
among all regions. Strengths in the region regional scores in these pillars. Although than half of the seven GII pillars. Countries that
remain in areas considered crucial for the larger economies such as South Africa, meet both of these benchmarks are referred
expansion of innovation locally. Factors such Botswana, Namibia, and Kenya help foster to as innovation outperformers.
as improved business environments offer the the expansion in Infrastructure, others such as Although the number of countries fea-
necessary stimulus to maintain the positive Senegal, Mauritius, Rwanda, and Zimbabwe tured in the GII this year is similar to last
development seen in Sub-Saharan Africa over are helping to do so in Human capital and years, the number of countries identified
the past years. research. as innovation achievers is slightly higher.
Boosted by economies such as Mauritius, This box showcases the regional innova- Figure 5.1 shows the performance of all 25
Botswana, South Africa, Namibia, Rwanda, tion performance of Sub-Saharan Africa coun- economies in Sub-Sahara Africa. This year
and Burkina Faso, this year Sub-Saharan Africa tries by considering both the overall GII scores over 50% of innovation achievers come from
has its highest scores in Institutions and and those of the seven individual GII pillars. Sub-Saharan Africa, allowing this region to
Market sophistication, where these countries Countries are termed innovation achievers continue to lead in this metric. A total of
perform on par or better than some of their and said to outperform their peers if their nine economiesKenya, Rwanda, Uganda,
peers in Europe and South East Asia, East Asia, GII scores are higher than expected based Mozambique, Malawi, Senegal, Madagascar,
and Oceania. In addition to developments in on their level of economic development (as Burundi, and the United Republic of Tanzania

Figure 5.1: Innovation achievers in Sub-Saharan Africa



37 Innovation achiever
ZAF Performing at level of development
GII score

Performing below level of development

UGA SEN TZA Upper bound
MOZ MDG Trend line
22 BDI ETH ZWE Lower bound
17 GIN


750 2,025 5,468

GDP per capita in PPP$ (logarithmic scale)

Note: BDI = Burundi; BEN = Benin; BFA = Burkina Faso; BWA = Botswana; CIV = Cte dIvoire; CMR = Cameroon; ETH = Ethiopia; GIN = Guinea; KEN = Kenya; MDG = Madagascar; MLI = Mali; MOZ = Mozambique; MUS =
Mauritius; MWI = Malawi; NAM = Namibia; NER = Niger; NGA = Nigeria; RWA = Rwanda; SEN = Senegal; TGO = Togo; TZA = Tanzania, United Republic of; UGA = Uganda; ZAF = South Africa; ZMB = Zambia; ZWE = Zimbabwe.

(Continued on next page)


to preserve its current innovation Among these, only Botswana and lower than 100. Eight of them have

1: The Global Innovation Index 2017

momentum. the United Republic of Tanzania improved since 2016: Benin (116th),
This year South Africa takes the improve their GII ranking compared Cameroon (117th), Burkina Faso
top spot among all economies in the to 2016, while Kenya remains stable (120th), Burundi (122nd), Niger
region (57th), followed by Mauritius and the other four economies (South (123rd), Zambia (124th), Togo
(64th), Kenya (80th), Botswana Africa, Mauritius, Namibia, and (125th), and Guinea (126th). See
(89th), the United Republic of Rwanda) lose positions. Box5 for more details.
Tanzania (96th), Namibia (97th), The remaining 17 economies in Because of issues with data cover-
Rwanda (99th), and Senegal (100th). this region can be found at ranks age, Ghana drops out of the GII this

Box 5: Sub-Saharan Africa: The innovation momentum in the most promising region continues (continued)

(Tanzania)perform better than their level of more than half of the seven GII pillars and thus and Mozambiqueare labelled innovation
development would predict (see Figure 5.1 are also labelled pillar outperformers. outperformers within the Sub-Saharan Africa
for details). The innovation achiever econo- Most of these innovation achiever region. Table 5.1 shows the full list of achiev-
mies are shown in red and located above the economies outperform in Institutions, ers and outperformers in this region.
upper-bound, farthest from the trend line. Infrastructure, and Market sophistication; they However, although the regions relatively
A total of eight economies are identified as outperform this year also in Human capital strong performance in innovation signals
performing at development (yellow). In the and research and in Business sophistication, strengths, differences between the innova-
same way, the remaining eight are signalled but not as much as they could. Uganda tion levels of some of its economies still
as performing below development (blue). outperforms in all seven pillars, followed show large disparities. Because, since the big
Kenya, Mozambique, Malawi, Rwanda, by Kenya and Rwanda that do so in six. dip experienced in parts of the region last
Uganda, and Senegal stand out for being South Africa and Tanzania outperform in five; year, economies in Africa aim for economic
innovation achievers at least five times in the while Mauritius, Mozambique, and Niger only recovery in 2017 and in the years following,
past six years. Kenya, the chief innovation in four. Malawi outperforms in three, while and while commodity prices are recovering,
achiever in the region, has been credited Madagascar and Burundi do so in two and it will be important for other less-developed
as such every year since 2011including in therefore are the only innovation achievers economies to keep improving their innova-
2017. With the exception of Malawi, these that are not pillar outperformers. tion performance to maintain the momen-
economies, along with Mauritius, South Africa, This year four of the innovation achievers tum of the regions innovation efforts.
Tanzania, and Niger, outperform their peers in mentioned aboveKenya, Rwanda, Uganda,

Table 5.1: Sub-Saharan Africa: Innovation achievers, pillar outperformers, and innovation outperformers, 201117

Economy Income group Years as an innovation achiever (total) Years as a pillar outperformer (total) outperformer

Kenya Lower-middle income 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7) 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7) Yes
Rwanda Low income 2017, 2016, 2015, 2014, 2012 (5) 2017, 2016, 2015, 2014, 2013, 2012, 2011 (7) Yes
Uganda Low income 2017, 2016, 2015, 2014, 2013 (5) 2017, 2016, 2015, 2014, 2013 (5) Yes
Mozambique Low income 2017, 2016, 2015, 2014, 2012 (5) 2017, 2016, 2015, 2014, 2013, 2012 (6) Yes

Malawi Low income 2017, 2016, 2015, 2014, 2012 (5) 2016, 2015, 2014, 2012, 2011 (5) No
Senegal Low income 2017, 2015, 2014, 2013, 2012 (5) 2017, 2015 (2) No
Madagascar Low income 2017, 2016 (2) 2012 (1) No
Burundi Low income 2017 (1) No
Tanzania, United Rep. Low income 2017 (1) 2017, 2014 (2) No

Note: World Bank Income Group Classification (July 2016): LI = low income; LM = lower-middle income; UM = upper-middle income; and HI = high income. This table includes GII 2017. Economies identified as an innovation
achiever and a pillar outperformer for two or more consecutive years, including 2016 and 2015, are also considered innovation outperformers.

Notes for this box appear at the end of the chapter.

year, while Zimbabwe is added (see Innovation Input Sub-Index and Creative outputs, gains are seen in
1: The Global Innovation Index 2017

Annex 2). Innovation Output Sub-Index, Intangible assets and Online creativ-
respectively, with a place in the top ity, primarily in ICTs and business
Latin America and the Caribbean (18 50 economies across five pillars: model creation, Wikipedia edits, and
economies) Institutions (41st), Infrastructure video uploads on YouTube. Although
Latin America and the Caribbean (47th), Market sophistication (50th), Business environment and Tertiary
includes only upper- and lower- Business sophistication (46th), and education still have room for improve-
middle-income economies, with Knowledge and technology outputs ment, Brazil is also relatively weak in
three exceptions: Chile, Uruguay, (49th). Its improvements in 2017 Credit and Knowledge impact. Some
and Trinidad and Tobago, which lie in Knowledge and technology indicators where the economy could
are all high-income economies. Still outputs, where it gains 10 positions, improve further include PISA results,
leading the region in the GII rank- and Human capital and research graduates in science and engineering,
ings for another year, Chile (46th) (61st), where it advances one spot. tertiary inbound mobility, gross capi-
loses two positions, and is followed In Knowledge and technology out- tal formation, JV-strategic alliance
by Costa Rica (53rd, down by eight) puts, major improvements are in deals, and growth rate of GDP per
and Mexico (58th, up by three). Knowledge diffusion (34th), with worker. Persistence will be needed
Following these countries, and better rankings in IP receipts and in a time of political and economic
ranking in the top half of the GII this FDI net outf lows, and in a number uncertainty to benefit from the
year, is Panama (63rd). The top 100 of individual indicators, including economic uptick as described at the
economies overall include Colombia PCT patent applications, scientific outset of the chapter.
(65th), Uruguay (67th), Brazil (69th), and technical articles, and growth
Peru (70th), Argentina (76th), rate of GDP per worker. In Human Central and Southern Asia (9 economies)
Dominican Republic (79th), Jamaica capital and research, Chile improves Economies of the Central and
(84th), Paraguay (85th), Trinidad and mainly in Education (65th), gaining Southern Asia region have seen fur-
Tobago (91st), Ecuador (92nd), and eight positions since last year and see- ther improvements in their rankings
Guatemala (98th). The remaining ing its ranking in every indicator in since 2016, with seven economies
economies in the region rank below this sub-pillar improve. Tertiary edu- improving their rankings and with
100 in the GII this year: El Salvador cation (55th) also gains one position, India moving into the top half of the
(103rd), Honduras (104th), and the as Chile becomes the 5th economy GII this year.
Plurinational State of Bolivia (106th). in the world in tertiary enrolment. India maintains its top place in
Although important regional Despite the improvements, Chile the region, moving up six spots
potential exists, the GII rankings of still shows areas of weakness in pil- from 66th last year to 60th this year
countries in Latin America relative lar 2, Human capital and research, in overall. The Islamic Republic of Iran
to other regions have not steadily a total of four indicators including becomes 2nd in the region, mov-
improved. In recent years and in government expenditure in educa- ing from 78th to 75th and leaving
2017, no economies from this region tion (60th), pupil-teacher ratio (83rd), its 78th spot to Kazakhstan, which
are identified as innovation achievers tertiary inbound mobility (96th), and drops three positions from 2016. The
(see Box4 in the 2015 edition of the global companies by R&D (43rd). remaining economies rank in order
GII). Brazil is ranked 69th in the GII within the region as follows: Sri Lanka
As previously mentioned, the 2017, the same position as last year. shows a one-position improvement
minimum data coverage threshold Brazils strongest pillar ranking is in this year (90th); this is followed by
rule was adjusted this year to retain Business sophistication (43rd), where Tajikistan (94th), Kyrgyzstan (95th),
only those economies with suffi- it sees one of its highest rankings in Nepal (109th), Pakistan (113th),
cient data coverage in the GII. As a IP payments (8th). Brazils biggest and Bangladesh (114th). Despite the

result, Nicaragua and the Bolivarian improvements are in Human capital improvements in data coverage in
Republic of Venezuela drop from the and research (50th, up by 10) and the region, Bhutan does not meet
GII 2017 (see Annex 2). Creative outputs (83rd, up by 7). In the 66% data coverage threshold (see
Chile ranks 46th in the GII this Human capital and research, Brazil Annex 2) and is thus excluded from
year, at the top spot in the region, improved its rank in all sub-pillars, in the 2017 GII.
but down two positions since 2016. particular in expenditure on educa- India remains 1st in the region
It is ranked 42nd and 53rd in the tion and QS university ranking. In and 6th among lower-middle-income

1: The Global Innovation Index 2017

India ahead of lower-middle and upper-middle income
Figure 6: India ahead of average lower-middle- and upper-middle-income economies


l India
l Upper-middle income
l Lower-middle income
GII 2017 score



Quality of High- & Research Graduates State Gross Global Growth ICT High-tech GERD Intellectual Patent
scientific medium-high talent in in science & of capital R&D rate of services exports performed property families
publications tech business engineering cluster formation companies PPP$ exports by receipts in 2+
manufactures enterprise development GDP/worker business offices

Source: GII 2017 data.

economies. India has also outper- year, including governments online results, and tertiary inbound mobil-
formed on innovation relative to its services, e-participation, logistics ity Indian scores are lower than the
GDP per capita for many years in a performance, gross capital formation, average for lower-middle-income
row (see Figure4). India ranks 60th high-tech imports, and industrial economies. The same is true for
overall in the GII this year, is also designs. Also worth mentioning is other Human capital and research
among the top 50 economies in two the six-position gain in global R&D indicators, including researchers and
pillars: Market sophistication (39th) companies, where India ranks 14th, tertiary enrolment, and for FDI net
and Knowledge and technology considerably better than the respec- inflows. On the output side, a number
outputs (38th). It improves its rank- tive groups of lower- and upper-mid- of indicatorssuch as scientific and
ings in five pillars: Institutions (up 4 dle-income economies on average. technical articles and trademarks by
spots), Infrastructure (up 14 spots), Other such areas in which India does originare lower than upper-mid-
Business sophistication (up 2 spots), far better than most middle-income dle-income economy averages. Other
Knowledge and technology outputs economies include graduates in sci- indicators on the output side that
(up 5 spots), and Creative outputs (up ence and engineering, gross capital show room for improvement include
9 spots). By contrast, Human capi- formation, state of cluster develop- indicators measuring new businesses
tal and research (64th) and Market ment, GERD performed by business, and industrial design filings.
sophistication lose one and six posi- research talent, and patent families in In the same way as other coun-
tions respectively. At the sub-pillar two or more offices on the input side; tries (on Viet Nam, see Box6), India
level, India enjoys its largest gains in and quality of scientific publications, has worked intensively to improve its
areas such as Knowledge absorption, growth rate of GDP per worker, innovation performance, including
Knowledge impact, and Intangible high-tech and ICT services exports, by hosting innovation workshops

assets. Despite remaining a weak sub- high-tech manufactures, and IP and instituting important work in
pillar, India improves in Education, receipts on the output side (Figure6). recent years with the use of the GII,
where it advances four positions India still has more potential. and by instituting a high-level Task
because of better relative government Business environment (121st) is an Force on Innovation to suggest ways
expenditure by pupil. area where the country can improve the country can improve its innova-
At the indicator level, India on most indicators. On the input side, tion eco-system.39 In this context,
improves in a number of areas this in environmental performance, PISA India has considerably improved its

data coverage in the 2016 and 2017 East Asia, East Asia, and Oceania the GII, the Innovation Input Sub-
1: The Global Innovation Index 2017

editions of the GII. Work is ongoing region are ranked within the top Index, and the Innovation Output
to overcome other data issuesfor 100 in the GII. With the exception Sub-Index: Singapore (7th), Korea
example, issues with R&D-related also of Cambodia and of Brunei (11th), Japan (14th), Hong Kong
indicators, such as GERD performed Darussalam, which enters the GII (China) (16th), and New Zealand
by business data dating from 2011 (see this year thanks to improved data (21st). China ranks next (22nd),
Indias Country/Economy Profile for coverage, all other economies in the being the third most efficient econ-
missing or outdated variables). region are also in the top 100 in the omy in the world; Australia follows
Innovation Input Sub-Index, the (at 23rd).
South East Asia, East Asia, and Oceania (15 Innovation Output Sub-Index, and Malaysia moves down two
economies) the Innovation Efficiency Ratio. positions to 37th, due mostly to a
This year all economies but The top f ive economies in the 10-position drop in Institutions
Cambodia (101st) within the South region rank in the top 25 overall for (53rd), a drop driven by lower

Box 6: ASEAN: Singapore and the new Asian Tigers?

Ten out of the 15 economies in the South Through this resolution, the Vietnamese areas of excellence are emerging, while oth-
East Asia, East Asia, and Oceania region are government has assigned responsibilities to ers are still works in progress. For example,
members of the Association of Southeast ministries, agencies, and local governments Viet Nam shows the best score of the group
Asian Nations (ASEAN). These economies to undertake actions to improve Viet Nams in expenditure on education and is also
are Brunei Darussalam, Cambodia, Indonesia, performance, and the Ministry of Science and performing well in ICT use, gross capital
Lao Peoples Democratic Republic, Malaysia, Technology (MOST) has been tasked with formation, and FDI net inflows; at the same
Myanmar, the Philippines, Singapore, coordinating these efforts. A MOST workshop time, it has some of the lowest scores in
Thailand, and Viet Nam. In 2015, intra-ASEAN in cooperation with the World Intellectual tertiary enrolment, state of cluster develop-
exports represented the 26% of exports by Property Organization (WIPO) was organized ment, university/industry research collabora-
ASEAN countries. Electrical machinery and in Hanoi in March 2017 to address missing tion, and knowledge-intensive employment.
equipment is the single most exported com- and outdated data and to help leverage Viet Malaysia ranks second in the ASEAN group
modity within ASEAN. Nams innovation strengths and overcome in expenditure on education, state of cluster
Since the 1980s, Singaporealong with related weaknesses. development, university/industry research
Hong Kong (China), the Republic of Korea, In the broader ASEAN analysis, both dif- collaboration, and ICT use, but has low scores
and, to some extent, Malaysiahas been ferences and similarities in innovation perfor- in PISA scores in reading, maths, and science;
labelled one of the Asian Tigers. Singapore mance are evident across ASEAN economies. tertiary enrolment; and knowledge-intensive
has managed to sustain its high economic Figures 6.1 and 6.2 show the scores of these employment.
growth rate to become one of the richest economies in selected innovation input and Third, the distance between the top
economies in the world. Correspondingly, it output indicators. Three findings emerge performer and the other ASEAN economies
has ranked in the top 10 since the first edition from these figures. First, a certain stability in output indicators is much larger than the
of the GII. In comparison, the other ASEAN exists at the top of the ASEAN rankings. distance in inputs. It takes time for economies
members are less rich and advanced. Singapore has the highest scores among to create the conditions and accumulate
However, some of the ASEAN ASEAN members in all selected indicators, the capabilities required to convert a fertile
economiesin particular, Indonesia, the except for expenditure on education (topped innovation environment and solid innova-
Philippines, Thailand, and Viet Namare by Viet Nam), gross capital formation (topped tion inputs into tangible innovation outputs
now considered to be new Asian Tigers on by Brunei Darussalam), ICT service exports and outcomes. Among ASEAN economies,
the rise. These economies participate more (topped by the Philippines), and trademarks Singapore is the top performer in the selected
and more in a number of regional and global by origin (topped by Thailand). Cambodia is innovation outputs, with two exceptions: ICT
value chains, including some in relatively relatively new in terms of economic catch- services exports, where the Philippines leads;
high-tech sectors. These countries have also up. Although improving, it lags behind in and trademarks by origin, where Viet Nam

become active in improving their innova- most of the input indicators selected here, presents the highest score in the group.
tion performance, sometimes in showcasing although it is second in FDI net inflows Malaysia has the second highest scores in
best practice use of the GII findings, paired among ASEAN economies, foreshadowing patents by origin, scientific and technical
with remarkable innovation results. In 2017, welcome development ahead. articles, and ICT services exports. Thailands
for example, the Vietnamese government Second, each economy is making an strengths are in citable documents and
mandated Resolution 19-2017/NQ-CP. effort to build its innovation system: in each, trademarks by origin, where it places 2nd.

(Continued on next page)


1: The Global Innovation Index 2017

Box 6: ASEAN: Singapore and the new Asian Tigers? (continued)

Figure 6.1: ASEAN scores in selected input indicators



GII 2017 score



Expenditure Tertiary State University/industry ICT use Gross capital Foreign direct Knowledge-intensive
on education enrolment of cluster research collaboration formation investment employment
development net inflows

Figure 6.2: ASEAN scores in selected output indicators



GII 2017 score



Patents by origin Quality of scientific publications Scientific and ICT services export Trademark application
technical publications class count by origin

l Singapore l Viet Nam l Brunei Darussalam l Indonesia

l Malaysia l Thailand l Philippines l Cambodia

Source: GII 2017 data.

Note: No data are available for Lao Peoples Democratic Republic or Myanmar, which are also omitted from the GII 2017.

1 Among other objectives, ASEAN aims to acceler- 2 Data from ASEANstats, available at http://asean. 4 For more information, see Viet Nams Ministry
ate economic growth and socioeconomic devel- org/storage/2016/11/Table18_as-of-6-dec-2016. of Planning and Investment website at
opment, promoting active collaboration and pdf.
mutual assistance on matters of common inter- aspx?idTin=35994&idcm=121.
3 Data from ASEANstats, available at http://asean.
est, including trade. Details are available at http://

rankings in Business environment output indicators, including inten- Israel is the only economy in the
1: The Global Innovation Index 2017

(50th, down by 22) and a 19- sity of local competition, GERD region to rank in the top 10 for any
position fall in Business sophistica- financed by business, patent families pillar (5th, Business sophistication,
tion (48th)driven mainly by its in two or more off ices, patents by up one spot; and 9th, Knowledge
rank in Knowledge workers, where origin, and PCT patent applications. and technology outputs, up three).
Malaysia moved from 35th to 93rd Opportunities for further improve- The country ranks 20th and 14th
this year (see also Box6). The lat- ment still exist, including in ease of in the Innovation Input Sub-Index
ter move is affected by the use of getting credit, growth rate of GDP and Innovation Output Sub-Index,
two more recent data points for per worker, new businesses, and cul- respectively, seeing the most gains
firms offering formal training (from tural and creative services exports. in Tertiary education (62nd, up 11
25th to 79th) and GERD financed spots), Knowledge absorption (9th,
by business (from 11th to 75th). Northern Africa and Western Asia (19 up 7 spots), and Knowledge diffu-
Malaysia is also among the middle- economies) sion (8th, up 6 spots). Israel keeps
income economies that are the clos- Israel (17th) and Cyprus (30th) its 1st place in researchers, venture
est to the top 25 this year (see Box4 achieve the top two spots in the capital deals, GERD performed
on the innovation divide). region for the fifth consecutive year, by business, and research talent in
Viet Nam, by contrast, gains 12 improving by four and one positions, business enterprise. It also gains top
positions this year, ranking 47th. respectively. Third in the region is 3 positions in gross expenditure
Viet Nam maintains its top place the United Arab Emirates (35th) on R&D (1st), university/industry
among lower-middle-income econ- which moves up six places from research collaboration (3rd), ICT
omies and enters the worlds top 10 last year, the most striking upward services export (1st), and Wikipedia
in the Innovation Efficiency Ratio move in the region. In the case of edits (3rd). Weaknesses for Israel are
(see Box 6). Thailand (51st) and the United Arab Emirates, data col- found in the input side of the GII
Mongolia (52nd) follow Viet Nam, laboration has also increased data and are more prominent in variables
ranking in the top half of the GII availability, reducing missing values such as gross fixed capital formation.
this year as well. Brunei Darussalam, from 17 last year to 11 this year. On the output side, two areas show
the Philippines, and Indonesia rank Important data points, however, are possibilities for improvement: the
71st, 73rd, and 87th, respectively. still missing, making it difficult to growth rate of GDP per worker and
Cambodia closes the rankings for evaluate certain pillars, most nota- trademarks by origin.
the region, coming in at 101st. bly in Education, where three out of
Japan has risen in the GII rank- five variables are not available, and Europe (39 economies)
ings each year for the last four years, Knowledge workers, with two out In this years edition of the GII,
moving up to 14th in 2017. Japan of five indicators missing. 15 of the top 25 economies come
ranks 11th overall in the Innovation Sixteen of the 19 economies in from Europe. This region is home
Input Sub-Index and 20th overall in the Northern Africa and Western to the top 3 economies of the GII
the Innovation Output Sub-Index, Asia region are in the top 100, 2017: Switzerland (1st), Sweden
up by four positions since 2016. including Turkey (43rd), Qatar (2nd), and the Netherlands (3rd).
This year Japan improves its rank in (49th), Saudi Arabia (55th), Kuwait Following these regional leaders
Institutions (13th) and Knowledge (56th), Armenia (59th), Bahrain among this group of top 25 are the
and technology outputs (12th), (66th), Georgia (68th), Morocco UK (5th), Denmark (6th), Finland
where it advances in all sub-pillars. (72nd), Tunisia (74th), Oman (77th), (8th), Germany (9th), Ireland (10th),
Japan ranks in the top 10 economies Lebanon (81st), Azerbaijan (82nd), Luxembourg (12th), Iceland (13th),
for six sub-pillars: Research and and Jordan (83rd). Of all the econo- France (15th), Norway (19th),
development (3rd), Information and mies in the region, Kuwait sees the Austria (20th), the Czech Republic

communication technologies (5th), most improvement in its overall GII (24th), and Estonia (25th). It should
Trade, competition, and market ranking, having moved up 11 spots. be noted that most of the economies
scale (3rd), Knowledge absorption Israel moves up four places, in this region have the fewest miss-
(8th), Knowledge creation (9th), and from 21st to 17th in 2017, remain- ing values, leading them to display
Knowledge diffusion (10th). Japan ing number 1 in the Northern the most accurate GII rankings (see
ranks 1st in a number of input and Africa and Western Asia region. Annex 2). This includes the following

economies with 100% data coverage creation. France loses the most posi- been whether the GII model

1: The Global Innovation Index 2017

in the Innovation Input Sub-Index, tions in Infrastructure (12th), and in c a n be appl ie d at t he sub -
the Innovation Output Sub-Index, all its sub-pillars, including losses national level to assess innova-
or both: Denmark, Finland, of the top spots in governments tion clusters more broadly. Vari-
Germany, France, Austria, the Czech online service and e-participation. ous countries have approached
Republic, Italy, Portugal, Bulgaria, Furthermore, France becomes rela- the GII co-publishers to create
Poland, Hungary, Romania, and the tively weak in pupil-teacher ratio, regional innovation indices on
Russian Federation. while retaining all the other areas the basis of the GII model. In
Eighteen economies follow of weaknesses that it presented last January 2017, the Indian gov-
among the top 50 and have main- year. ernment decided to rank the
tained relatively stable rankings since performance of Indian states in
2014: Malta (26th), Belgium (27th), the India Innovation Index.40
Spain (28th), Italy (29th), Portugal Assessing regional innovation clusters
A shared conviction underlying
(31st), Slovenia (32nd), Latvia (33rd), This year the GII makes a f irst
both points is that the interaction of
Slovakia (34th), Bulgaria (36th), attempt at assessing sub-national
critical innovation inputs and out-
Poland (38th), Hungary (39th), innovation clusters. The Special
puts happens at the local level, and
Lithuania (40th), Croatia (41st), Section on Clusters in this report sets
this phenomenon requires improved
Romania (42nd), Greece (44th), out the approach and main findings
metrics. Yet this is where the prob-
the Russian Federation (45th), in more detail.
lem lies, as shown in Table7.41
Montenegro (48th, which joins The importance of innovation
Despite the progress that has
the top 50 this year), and Ukraine hubs at the sub-national and interna-
been made, measuring the territorial
(which joins the top 50 this year at tional level has been at the forefront
dimension of innovation remains
the 50th position, moving up by six). of GII discussions for the last 10 years
challenging. Only a few GII indi-
The remaining European econo- for two main reasons.
cators are readily available at the
mies remain among the top 100 econ- First, successful innovation clus-
regional or city level for a large set
omies overall. The regions rankings ters, and thus agglomerations of
of countries. A case in point is that,
continue as follows: the Republic innovation activity, are consid-
at this time, the GII model relies on
of Moldova (54th), the Former ered essential for national inno-
a survey-based question to assess the
Yugoslav Republic of Macedonia vation performance. By pooling
state of cluster development (indi-
(61st), Serbia (62nd), Bosnia and talent, know-how, research labs,
cator 5.2.1) rather than official data.
Herzegovina (86th), Belarus (88th), and manufacturing capabilities
As a testament to imperfect data
and Albania (93rd), with Serbia and they constitute spikes or peaks
availability on this critical innova-
Bosnia and Herzegovina as the only of excellence with critical inno-
tion dimension, efforts to replace
improving economies in this group. vation linkages. A discussion on
this variable with hard data from
France moves up another three this issue has been at the fore-
recognized sources have so far failed.
spots in 2017, from 18th to 15th. front of almost every GII edi-
Besides, clusters often do not stop
France ranks 15th in the Innovation tion. In particular, the GII 2013
at national borders. By def inition,
Input Sub-Index and gains one spot on the theme Local Dynamics
they thus do not map to nationally
in the Innovation Output Sub-Index of Innovation analysed clusters,
available data sources; the search for
(18th). It ranks in the top 25 econo- asking which kinds of linkages
readily available data is elusive.
mies in all pillars, showing improve- exist among them, and to what
To make progress on this front, a
ments in Institutions (24th), Market ex tent k nowledge spi l lovers
first step is to identify clusters in an
sophistication (11th), Knowledge occur. Impor tantly, some of
innovative way. The GII 2017 edi-
and technology outputs (20th), and these clusters are international
tion makes progress in this regard. In

Creative outputs (12th). Frances in nature. They do not coincide

the Special Section on Clusters at the
three most-improved sub-pillars with boundaries of sub-national
end of the report, Bergquist, Fink,
Investment (10th), Knowledge cities or regions; rather they
and Raffo propose a novel approach
impact (36th), and Intangible assets cross national borders.
to assess the inventive capacity in
(7th)gain positions in market cap-
Second, over the last 10 years, clusters based on patenting data. By
italization, growth rate of GDP per
one of the most frequent ques- the means of inventor addresses, and
worker, and ICT and business model
tions asked by countr ies has using underlying geo-coding, the

Table 7: Top cluster of countries or cross-border regions within the top 100 addressing this challenge will require
1: The Global Innovation Index 2017

a mix of technological and non-tech-

Rank Cluster name Territory(ies)
nological solutions: organizational
1 TokyoYokohama Japan
changes, public and private invest-
2 ShenzhenHong Kong (China) China/Hong Kong (China)
ment in R&D, and more effective
3 San JoseSan Francisco, CA United States
technology transfer mechanisms are
4 Seoul Korea, Rep.
all important elements of agri-food
10 Paris France
innovation systems.
12 FrankfurtMannheim Germany
Historically, innovation in agri-
18 Eindhoven Netherlands/Belgium
culture has proven not only feasible
21 London United Kingdom
but highly successful. Today, a new
22 Tel Aviv Israel
innovation drive is required among
24 Stockholm Sweden high-, middle-, and low-income
31 Zurich Switzerland/Germany economies. In high- and middle-
34 HelsinkiEspoo Finland income economies, a new innovation
35 Singapore Singapore wave is on the horizon: innovations
36 Basel Switzerland/France/Germany from other sectors are spilling over
39 Copenhagen Denmark to agricultural and food systems,
43 Bengaluru India making them smart and digital. In
44 Sydney Australia low-income economies, the focus
45 RotterdamThe Hague Netherlands is on reducing the bottlenecks of
47 Montreal, QC Canada agri-food innovation systems, while
52 Barcelona Spain speeding up innovation convergence
54 BrusselsLeuven Belgium with more productive economies. In
57 Moscow Russian Federation all economies, public policy is fun-
58 Milan Italy damental to promoting an enabling
65 Lausanne Switzerland/France environment that encourages tech-
71 Vienna Austria nology uptake, entrepreneurship and
82 Aachen Germany/Netherlands/Belgium skills, and innovation. The remaining
92 Kuala Lumpur Malaysia chapters of the report provide more
Source: Derived from Annex 2 of the Special Section on Clusters.
details on this years theme from aca-
demic, business, and particular coun-
try perspectives from leading experts
and decision makers.
This chapter has also presented
authors identify the largest inventive important component of the GII the main GII 2017 results, distill-
clusters as measured by PCT patent- and other innovation measurement ing main messages and noting some
ing activity, possibly up to the street efforts. important evolutions that have taken
level thanks to advanced mapping place since last year. Three main
techniques. Table7 presents some of findings stand out. Firstand in a
the leading innovation clusters that Conclusions turn of eventsa novel and more
result from this analysis. The theme for this years GII is sustained growth momentum is cur-
In the coming years, attempts Innovation Feeding the World. This rently in place. Second, more rapid
to foster data on local innovation chapter has provided an overview of economic growth can lay the founda-

clusters should receive increased the current trends, strategies, and tion for innovation-driven economic
attention, and consideration of clus- policies for innovation in agriculture development, but more investment
ters may possibly become a more and food systems. Within agri-food would be needed to boost productiv-
systems, innovation needs to be a ity growth, which is still at historic
priority to achieve sustainable pro- lows. To this end, R&D efforts from
ductivity growth and address the both the public and private sector
global food challenge. Successfully would also need to be intensified.

Third, while the GII results point 3 The general trend line is defined by the 10 Fernald, 2014. See also Chapter 1 in WIPO

1: The Global Innovation Index 2017

scores and economic development level 2015.
to a certain stability at the top, new of all countries considered in the GII. The
opportunities are emerging: new threshold bounds are defined as 10% above 11 WTO, 2017.
and 10% below the scores defined by trend
Asian Tigers are active in improving line (see Box 2 in Escalona Reynoso et al.,
12 UNCTAD, 2016, 2017.

their innovation performance, and 2015). 13 Cornell et al., 2016; WIPO, 2015, 2017
new innovation actors from vari- (forthcoming). On slowing technology
4 In addition to these 9 Sub-Saharan Africa diffusion see also Andrews et al., 2015; Decker
ous regions are climbing in the GII countries, 26 countries (35 total) were et al., 2016; Haltiwanger, 2011; Haltiwanger et
identified as pillar outperformers this year.
rankings. These come from Europe (9); South East Asia,
al., 2014; OECD, 2015.

Over the last years, the GII has East Asia, and Oceania (6); Latin America 14 See Lee, 2016, for the case of Korea, for
and the Caribbean (5); Northern Africa and instance.
established itself as a leading refer- Western Asia (4); and Central and Southern
ence on innovation, becoming a Asia (2). 15 IMF, 2017; UNCTAD, 2017; WTO, 2017.
The productivity forecast draws on The
tool for action for decision makers 5 This can be partially attributed to the higher Conference Board, Total Economy Database
wishing to improve their countries overall average scores in both of these (adjusted version), May 2017 release, available
indicators displayed by the region, which at
innovation performance. Numerous makes it harder for individual countries to economydatabase/.
workshops in different countries have perform above that level.
16 OECD, 2009, 2017a.
brought innovation actors together, 6 For a country to be labelled an innovation
17 IMF, 2016.
helped improve data availability, and outperformer it has to be identified as an
innovation achiever and it must also score 18 These estimates are based on preliminary
contributed to designing effective above its income group average in four calculations using GDP, GERD, and BERD
innovation policies. These exchanges or more GII pillars for two or more years, figures at constant $PPP 2005 prices from
including the two most recent2015
on the ground also generate feedback and 2016. In 2017, 10 economies were
the UNESCO-UIS Science & Technology Data
Center, updated March 2017. Economies
that, in turn, improves the GII and identified as innovation outperformers. The included: Afghanistan, Albania, Algeria,
other countries identified as innovation
assists the journey towards improved outperformers this year are Viet Nam, the
Angola, Antigua and Barbuda, Argentina,
Armenia, Australia, Austria, Azerbaijan,
innovation measurement and policy. Republic of Moldova, India, Armenia, Ukraine, Bahamas, Bahrain, Bangladesh, Barbados,
This valuable feedback will continue and Tajikistan. See Escalona Reynoso et al. Belarus, Belgium, Belize, Benin, Bermuda,
(2015) for more details. Bhutan, Bolivia (Plurinational State of),
to be integrated into future iterations Bosnia and Herzegovina, Botswana, Brazil,
of this lead chapter of the GII in the Brunei Darussalam, Bulgaria, Burkina Faso,
years to come. Burundi, Cabo Verde, Cambodia, Cameroon,
Notes for Chapter 1 Canada, Central African Republic, Chad,
Chile, China, China (Hong Kong Special
1 Conference Board, 2017; IMF, 2017; OECD, Administrative Region), China (Macao Special
2017a. According to the World Bank (2017), Administrative Region), Colombia, Comoros,
Notes for Box 5 the world economy will grow at 2.7% in 2017, Congo, Costa Rica, Croatia, Cuba, Cyprus,
1 In 2011 most innovation achievers were up by 0.4% from 2016, with a downward Czech Republic, Cte dIvoire, Democratic
located in the South East Asia, East Asia, and revision of 0.1% from June 2016. For 2018, Republic of the Congo, Denmark, Djibouti,
Oceania region. In 2012 and 2013 Europe and the OECD (2017a) and IMF (2017) forecast a Dominica, Dominican Republic, Ecuador,
Sub-Saharan Africa shared the same number growth rate of 3.6% without recent revisions. Egypt, El Salvador, Equatorial Guinea, Eritrea,
of innovation achievers: six and four in each The World Bank (2017) predicted global Estonia, Ethiopia, Fiji, Finland, France, Gabon,
year, respectively. GDP growth at 2.9%, and recently revised it Gambia, Georgia, Germany, Ghana, Greece,
downward by 0.1%. Grenada, Guatemala, Guinea, Guinea-Bissau,
2 This can be partially attributed to Guyana, Haiti, Honduras, Hungary, Iceland,
improvements to data coverage. A stricter 2 IMF, 2017.
India, Indonesia, Iran (Islamic Republic of),
cut-off rule that increases the minimum 3 IMF, 2017; OECD, 2017a; World Bank, 2017. Iraq, Ireland, Israel, Italy, Jamaica, Japan,
required threshold for all countries in the Jordan, Kazakhstan, Kenya, Kiribati, Kuwait,
GII to at least 66% of all indicators in each 4 IMF, 2017, with Russian GDP growth recently Kyrgyzstan, Lao Peoples Democratic
of the sub-indices was introduced this year revised upwards. Republic, Latvia, Lebanon, Lesotho, Liberia,
(see Appendix IV: Technical Notes for more Libya, Lithuania, Luxembourg, Madagascar,
details). This procedure translates into more 5 World Bank, 2017.
Malawi, Malaysia, Maldives, Mali, Malta,
precise measurements of the innovation 6 Adler et al., 2017; OECD, 2017a; WIPO, 2015; Marshall Islands, Mauritania, Mauritius,
performance of each country and thus into World Bank, 2017. Mexico, Micronesia (Federated States
a better identification of those that can of), Mongolia, Montenegro, Morocco,
be identified as innovation achievers. As a 7 World Bank, 2017. Mozambique, Namibia, Nepal, Netherlands,
result of this improvement, however, two New Zealand, Nicaragua, Niger, Nigeria,
8 Adler et al., 2017; Cornell et al., 2016.

economies from this region identified as Norway, Oman, Pakistan, Palau, Palestine,
innovation achievers in previous years are no Estimates indicate that worldwide
productivity growth slowed down in 2015 Panama, Papua New Guinea, Paraguay,
longer in the GII ranks: Gambia (2014) and Peru, Philippines, Poland, Portugal, Puerto
Ghana (2011). and remained at the same modest rate of
1.5% in 2016 (Conference Board, 2016, 2017). Rico, Qatar, Republic of Korea, Republic
of Moldova, Romania, Russian Federation,
9 The Conference Board, Total Economy Rwanda, Saint Kitts and Nevis, Saint Lucia,
Database (adjusted version), May 2017 Saint Vincent and the Grenadines, Samoa, Sao
release, available at http://www.conference- Tome and Principe, Saudi Arabia, Senegal, Serbia, Seychelles, Sierra Leone, Singapore,
Slovakia, Slovenia, Solomon Islands, South
Africa, Spain, Sri Lanka, Sudan, Suriname,

Swaziland, Sweden, Switzerland, Taiwan 34 On informal actors, see Kraemer-Mbula and Cornell University, INSEAD, and WIPO. 2015.
1: The Global Innovation Index 2017

(China), Tajikistan, Thailand, The Former Wunsch-Vincent, 2016. The Global Innovation Index 2015: Effective
Yugoslav Republic of Macedonia, Timor-Leste, Innovation Policies for Development, eds. S.
Togo, Tonga, Trinidad and Tobago, Tunisia, 35 Economies are grouped according to the Dutta, B. Lanvin, and S. Wunsch-Vincent.
Turkey, Turkmenistan, Tuvalu, Uganda, World Bank classification (July 2016) gross Ithaca, Fontainebleau, and Geneva: Cornell,
Ukraine, United Arab Emirates, United national income (GNI) per capita, calculated INSEAD, and WIPO.
Kingdom of Great Britain and Northern using the World Bank Atlas method. The
Ireland, United Republic of Tanzania, United groups are: low income, US$1,025 or . 2016. The Global Innovation Index 2016:
States of America, Uruguay, Uzbekistan, less; lower-middle income, US$1,026 to Winning with Global Innovation, eds. S. Dutta,
Vanuatu, Venezuela (Bolivarian Republic of), US$4,035; upper-middle income, US$4,036 to B. Lanvin, and S. Wunsch-Vincent. Ithaca,
Viet Nam, Yemen, Zambia, Zimbabwe. US$12,475; and high income, US$12,476 or Fontainebleau, and Geneva: Cornell, INSEAD,
more. and WIPO.
19 The top three spenders relative to GDP are
Israel, Korea, and Japan, with Israel overtaking 36 Since 2012, the regional groups have been Decker, R., J. Haltiwanger, R.S. Jarmin, and J. Miranda.
Korea in 2015. Based on our estimates, China based on the United Nations Classification: 2016. Where Has All the Skewness Gone?
is the only emerging economy with R&D EUR = Europe; NAC = Northern America; LCN The Decline in High-Growth (Young) Firms
intensity above the global average. Other = Latin America and the Caribbean; CSA = in the U.S., European Economic Review 86
middle-income economies, such as Malaysia, Central and Southern Asia; SEAO = South (July): 4-23.
Brazil, India, and South Africa, present lower East Asia, East Asia, and Oceania; NAWA =
Northern Africa and Western Asia; and SSF = Dutta, S., D. Benavente, B. Lanvin, and S. Wunsch-
R&D intensities, between 1.3% and 0.7%. Vincent. 2013. The Global Innovation Index
Sub-Saharan Africa.
20 Cornell et al., 2016; OECD, 2017b. 2013: Local Dynamics Keep Innovation Strong
37 To address their inherent volatility (see in the Face of Crisis. In The Global Innovation
21 Despite these aggregate figures, some previous GII editions) and thus reduce the Index 2013: The Local Dynamics of Innovation,
surveys indicate that top world R&D swings in the ranking induced by FDI flows, eds. S. Dutta and B. Lanvin. Ithaca and
companies raised their R&D expenditures in this year the GII takes 3-year averages of FDI Fontainebleau: Cornell, INSEAD. 3-67.
2015 and 2016 (European Commission, 2016; net inflows and outflows (see Annex 2).
Strategy&, 2016). Dutta, S., R. Escalona Reynoso, A. Bernard, B. Lanvin,
38 Note that any assessment of how the UKs and S. Wunsch-Vincent. 2015. The Global
22 WIPO, 2016. At the same time, worldwide planned withdrawal from the European Innovation Index 2015: Effective Innovation
patent applications under WIPOs Patent Union affected the countrys GII rank would Policies for Development. In The Global
Cooperation Treaty (PCT) saw a 1.4% increase be speculative, at best. First, most of the data Innovation Index 2015: Effective Innovation
in 2015; a significant fall in growth compared predate the actual related referendum. As is Policies for Development, eds. S. Dutta, B.
with previous years (WIPO, 2016). the case with other high-income countries, Lanvin, and S. Wunsch-Vincent. Geneva,
37% of the UKs indicators are from 2016; Ithaca, and Fontainebleau: Cornell, INSEAD,
23 OECD, 2009, 2017b; WIPO, 2015. the remaining 63% reflect 2015 and earlier and WIPO. 363.
24 A recent IMF analysis shows that, if advanced years. Second, the causal relations between
plans or the actual withdrawal from the EU Escalona Reynoso, R., A. L. Bernard, M. Saisana, M.
economies increased private R&D by 40% on Schaaper, F. Guadagno, and S. Wunsch-
average, they could increase their GDP by 5% and the 2016 GII indicators are complex and
uncertain in size and direction. Vincent. 2015. Benchmarking Innovation
in the long term (IMF, 2016). Performance at the Global and Country
25 FAO, 2016. 39 See the Preface to this report by the Levels. In The Global Innovation Index 2015:
Confederation of Indian Industry. Effective Innovation Policies for Development,
26 FAO et al., 2015. eds. S. Dutta, B. Lanvin, and S. Wunsch-
40 Government of India, Press Information Vincent. Geneva, Ithaca, and Fontainebleau:
27 FAO et al., 2015. Bureau, 2017. Cornell, INSEAD, and WIPO. 6580.
28 Malnutrition manifests itself in various forms 41 See also Dutta et al., 2013; Hollanders, 2013; European Commission. 2016. The 2016 EU Industrial
beyond undernutrition, such as micronutrient Primi, 2013. R&D Investment Scoreboard. Authors Hctor
malnutrition, obesity, calorie deficiencies, Hernndez, Alexander Tbke, Fernando
anemia, or diabetes (IFPRI, 2016). See also Hervs, Antonio Vezzani, Mafini Dosso, Sara
Chapter 6. Amoroso, and Nicola Grassano. Seville, Spain:
29 Pingali, 2012. References and sources European Commission, Joint Research Centre.

30 It was estimated that in the absence of the Adler, G. R. Duval, D. Furceri, S. Kili elik, K. Evenson, R. E., and D. Gollin. 2003. Assessing the
green revolution, crop yields in developing Koloskova, and M. Poplawski-Ribeiro. Impact of the Green Revolution, 1960 to
countries would have decreased by 23.5%, 2017. Gone with the Headwinds: Global 2000. Science 300: 75862.
with prices between 35% and 66% higher Productivity. IMF Staff Discussion Note 17/04.
Washington, DC: IMF. FAO (Food and Agriculture Organization of the
in 2000. Caloric intake would have fallen by United Nations). 2016. The State of Food and
14.4%, and the percentage of malnourished Andrews, D., C. Criscuolo, and P. Gal. 2015. Frontier Agriculture 2016: Climate Change, Agriculture
children would have increased by 8% firms, technology diffusion and public policy: and Food Security. Rome: FAO.
(Evenson and Gollin, 2003). micro evidence from OECD countries. OECD
Productivity Working Papers No. 2. Paris, OECD FAO, IFAD, and WFP (Food and Agriculture
31 Juma, 2011, 2015; Juma and Gordon, 2015. Organization of the United Nations,
32 See Dutta et al., 2015. International Fund for Agricultural

Conference Board. 2016. Global Economic Outlook Development, and World Food Programme).
33 See, for example, WIPO, 2011. See also the 2016: The Global Economy in a Holding Pattern. 2015. The State of Food Insecurity in the World
ongoing WIPO project on International November 2015. New York: The Conference 2015. Meeting the 2015 International Hunger
Comparison of Knowledge Transfer Policies Board. Targets: Taking Stock of Uneven Progress.
and Practices in collaboration with the Rome: FAO.
Chinese Ministry of Science and Technology . 2017. Global Economic Outlook 2017: Bucking
(MOST); further details are available at http:// the TrendOvercoming Uncertainty, Shocks, Fernald, J. 2014. Productivity and Potential Output and Disruption with Qualitative Growth. before, during, and after the Great Recession.
studies/. November 2016. New York: The Conference NBER Working Paper 20248. Cambridge, MA:
Board. National Bureau of Economic Research.

Government of India, Press Information Bureau. Pingali, P. L. 2012. Green Revolution: Impacts,

1: The Global Innovation Index 2017

2017. Amitabh Kant Launches India Limits, and the Path Ahead. Proceedings of
Innovation Index. Press Information Bureau, the National Academy of Sciences of the United
Government of India, NITI Aayog. 2 February States of America 109 (31): 1230208.
2017. Available at
PrintRelease.aspx?relid=157941. Primi, A. 2013. The Evolving Geography of
Innovation: A Territorial Perspective. In
Haltiwanger, J. 2011. Firm Dynamics and The Global Innovation Index 2013: The Local
Productivity Growth. EIB Papers 16 (1): Dynamics of Innovation, eds. S. Dutta and B.
11636. Lanvin. Ithaca and Fontainebleau: Cornell,
INSEAD. 6978.
Haltiwanger, J., I. Hathaway, and J. Miranda. 2014.
Declining Business Dynamism in the U.S. Strategy&. 2016. 2016 Global Innovation 1000:
High-Technology Sector. The Kauffman Software-as-a-Catalyst. Fact Pack. October
Foundation. Available at http://www. 2016. PwC. Available at https://www.
research%20reports%20and%20 Global-Innovation-1000-Fact-Pack.pdf.
dynamism_in_us_high_tech_sector.pdf. UNCTAD (United Nation Conference on Trade
and Development). 2016. Global Investment
Hollanders, H. 2013. Measuring Regional Innovation: Trends Monitor No. 24. October 2016. Geneva
A European Perspective. In The Global and New York: UNCTAD. Available at
Innovation Index 2013: The Local Dynamics of
Innovation, eds. S. Dutta and B. Lanvin. Ithaca webdiaeia2016d3_en.pdf.
and Fontainebleau: Cornell, INSEAD. 7986.
. 2017. Global Investment Trends Monitor No.
IFPRI (International Food Policy Research Institute). 25. February 2017. Geneva and New York:
2016. Global Nutrition Report 2016: From UNCTAD. Available at
Promise to Impact: Ending Malnutrition by 2030. PublicationsLibrary/webdiaeia2017d1_en.pdf.
Washington, DC: IFPRI.
WIPO (World Intellectual Property Organization)
IMF (International Monetary Fund). 2016. Fiscal 2011. Harnessing Public Research for
Policies for Innovation and Growth. In Fiscal Innovation: The Role of Intellectual Property.
Monitor: Acting Now, Acting Together. April In World Intellectual Property Report 2011: The
2016. Washington, DC: IMF. Chapter 2. Changing Face of Innovation. Geneva: WIPO.
Chapter 4.
. 2017. Global Prospects and Policies. World
Economic Outlook (WEO): Gaining Momentum? . 2015. World Intellectual Property Report:
April 2017. Washington, DC: IMF. Breakthrough Innovation and Economic
Growth. Geneva: WIPO.
Juma, C. 2011. Preventing Hunger: Biotechnology Is
Key. Nature 479: 47172. . 2016. World Intellectual Property Indicators
2016. Geneva: WIPO.
. 2015. The New Harvest: Agricultural Innovation
in Africa. New York: Oxford University Press. . 2017 (forthcoming). World Intellectual
Property Report: Intangible Assets and Global
Juma, C. and K. Gordon. 2015. Taking Root: Global Value Chains. Geneva: WIPO.
Trends in Agricultural Biotechnology.
Discussion Paper, Belfer Center for Science World Bank. 2017. Global Outlook: Subdued
and International Affairs. Growth, Shifting Policies, Heightened
Uncertainty. Global Economic Prospects
Kraemer-Mbula, E. and S. Wunsch-Vincent. 2016. 2017: Weak Investment in Uncertain Times.
The Informal Economy in Developing Nations: Washington, DC: World Bank Group.
Hidden Engine of Innovation? New York:
Cambridge University Press. WTO (World Trade Organization) 2017. Trade
Recovery Expected in 2017 and 2018, Amid
Lee, K. 2016. Economic Catch-Up and Technological Policy Uncertainty. World Trade Organization
Leapfrogging: The Path to Development and Press Release. PRESS/793. 12 April 2017.
Macroeconomic Stability in Korea. Cheltenham, Geneva: WTO.
UK: Edward Elgar Publishing.

OECD (Organisation for Economic Co-operation

and Development). 2009. Policy Responses
to the Economic Crisis: Investing in Innovation
for Long-Term Growth, eds. D. Guellec and S.
Wunsch-Vincent. Paris: OECD Publishing.

. 2015. The Future of Productivity. Paris: OECD



. 2017a. OECD Interim Economic Outlook.

March 2017. Paris: OECD Publishing.

. 2017b. Main Science and Technology

Indicators (MSTI). Last update: MSTI 7
February 2017. Available at http://www.oecd.
ANNEX 1 47

The Global Innovation Index (GII) Conceptual Framework

Annex 1: The GII Conceptual Framework

rich database of detailed metrics for in the Oslo Manual developed by
The rationale for the Global Innovation
refining innovation policies. the European Communities and
The GII is not meant to be the the Organisation for Economic
The Global Innovation Index (GII)
ultimate and definitive ranking of Co-operation and Development
project was launched by Professor
economies with respect to innova- (OECD):2
Dutta at INSEAD in 2007 with the
tion. Measuring innovation outputs An innovation is the implementation
simple goal of determining how to
and impacts remains difficult, hence of a new or significantly improved
f ind metrics and approaches that product (good or service), a new
great emphasis is placed on measur-
better capture the richness of inno- process, a new marketing method, or a
ing the climate and infrastructure new organizational method in business
vation in society and go beyond such
for innovation and on assessing practices, workplace organization, or
traditional measures of innovation as external relations.
related outcomes.
the number of research articles and
Although the end results take the This definition ref lects the evo-
the level of research and develop-
shape of several rankings, the GII is lution of the way innovation has
ment (R&D) expenditures.1
more concerned with improving been perceived and understood over
There were several motivations
the journey to better measure and the last two decades.3
for setting this goal. First, innovation
understand innovation and with Economists and policy makers
is important for driving economic
identifying targeted policies, good used to focus on R&D-based tech-
progress and competitiveness
practices, and other levers that foster nological product innovation, largely
both for developed and developing
innovation. The rich metrics can be produced in-house and mostly in
economies. Many governments are
usedon the level of the index, the manufacturing industries. This
putting innovation at the centre of
sub-indices, or the actual raw data type of innovation was performed
their growth strategies. Second, the
of individual indicatorsto moni- by a highly educated labour force
definition of innovation has broad-
tor performance over time and to in R&D-intensive companies. The
enedit is no longer restricted to
benchmark developments against process leading to such innovation
R&D laboratories and to published
countries in the same region or was conceptualized as closed, inter-
scientific papers. Innovation could
income classification. nal, and localized. Technological
be and is more general and hori-
Drawing on the expertise of breakthroughs were necessarily
zontal in nature, and includes social
the GIIs Knowledge Partners and radical and took place at the global
innovations and business model
its prominent Advisory Board, the knowledge frontier. This charac-
innovations as well as technical
GII model is continually updated terization implied the existence of
ones. Last but not least, recogniz-
to ref lect the improved availability leading and lagging countries, with
ing and celebrating innovation in
of statistics and our understanding low- or middle-income economies
emerging markets is seen as critical
of innovation. This year the model only catching up.
for inspiring peopleespecially the

continues to evolve, although its Today innovation capability is

next generation of entrepreneurs and
mature state now requires only seen more as the ability to exploit
minor updates (refer to Annex 2). new technological combinations; it
Now in its 10th edition, the GII
helps to create an environment in embraces the notion of incremental
An inclusive perspective on innovation innovation and innovation with-
which innovation factors are under
The GII adopts a broad notion of out research. Non-R&D innova-
continual evaluation, and it provides
innovation, originally elaborated tive expenditure is an important
a key tool for decision makers and a

Figure 1: Framework of the Global Innovation Index 2017

Annex 1: The GII Conceptual Framework

Global Innovation Index


Innovation Efficiency Ratio


Innovation Input Innovation Output

Sub-Index Sub-Index

Human Knowledge and

capital and Market Business technology Creative
Institutions research Infrastructure sophistication sophistication outputs outputs

Political Knowledge Knowledge Intangible

environment Education ICTs Credit workers creation assets

Regulatory Tertiary General Innovation Knowledge Creative goods

environment education infrastructure Investment linkages impact and services

Business Research & Ecological competition, Knowledge Knowledge Online
environment development sustainability & market scale absorption diffusion creativity

component of reaping the rewards of scarce.5 For example, there are no obtained are from 2016, 38.1% are
technological innovation. Interest in official statistics on the amount of from 2015, 11.3% are from 2014,
understanding how innovation takes innovative activitydefined as the 5.7% from 2013, and the small
place in low- and middle-income number of new products, processes, remainder 6.3% from earlier years.6
countries is increasing, along with an or other innovationsfor any given
awareness that incremental forms of innovation actor, let alone for any The GII conceptual framework
innovation can impact development. given country (see Box 1, Annex 1 The GII is an evolving project that
Furthermore, the process of innova- of Chapter 1 in the GII 2013). Most builds on its previous editions while
tion itself has changed significantly. measures also struggle to appropri- incorporating newly available data
Investment in innovation-related ately capture the innovation outputs and that is inspired by the latest
activity has consistently intensified of a wider spectrum of innovation research on the measurement of
at the f irm, country, and global actors, such as the services sector or innovation. This year the GII model
levels, adding both new innovation public entities. includes 127 countries/economies,
actors from outside high-income The GII aims to move beyond which represent 92.5% of the worlds
economies and nonprofit actors. The the mere measurement of such population and 97.6% of the worlds
structure of knowledge production simple innovation metrics. To do so GDP (in current US dollars). The

activity is more complex and geo- will require the integration of new GII relies on two sub-indicesthe
graphically dispersed than ever. variables, with a trade-off between Innovation Input Sub-Index and the
A key challenge is to find metrics the quality of the variable on the one Innovation Output Sub-Index
that capture innovation as it actually hand and achieving good country each built around pillars. Four mea-
happens in the world today.4 Direct coverage on the other hand. sures are calculated (see Figure 1):
official measures that quantify inno- The timeliest possible indicators
vation outputs remain extremely are used for the GII: 38.7% of data

1. Innovation Input Sub-Index: Table 1a: Institutions pillar

Annex 1: The GII Conceptual Framework

Average value by income group
Five input pillars capture ele-
High Upper-middle Lower-middle Low
ments of the national economy Indicator income income income income Mean
that enable innovative activities. 1 Institutions
1.1 Political environment
2. Innovation Output Sub-Index: 1.1.1 Political stability and safety* ................................................0.69 ..............0.23 ..............0.80 ..............0.66 ..............0.06
Innovation outputs are the re- 1.1.2 Government effectiveness* .................................................1.21 .................0.04 ..............0.50 ..............0.78 .................0.26
sults of innovative activities 1.2 Regulatory environment
within the economy. Although 1.2.1 Regulatory quality*a ..................................................................1.19 .................0.03 ..............0.45 ..............0.63 .................0.28
1.2.2 Rule of law*a ...................................................................................1.1.9 .............0.22 ..............0.60 ..............0.64 .................0.18
the Output Sub-Index includes 1.2.3 Cost of redundancy dismissal, salary weeksb..........14.60 ..............17.95 ..............26.60 ..............16.18 ..............18.29
only two pillars, it has the 1.3 Business environment
same weight in calculating the 1.3.1 Ease of starting a business* ...............................................90.29 ..............84.76 ..............82.13 ..............79.87 ..............85.64
overall GII scores as the Input 1.3.2 Ease of resolving insolvency* ...........................................68.24 ..............51.63 ..............39.85 ..............38.80 ..............53.69
1.3.3 Ease of paying taxes*.............................................................83.83 ..............69.51 ..............59.52 ..............57.51 ..............71.19
Note: (*) index, () survey question, (a) half weight, (b) higher values indicate worse outcomes.
3. The overall GII score is the
simple average of the Input and
Output Sub-Indices. identified with an asterisk *, survey to innovation. The Institutions pillar
4. The Innovation Eff iciency questions with a dagger , and the captures the institutional framework
Ratio is the ratio of the Output remaining indicators are hard data); of a country (Table1a).
Sub-Index to the Input Sub- their weight in the index (indicators The Political environment sub-
Index. It shows how much inno- with half weight are identified with pillar includes two indices: one that
vation output a given country is the letter a); and the direction of ref lects perceptions of the likelihood
getting for its inputs. their effect (indicators for which that a government might be destabi-
higher values imply worse out- lized; and one that ref lects the qual-
Each pillar is divided into three comes are identified with the letter ity of public and civil services, policy
sub-pillars, each of which is com- b). The table then provides each formulation, and implementation.
posed of individual indicators, for indicators average values (in their The Regulatory environment
a total of 81 indicators this year. respective units) per income group sub-pillar draws on two indices
The GII pays special attention to (World Bank classification) and for aimed at capturing perceptions on
presenting a scoreboard for each the whole sample of 127 countries/ the ability of the government to
economy that includes strengths and economies retained in the f inal formulate and implement cohesive
weaknesses (Appendix I Country/ computation (Tables 1a through 1g). policies that promote the develop-
Economy Profiles), making accessi- ment of the private sector and at
ble the data series (Appendix II Data evaluating the extent to which the
Tables), and providing data sources The Innovation Input Sub-Index rule of law prevails (in aspects such
and definitions (Appendix III) and The first sub-index of the GII, the as contract enforcement, property
detailed technical notes (Appendix Innovation Input Sub-Index, has five rights, the police, and the courts).
IV). Adjustments to the GII frame- enabler pillars: Institutions, Human The third indicator evaluates the
work, including a detailed analysis capital and research, Infrastructure, cost of redundancy dismissal as the
of the factors inf luencing year-on- Market sophistication, and Business sum, in salary weeks, of the cost of
year changes, are detailed in Annex sophistication. Enabler pillars advance notice requirements added
2. In addition, since 2011 the GII def ine aspects of the environment to severance payments due when
has been submitted to an indepen- conducive to innovation within an terminating a redundant worker.
dent statistical audit performed by economy. The Business environment sub-
the Joint Research Centre of the pillar expands on three aspects that

European Union (results are detailed Pillar 1: Institutions directly affect private entrepreneur-
in Annex 3). Nurturing an institutional frame- ial endeavours by using the World
A table is included here for each work that attracts business and Bank indices on the ease of start-
pillar. That table provides a list of fosters growth by providing good ing a business; the ease of resolving
the pillars indicators, specifying governance and the correct levels of insolvency (based on the recovery
their type (composite indicators are protection and incentives is essential rate recorded as the cents on the

Table 1b: Human capital & research pillar mathematics, and science, as well as
Annex 1: The GII Conceptual Framework

Average value by income group

the pupil-teacher ratio.
High Upper-middle Lower-middle Low
Indicator income income income income Mean Higher education is crucial for
2 Human capital and research economies to move up the value
2.1 Education chain beyond simple production
2.1.1 Expenditure on education, % GDP ..................................5.49 .................4.56 .................4.21 .................4.75 .................4.75 processes and products. The sub-
2.1.2 Govt expend. on edu./pupil, secondary...................24.86 ..............17.65 ..............17.97 ..............25.17 ..............21.17
pillar on tertiary education aims at
2.1.3 School life expectancy, years............................................16.56 ..............14.31 ..............11.86 .................9.67 ..............13.95
2.1.4 PISA scales in reading, maths & sciencea................489.53 ...........416.63 ...........405.24 ............n/a.................. 459.98 capturing coverage (tertiary enrol-
2.1.5 Pupil-teacher ratio, secondarya,b.....................................11.25 ..............15.06 ..............20.07 ..............27.26 ..............16.52 ment); priority is given to the sectors
2.2 Tertiary education traditionally associated with innova-
2.2.1 Tertiary enrolment, % grossa .............................................66.29 ..............47.38 ..............28.27 .................7.28 ..............44.83
tion (with a series on the percent-
2.2.2 Graduates in science & engineering, % .....................22.76 ..............21.04 ..............22.06 ..............14.44 ..............21.32
2.2.3 Tertiary inbound mobility, %a ..............................................9.96 .................3.45 .................1.53 .................3.4.2 ................5.77 age of tertiary graduates in science,
2.3 Research and development (R&D) engineering, manufacturing, and
2.3.1 Researchers, FTE/mn pop ............................................3,680.04 ...........792.86 ...........449.14 ..............68.47 ....... 1,938.71 construction); and the inbound and
2.3.2 Gross expenditure on R&D, % GDP ..................................1.65 .................0.55 .................0.34 .................0.36 .................0.96
mobility of tertiary students, which
2.3.3 Global R&D firms, avg. exp. top 3, mn $US........1,332.33 ...........154.67 ..............37.95 .................0.00 ........... 554.25
2.3.4 QS university ranking, average score top 3* ...........39.97 ..............18.48 .................6.93 .................0.18 ..............21.70 plays a crucial role in the exchange
Note: (*) index, () survey question, (a) half weight, (b) higher values indicate worse outcomes. FTE = full-time equivalence. of ideas and skills necessary for
1 Scaled by percent of GDP per capita. innovation.
The last sub-pillar, on R&D,
measures the level and quality of
Table 1c: Infrastructure pillar R&D activities, with indicators on
Average value by income group researchers (full-time equivalence),
High Upper-middle Lower-middle Low
Indicator income income income income Mean
gross expenditure, the R&D expen-
3 Infrastructure
ditures of top global R&D spend-
3.1 Information and communication technologies (ICTs) ers, and the quality of scientific and
3.1.1 ICT access* ......................................................................................8.08 .................5.98 .................4.41 .................2.68 .................6.01 research institutions as measured
3.1.2 ICT use* .............................................................................................6.86 .................4.36 .................2.30 .................0.86 .................4.41 by the average score of the top
3.1.3 Governments online service* .............................................0.77 .................0.57 .................0.46 .................0.28 .................0.58
3.1.4 E-participation* ............................................................................0.75 .................0.57 .................0.49 .................0.30 .................0.59
three universities in the QS World
3.2 General infrastructure University Ranking of 2016. The
3.2.1 Electricity output, kWh/capa ......................................9,396.97 ....... 3,285.84 ....... 1,135.44 ........... 221.18 ....... 5,031.15 R&D expenditures of the top three
3.2.2 Logistics performance*a .........................................................3.60 .................2.83 .................2.64 .................2.56 .................3.04 firms in a given country looks at the
3.2.3 Gross capital formation, % GDP ......................................21.81 ..............25.33 ..............22.27 ..............24.49 ..............23.22
average expenditure of these three
3.3 Ecological sustainability
3.3.1 GDP/unit of energy use, 2010 PPP$/kg oil eq........10.15 .................9.73 .................8.84 .................4.36 .................9.29 firms that are part of the top 2,500
3.3.2 Environmental performance* ..........................................82.18 ..............74.11 ..............65.77 ..............47.86 ..............72.08 R&D spenders worldwide. The QS
3.3.3 ISO 14001 environ. certificates/bn PPP$ GDPa .........4.45 .................2.73 .................0.56 .................0.23 .................2.60 university rankings indicator gives
Note: (*) index, () survey question, (a) half weight, (b) higher values indicate worse outcomes. KwH = kilowatt hours.
the average scores of the countrys
top three universities that belong to
the top 700 universities worldwide.
dollar recouped by creditors through achievements at the elementary
These indicators are not aimed at
reorganization, liquidation, or debt and secondary education levels.
assessing the average level of all insti-
enforcement/foreclosure proceed- Education expenditure and school
tutions within a particular economy.
ings); and the ease of paying taxes. life expectancy are good proxies for
coverage. Government expenditure
Pillar 3: Infrastructure
Pillar 2: Human capital and research per pupil, secondary gives a sense of
The third pillar includes three sub-
The level and standard of education the level of priority given to second-
pillars: Information and communi-

and research activity in a country are ary education by the state. The qual-
cation technologies (ICTs), General
prime determinants of the innova- ity of education is measured through
infrastructure, and Ecological sus-
tion capacity of a nation. This pillar the results to the OECD Programme
tainability (Table1c).
tries to gauge the human capital of for International Student Assessment
Good and ecologically friendly
countries (Table1b). (PISA), which examines 15-year-old
communication, transport, and
The f irst sub-pillar includes a students performances in reading,
energy infrastructures facilitate the
mix of indicators aimed at capturing

production and exchange of ideas, Table 1d: Market sophistication pillar

Annex 1: The GII Conceptual Framework

Average value by income group
services, and goods and feed into the
High Upper-middle Lower-middle Low
innovation system through increased Indicator income income income income Mean
productivity and eff iciency, lower 4 Market sophistication
transaction costs, better access to 4.1 Credit
markets, and sustainable growth. 4.1.1 Ease of getting credit* ..........................................................59.79 ..............60.29 ..............55.74 ..............36.76 ..............55.98
4.1.2 Domestic credit to private sector, % GDP ................99.09 ..............59.83 ..............41.61 ..............23.82 ..............66.31
The ICTs sub-pillar includes four
4.1.3 Microfinance gross loans, % GDP .....................................0.15 .................0.95 .................3.63 .................0.98 .................1.79
indices developed by international 4.2 Investment
organizations on ICT access, ICT 4.2.1 Ease of protecting minority investors* .......................62.98 ..............58.86 ..............53.33 ..............43.63 ..............57.20
use, online service by governments, 4.2.2 Market capitalization, % GDPa..........................................93.18 ..............41.80 ..............28.10 ..............21.82 ..............60.25
4.2.3 Venture capital deals/bn PPP$ GDPa ..............................0.11 .................0.02 .................0.02 .................0.03 .................0.06
and online participation of citizens.
4.3 Trade, competition, and market scale
The sub-pillar on general infra- 4.3.1 Applied tariff rate, weighted mean, %a,b.......................1.84 .................3.79 .................5.35 .................8.99 .................4.08
structure includes the average of 4.3.2 Intensity of local competition ..........................................5.42 .................5.01 .................4.88 .................4.67 .................5.10
electricity output in kWh per capita; 4.3.3 Domestic market scale, bn PPP$.............................1,120.76 ....... 1,183.87 ...........700.32 ..............48.06 ........... 905.18

a composite indicator on logistics Note: (*) index, () survey question, (a) half weight, (b) higher values indicate worse outcomes.

performance; and gross capital for-

mation, which consists of outlays on
additions to the fixed assets and net The Credit sub-pillar includes a unsuccessful. Domestic market
inventories of the economy, includ- measure on the ease of getting credit scale, as measured by an economys
ing land improvements (fences, aimed at measuring the degree to GDP, was incorporated in 2016, so
ditches, drains); plant, machinery, which collateral and bankruptcy the last sub-pillar takes into consid-
and equipment purchases; and the laws facilitate lending by protecting eration the impact that the size of
construction of roads, railways, and the rights of borrowers and lenders, an economy has on its capacity to
the like, including schools, offices, as well as the rules and practices introduce and test innovations in the
hospitals, private residential dwell- affecting the coverage, scope, and market place.
ings, and commercial and industrial accessibility of credit information.
buildings. Transactions are given by the total Pillar 5: Business sophistication
The sub-pillar on ecological value of domestic credit and, in an The last enabler pillar tries to cap-
sustainability includes three indi- attempt to make the model more ture the level of business sophistica-
cators: GDP per unit of energy applicable to emerging markets, by tion to assess how conducive firms
use (a measure of efficiency in the the gross loan portfolio of microfi- are to innovation activity (Table1e).
use of energy), the Environmental nance institutions. The Human capital and research
Performance Index of Yale and The Investment sub-pillar pillar (pillar 2) made the case that
Columbia Universities, and the includes the ease of protecting the accumulation of human capi-
number of certif icates of confor- minority investors index as well as tal through education, particularly
mity with standard ISO 14001 on two indicators on the level of trans- higher education and the prioritiza-
environmental management systems actions. These two indicators look tion of R&D activities, is an indis-
issued. at whether market size is matched pensable condition for innovation to
by market dynamism and provide a take place. That logic is taken one
Pillar 4: Market sophistication hard data metric on venture capital step further here with the assertion
The availability of credit and an deals. that businesses foster their produc-
environment that supports invest- The last sub-pillar tackles trade, tivity, competitiveness, and innova-
ment, access to the international competition, and market scale. The tion potential with the employment
market, competition, and market market conditions for trade are given of highly qualified professionals and
scale are all critical for businesses to

in the f irst indicator measuring technicians.

prosper and for innovation to occur. the average tariff rate weighted by The f irst sub-pillar includes
The Market sophistication pillar has import shares. The second indica- four quantitative indicators on
three sub-pillars structured around tor is a survey question that ref lects knowledge workers: employment
market conditions and the total level the intensity of competition in local in knowledge-intensive services;
of transactions (Table1d). markets. Efforts made at finding hard the availability of formal training at
data on competition so far remain the firm level; R&D performed by

Table 1e: Business sophistication pillar reveal how good economies are at
Annex 1: The GII Conceptual Framework

Average value by income group

absorbing and diffusing knowledge.
High Upper-middle Lower-middle Low
Indicator income income income income Mean Sub-pillar 5.3 includes f ive
5 Business sophistication metrics that are linked to sectors
5.1 Knowledge workers with high-tech content or are key
5.1.1 Knowledge-intensive employment, %.......................38.87 ..............23.03 ..............17.99 .................3.73 ..............27.37 to innovation: intellectual property
5.1.2 Firms offering formal training, % firms .......................40.37 ..............38.43 ..............32.05 ..............28.41 ..............35.00
payments as a percentage of total
5.1.3 GERD performed by business, % GDPa..........................1.06 .................0.28 .................0.10 .................0.04 .................0.63
5.1.4 GERD financed by business, %a.......................................43.84 ..............25.65 ..............15.82 .................5.87 ..............31.32 trade; high-tech net imports as a
5.1.5 Females emp. w/adv. degrees, % tot. emp.a ...........18.81 ..............13.01 ..............10.02 .................2.27 ..............14.54 percentage of total imports; imports
5.2 Innovation linkages of communication, computer and
5.2.1 University/industry research collaborationa ...........4.26 .................3.40 .................3.21 .................3.13 .................3.66
information services as a percentage
5.2.2 State of cluster development ...........................................4.37 .................3.64 .................3.48 .................3.29 .................3.85
5.2.3 GERD financed by abroad, % ............................................14.14 .................9.09 .................8.98 ..............30.63 ..............13.49 of total trade; and net inf lows of
5.2.4 JV-strategic alliance deals/bn PPP$ GDPa ....................0.07 .................0.02 .................0.02 .................0.02 .................0.04 foreign direct investment (FDI) as a
5.2.5 Patent families filed in 2+ offices/bn PPP$ GDPa ....3.38 .................0.16 .................0.09 .................0.07 .................1.44
percentage of GDP (three-year aver-
5.3 Knowledge absorption
age). To strengthen the sub-pillar,
5.3.1 Intellectual property payments, % total tradea ........1.90 .................0.69 .................0.44 .................0.13 .................1.00
5.3.2 High-tech imports less re-imports, % total trade 10.27 .................9.81 .................7.98 .................7.91 .................9.36 the percentage of research talent in
5.3.3 ICT services imports, % total trade...................................1.67 .................0.93 .................0.86 .................1.71 .................1.30 business was added in 2016 to pro-
5.3.4 FDI net inflows, % GDP ............................................................5.32 .................3.94 .................3.18 .................5.32 .................4.49
vide a measurement of professionals
5.3.5 Research talent, % in business enterprise ................42.75 ..............23.24 ..............20.23 ..............17.04 ..............32.44
engaged in the conception or cre-
Note: (*) index, () survey question, (a) half weight, (b) higher values indicate worse outcomes. GERD = gross domestic expenditure on R&D.
ation of new knowledge, products,
processes, methods and systems,
including business management.
business enterprise (GERD) as a per- abroad, and the number of deals
centage of GDP (i.e., GERD over on joint ventures and strategic alli-
GDP); and the percentage of total ances. In addition, the total number The Innovation Output Sub-Index
gross expenditure of R&D that is of Patent Cooperation Treaty (PCT) Innovation outputs are the results
financed by business enterprise. In and national office published patent of innovative activities within the
addition, the sub-pillar includes an family applications filed by residents economy. Although the Output
indicator related to the percentage in at least two off ices proxies for Sub-Index includes only two pillars,
of females employed with advanced international linkages. it has the same weight in calculating
degrees. This indicator, in addition In broad terms, pillar 4 on mar- the overall GII scores as the Input
to providing a glimpse into the gen- ket sophistication makes the case Sub-Index. There are two output
der labour distributions of nations, that well-functioning markets con- pillars: Knowledge and technology
offers more information about the tribute to the innovation environ- outputs and Creative outputs.
degree of sophistication of the local ment through competitive pressure,
human capital currently employed. efficiency gains, and economies of Pillar 6: Knowledge and technology outputs
Innovation linkages and public/ transaction and by allowing supply This pillar covers all those vari-
private/academic partnerships are to meet demand. Markets that are ables that are traditionally thought
essential to innovation. In emerg- open to foreign trade and investment to be the fruits of inventions and/
ing markets, pockets of wealth have have the additional effect of expos- or innovations (Table1f ). The first
developed around industrial or tech- ing domestic firms to best practices sub-pillar refers to the creation of
nological clusters and networks, in around the globe, which is critical knowledge. It includes five indica-
sharp contrast to the poverty that to innovation through knowledge tors that are the result of inventive
may prevail in the rest of the terri- absorption and diffusion, which are and innovative activities: patent
applications filed by residents both

tory. The Innovation linkages sub- considered in pillars 5 and 6. The

pillar draws on both qualitative and rationale behind sub-pillars 5.3 on at the national patent office and at
quantitative data regarding business/ knowledge absorption (an enabler) the international level through the
university collaboration on R&D, and 6.3 on knowledge diffusion (a PCT; utility model applications filed
the prevalence of well-developed result)two sub-pillars designed to by residents at the national off ice;
and deep clusters, the level of gross be mirror images of each otheris scientif ic and technical published
R&D expenditure f inanced by precisely that together they will articles in peer-reviewed journals;

and an economys number of articles Table 1f: Knowledge & technology outputs pillar

Annex 1: The GII Conceptual Framework

Average value by income group
(H) that have received at least H
High Upper-middle Lower-middle Low
citations. Indicator income income income income Mean
The second sub-pillar, on knowl- 6 Knowledge and technology outputs
edge impact, includes statistics rep- 6.1 Knowledge creation
resenting the impact of innovation 6.1.1 Patents by origin/bn PPP$ GDPa .......................................7.65 .................3.02 .................1.27 .................0.25 .................4.10
6.1.2 PCT patent applications/bn PPP$ GDPa .......................2.50 .................0.23 .................0.10 .................0.06 .................1.17
activities at the micro- and macro-
6.1.3 Utility models by origin/bn PPP$ GDP ..........................1.26 .................3.23 .................3.19 .................0.19 .................2.40
economic level or related proxies: 6.1.4 Scientific & technical articles/bn PPP$ GDPa ..........30.01 ..............10.82 .................7.22 .................8.66 ..............16.94
increases in labour productivity, 6.1.5 Citable documents H index*a........................................422.21 ...........166.28 ...........120.37 ..............78.91 ........... 241.56
the entry density of new f irms, 6.2 Knowledge impact
6.2.1 Growth rate of PPP$ GDP/worker, % ..............................0.70 .................0.69 .................1.19 .................2.32 .................0.97
spending on computer software, the
6.2.2 New businesses/th pop. 1564a ........................................6.12 .................3.28 .................1.00 .................0.45 .................3.64
number of certificates of conformity 6.2.3 Computer software spending, % GDPa .........................0.42 .................0.21 .................0.19 .................0.07 .................0.26
with standard ISO 9001 on quality 6.2.4 ISO 9001 quality certificates/bn PPP$ GDPa............14.69 .................9.35 .................2.73 .................1.33 .................8.89
6.2.5 High- & medium-high-tech manufactures, %a......33.74 ..............21.97 ..............15.83 .................8.68 ..............25.05
management systems issued, and
6.3 Knowledge diffusion
the measure of high- and medium-
6.3.1 Intellectual property receipts, % total tradea .............1.20 .................0.08 .................0.11 .................0.05 .................0.51
high-tech industrial output over 6.3.2 High-tech exports less re-exports, % total tradea ...6.87 .................4.55 .................2.15 .................0.34 .................4.39
total manufactures output. 6.3.3 ICT services exports, % total tradea..................................2.99 .................1.73 .................2.34 .................2.34 .................2.42
6.3.4 FDI net outflows, % GDP ........................................................3.59 .................0.95 .................0.22 .................0.52 .................1.75
The third sub-pillar, on knowl-
Note: (*) index, () survey question, (a) half weight, (b) higher values indicate worse outcomes.
edge diffusion, is the mirror image
of the knowledge absorption sub-
pillar of pillar 5, with the exception
of indicator 5.3.5. It includes four Table 1g: Creative outputs pillar
Average value by income group
statistics all linked to sectors with High Upper-middle Lower-middle Low
high-tech content or that are key Indicator income income income income Mean

to innovation: intellectual property 7 Creative outputs

receipts as a percentage of total 7.1 Intangible assets
7.1.1 Trademarks by origin/bn PPP$ GDP.............................56.96 ..............56.80 ..............45.97 ..............16.72 ..............49.60
trade; high-tech net exports as a
7.1.2 Industrial designs by origin/bn PPP$ GDPa ................5.3.5 ................3.09 .................4.48 .................1.26 .................4.10
percentage of total exports; exports 7.1.3 ICTs & business model creation ......................................5.28 .................4.51 .................4.25 .................3.89 .................4.68
of ICT services as a percentage of 7.1.4 ICTs & organizational model creation .........................4.93 .................4.04 .................3.87 .................3.40 .................4.28
total trade; and net outf lows of FDI 7.2 Creative goods and services
7.2.1 Cultural & creative services exp., % total tradea .......0.85 .................0.58 .................0.08 .................0.23 .................0.54
as a percentage of GDP (three-year
7.2.2 National feature films/mn pop. 1569a .........................9.35 .................3.30 .................2.90 .................1.30 .................5.52
average). 7.2.3 Global ent. & media market/th pop. 1569a ..............1.26 .................0.19 .................0.05 ............n/a........................0.78
7.2.4 Printing & publishing manufactures, % ........................2.21 .................1.62 .................1.12 .................1.55 .................1.78
7.2.5 Creative goods exports, % total trade............................1.90 .................1.70 .................0.86 .................0.07 .................1.39
Pillar 7: Creative outputs
7.3 Online creativity
The role of creativity for innovation
7.3.1 Generic TLDs/th pop. 1569 .............................................33.42 .................5.79 .................1.37 .................0.32 ..............14.56
is still largely underappreciated in 7.3.2 Country-code TLDs/th pop. 1569 ...............................31.69 .................6.52 .................0.91 .................0.77 ..............14.07
innovation measurement and policy 7.3.3 Wikipedia yearly edits/mn pop. 1569.......................60.37 ..............46.09 ..............33.69 .................9.93 ..............44.01
7.3.4 Video uploads on YouTube/pop. 1569 ....................48.20 ..............25.32 ..............11.43 .................0.94 ..............35.41
debates. Since its inception, the GII
Note: (*) index, () survey question, (a) half weight, (b) higher values indicate worse outcomes. Scores rather than values are presented for indicators 7.3.1, 7.3.2,
has always emphasized measuring 7.3.3, and 7.3.4. TLDs = top-level domains.
creativity as part of its Innovation
Output Sub-Index. The last pillar,
on creative outputs, has three sub- models, new areas that are increas- composite was added. In addition,
pillars (Table1g). ingly linked to process innovations the indicator on audio-visual and
The f irst sub-pillar on intan- in the literature. related services exports was renamed
gible assets includes statistics on

The second sub-pillar on cre- Cultural and creative services

trademark applications by residents ative goods and services includes exports and expanded to include
at the national off ice; industrial proxies to get at creativity and the information services, advertising,
designs included in applications at a creative outputs of an economy. market research and public opinion
regional or national office, and two In 2014, in an attempt to include polling, and other personal, cultural,
survey questions regarding the use of broader sectoral coverage, a global and recreational services (as a per-
ICTs in business and organizational entertainment and media output centage of total trade). These two

indicators complement the remain- INSEAD. 2011. The Global Innovation Index 2011:
Annex 1: The GII Conceptual Framework

Accelerating Growth and Development, ed. S.

der of the sub-pillar, which measures Dutta. Fontainebleau: INSEAD.
national feature films produced in
OECD (Organisation for Economic Co-operation and
a given country (per capita count); Development). 2010. The OECD Innovation
printing and publishing output (as Strategy: Getting a Head Start on Tomorrow.
Paris: OECD.
a percentage of total manufactures
. 2011. OECD Science, Technology and Industry
output); and creative goods exports Scoreboard 2011. Paris: OECD.
(as a percentage of total trade), all
. 2013. OECD Science, Technology and Industry
of which are aimed at providing an Scoreboard 2013. Paris: OECD.
overall sense of the international OECD and Eurostat (Organisation for Economic
reach of creative activities in the Co-operation and Development and
country. Eurostat). 2005. Oslo Manual: Guidelines for
Collecting and Interpreting Innovation Data,
The third sub-pillar on online 3rd Edition. Paris: OECD Publishing.
creativity includes four indicators, WIPO (World Intellectual Property Organization).
all scaled by population aged 15 2011. The Changing Nature of Innovation
and Intellectual Property. In World Intellectual
through 69 years old: generic and Property Report 2011: The Changing Face
country-code top level domains, of Innovation, Chapter 1. Geneva: WIPO.
average yearly edits to Wikipedia; Available at
and video uploads on YouTube.
Attempts made to strengthen this
sub-pillar with indicators in areas
such as Internet and machine learn-
ing, blog posting, online gaming,
and the development of applications
have so far proved unsuccessful.

1 For a fuller introduction to the Global
Innovation Index, see the GII 2011.

2 OECD and Eurostat, 2005.

3 OECD, 2010; INSEAD, 2011; and WIPO, 2011.

4 INSEAD, 2011; OECD Scoreboard, 2013; WIPO,


5 INSEAD, 2011; OECD, 2011; WIPO, 2011.

6 For completeness, 2.0% of data points are

from 2012, 1.2% from 2011, 1.3% from 2010,
0.7% from 2009, 0.7% from 2008, 0.3% from
2007, and 0.1% from 2006. In addition, the
GII is calculated on the basis of 9,225 data
points (compared to 10,287 with complete
series), implying that 10.3% of data points
are missing. The Data Tables (Appendix II)
include the reference year for each data point
and mark missing data as not available (n/a).

Cornell University, INSEAD, and WIPO (World
Intellectual Property Organization). 2013.
The Global Innovation Index 2013: The Local
Dynamics of Innovation, eds. S. Dutta and B.
Lanvin. Geneva, Ithaca, and Fontainebleau:
Cornell, INSEAD, and WIPO.
ANNEX 2 55

Adjustments to the Global Innovation Index Framework

Annex 2: Adjustments and Year-on-Year Comparabilty

and Year-on-Year Comparability of Results

The Global Innovation Index (GII) is Table 1: Changes to the Global Innovation Index framework
a cross-country performance assess-
ment, compiled on an annual basis, GII 2016 Adjustment GII 2017
which continuously seeks to update 4.2.3 Total value of stocks traded,
and improve the way innovation is
measured. The GII report pays special 4.2.4 Venture capital deals/bn
Number changed
4.2.3 Venture capital deals/bn
attention to making accessible the sta-
tistics used in the Country/Economy 5.3.4 Foreign direct investment net Name and 5.3.4 Foreign direct investment net
inflows methodology changed inflows (3-year avg.)
Profiles and Data Tables, providing
6.3.4 Foreign direct investment net Name and 6.3.4 Foreign direct investment net
data sources and definitions, and outflows methodology changed outflows (3-year avg.)
detailing the computation method-
7.3.3 Wikipedia monthly edits Name and 7.3.3 Wikipedia yearly edits
ology (Appendices I, II, III, and IV, methodology changed
respectively). This annex summarizes Note: Refer to Annex 1 and Appendix III for a detailed explanation of terminologies. Indicators whose name did not change but methodology at the source did
the changes made this year and pro- are not part of this list. Refer to Appendix III for a detailed explanation of methodological changes at the source.

vides an assessment of the impact of

these changes on the comparability
of rankings.

Adjustments to the Global Innovation

Index framework
The GII model is revised every year the International Organization for its methodology are not identified in
in a transparent exercise. This year, no Standardization (ISO); IHS Global Table1.
change was made at either the pillar Insight; QS Quacquarelli Symonds The statistical audit performed
or the sub-pillar level. Ltd; Bureau van Dijk (BvD); by the JRC (see Annex 3) pro-
Beyond the use of World ZookNIC Inc; and Google to obtain vides a confidence interval for each
Intellectual Property Organization the best available data on innovation ranking following a robustness and
(WIPO) data, we collaborate with measurement globally. uncertainty analysis of the modelling
both public international bodies such Table1 provides a summary of assumptions.
as the International Energy Agency; adjustments to the GII 2017 frame-
the United Nations Educational, work for quick reference. A total of
Scientific and Cultural Organization five indicators were modified this Sources of changes in the rankings
(UNESCO); the United Nations year: one indicator was removed, The GII compares the performance

Industrial Development Organization one indicator changed its number as of national innovation systems across
(UNIDO); the International a result, and three indicators under- economies, and it also presents
Telecommunication Union (ITU); went methodological and name changes in economy rankings over
and the Joint Research Centre of the changes. Indicators that retained time.
European Commission (JRC) as well the same name as last year but are Importantly, scores and rankings
as with private organizations such as derived from a source that changed from one year to the next are not

directly comparable (see Annex 2 of is that it widens the set of data direct investment are now being mea-
Annex 2: Adjustments and Year-on-Year Comparabilty

the GII 2013 for a full explanation). points for cross-economy com- sured as an average of the most recent
Making inferences about absolute or parability. three years to produce a more stable
relative performance on the basis of reflection of these indicators datasets.
Normalization factor. Most
year-on-year differences in rankings The underlying methodology for
GII variables are normalized
can be misleading. Each ranking indicator 7.3.3 has also changed; it
using either GDP or population.
ref lects the relative positioning of that now measures edits within each econ-
This approach is also intended
particular country/economy on the omy by year rather than by month.
to enable cross-economy com-
basis of the conceptual framework,
parability. Yet, again, year-on-
the data coverage, and the sample of
year changes in individual vari-
economieselements that change Missing values
ables may be driven either by
from one year to another. Since its inception, the GII has had
the variables numerator or by its
A few particular factors inf lu- a positive inf luence on data avail-
ence the year-on-year ranking of a ability, increasing awareness of the
country/economy: Consistent data collection. importance of submitting timely data.
Finally, measuring year-on-year The number of data points submitted
the actual performance of the
performance changes relies on by economies to international data
economy in question;
the consistent collection of data agencies has substantially increased in
adjustments made to the GII over time. Changes in the defi- recent years. In the GII 2016, 12.8%
framework; nition of variables or in the data of data points were missing; this year,
collection process could create in the GII 2017, coverage improved
data updates, the treatment of
movements in the rankings that again, with only 10.3% of data points
outliers, and missing values; and
are unrelated to true perfor- missing.
the inclusion or exclusion of mance. When it comes to country cover-
cou nt r ies/econom ies i n t he age, the objective is to include as many
sample. A detailed economy study based economies as possible. However, it is
on the GII database and the country/ also important to maintain a good
Additionally, the following char- economy profile over time, coupled level of data coverage within each
acteristics complicate the time-series with analytical work on grounds that of these economies. Because the GII
analysis based on simple GII scores or include innovation actors and deci- results are linked to data availability
rankings: sion makers, yields the best results in (see the JRC Statistical Audit pre-
terms of grasping an economys inno- sented in Annex 3 for more details),
Missing values. The GII pro-
vation performance over time as well which affects the overall GII ranks,
duces relat ive i ndex scores,
as possible avenues for improvement. this year the minimum data coverage
which means that a missing value
threshold rule was strengthenedon
for one economy af fects the
the recommendation of the JRCto
index score of other economies.
Methodology and data maintain the significance of both the
Because the number of missing
The revision of the computation GII results and the country sample.
values decreases every year, this
methodology for certain individual To be included in the GII 2017, an
problem is reduced over time.
indicators has caused shifts in the economy must have a minimum
R e f e r e n c e ye a r. T he d at a results for several countries. symmetric data coverage of 36 indi-
underlying the GII do not refer For indicator 3.3.1, which mea- cators in the Innovation Input Sub-
to a single year but to several sures energy use, the constant PPP$ Index (66%) and 18 indicators in the
years, depending on the latest per kg of oil equivalent was updated Innovation Output Sub-Index (66%),
available year for any given vari- from 2005 PPP$ to 2010 PPP$. and it must have scores for at least two

able. In addition, the reference The methodology underpinning sub-pillars per pillar. Missing values
years for different variables are indicators 4.2.3 and 5.2.4 expanded are indicated with n/a and are not
not the same for each economy. to use datasets from previous years to considered in the sub-pillar score.
The motivation for this approach improve data coverage. This adjustment derives from a
For indicators 5.3.4 and 6.3.4, the sensitivity that is the result of the data
net inf lows and outf lows of foreign availability, which is less satisfactory

Table 2: GII economies with the most missing values in the case of the Output Sub-Index:

Annex 2: Adjustments and Year-on-Year Comparabilty

four countries that were part of the
Economy Number of missing values Economy Number of missing values
GII 2016 have data coverage below
Trinidad and Tobago 25 Brunei Darussalam 21 the 66% threshold in the 27 variables
Togo 23 Burkina Faso 20 in the Output Sub-Index. In contrast,
Burundi 22 Guinea 20 data coverage is satisfactory in all of
Niger 22 Nepal 20 these cases in the Input Sub-Index
Benin 21 (all of these economies have indicator
coverage of more than 66% over the
54 input variables). As a result, the fol-
Table 3: GII economies with the fewest missing values lowing countries included in the GII
Economy Number of missing values Economy Number of missing values
2016 dropped out this year: Bhutan,
Ghana, Nicaragua, and the Bolivarian
Colombia 0 Israel 3
Republic of Venezuela.1 The rules on
Hungary 0 Kazakhstan 3
missing data and the minimum cover-
Mexico 0 Netherlands 3
age necessary per sub-pillar will be
Romania 0 Serbia 3
progressively tightened, leading to
Bulgaria 1 Slovenia 3
the exclusion of countries that fail to
Chile 1 Spain 3
meet the desired minimum coverage
Czech Republic 1 Sweden 3
in any sub-pillar (see Appendix I for
Malaysia 1 Argentina 4
more details).
Poland 1 Croatia 4
Despite requiring minimum lev-
Russian Federation 1 Egypt 4
els of coverage, for several economies
Turkey 1 Latvia 4
the number of missing data points
Austria 2 Lithuania 4
remains very high. Table2 lists the
Brazil 2 Malta 4
countries that have the highest num-
France 2 Morocco 4
ber of missing data points (20 or more),
Italy 2 New Zealand 4
ranking them according to how many
Japan 2 Norway 4
data points are missing.
Korea, Rep. 2 Philippines 4
Conversely, Table 3 lists those
Portugal 2 Switzerland 4
economies with the best data cover-
Slovakia 2 Tunisia 4 age, ranking them according to the
South Africa 2 United Kingdom 4 least number of missed data points.
Thailand 2 Cyprus 5 These economies are missing at most
Ukraine 2 Georgia 5 only five data points; some are miss-
Australia 3 Greece 5 ing none at all.
Belgium 3 India 5
Costa Rica 3 Ireland 5
Denmark 3 Luxembourg 5 Note
Estonia 3 Moldova, Rep. 5 Conversely, Brunei Darussalam, Trinidad and
Tobago, and Zimbabwewhich were not
Finland 3 Panama 5
included in the GII 2016enter the GII this
Germany 3 Singapore 5 year with the required coverage in both
Indonesia 3 United States of America 5 sub-indices and sufficient data availability
per pillar.
ANNEX 3 59

Joint Research Centre Statistical Audit of the 2017 Global Innovation Index

Annex 3: JRC Statistical Audit of the GII

MICHAELA SAISANA, MARCOS DOMNGUEZ-TORREIRO , and DANIEL VERTESY, European Commission, Joint Research Centre (JRC), Ispra, Italy

Conceptual and practical challenges Indicators and Scoreboards at the at setting the foundation for a bal-
are inevitable when trying to under- Joint Research Centre ( JRC) in anced index. The entire process fol-
stand and model the fundamentals Ispra has been invited for the seventh lowed four steps (see Figure1).
of innovation at the national level consecutive year to audit the GII.
worldwide. In its 10th edition, As in previous editions, the present Step 1: Conceptual consistency
the 2017 Global Innovation Index JRC audit focuses on the statistical Eighty-one indicators were selected
(GII) considers these conceptual soundness of the multi-level structure for their relevance to a specific
challenges in Chapter 1 and deals of the index as well as on the impact innovation pillar on the basis of the
with practical challengesrelated of key modelling assumptions on the literature review, expert opinion,
to data quality and methodological results.2 The independent statistical country coverage, and timeliness. To
choicesby grouping country-level assessment of the GII provided by represent a fair picture of country dif-
data over 127 countries and across the JRC guarantees the transparency ferences, indicators were scaled either
81 indicators into 21 sub-pillars, 7 and reliability of the index for both at the source or by the GII team as
pillars, 2 sub-indices and, finally, policy makers and other stakeholders, appropriate and where needed.
an overall index. This annex offers thus facilitating more accurate prior-
detailed insights into the practical ity setting and policy formulation in Step 2: Data checks
issues related to the construction of this particular field. The most recently released data
the GII, analysing in depth the sta- As in past GII reports, the JRC within the period 200616 were
tistical soundness of the calculations analysis complements the country used for each economy: 77% of the
and assumptions made to arrive at rankings with confidence intervals available data refer to 2015 or more
the final index rankings. Statistical for the GII, the Innovation Input Sub- recent years. In past editions, coun-
soundness should be regarded as a Index, and the Innovation Output tries were included if data availability
necessary but not sufficient condi- Sub-Index in order to better appre- was at least 60% across all variables
tion for a sound GII, since the cor- ciate the robustness of these ranks to in the GII framework. A more strin-
relations underpinning the majority the computation methodology. In gent criterion was adopted this year,
of the statistical analyses carried out addition, the JRC analysis includes following the JRC recommendation
herein need not necessarily represent an assessment of the added value of of past GII audits. That is, countries
the real inf luence of the individual the GII and a measure of distance to were included if data availability was
indicators on the phenomenon being the efficient frontier of innovation by at least 66% within each of the two
measured.1 Consequently, the devel- using data envelopment analysis. sub-indices (i.e., 36 out of 54 vari-
opment of the GII must be nurtured ables within the Input Sub-Index
by a dynamic iterative dialogue and 18 out of the 27 variables in the
between the principles of statistical Conceptual and statistical coherence in Output Sub-Index) and at least two

and conceptual soundness or, to put the GII framework of the three sub-pillars in each pillar
it another way, between the theoreti- An earlier version of the GII model could be computed. This more strin-
cal understanding of innovation and was assessed by the JRC in April gent criterion for a countrys inclu-
the empirical observations of the data May 2017. Fine-tuning suggestions sion in the GII was introduced this
underlying the variables. were taken into account in the final year in order to ensure that country
The European Commissions computation of the rankings in an scores for the GII and for the two
Competence Centre on Composite iterative process with the JRC aimed Input and Output Sub-Indices are

Figure 1: Conceptual and statistical coherence in the GII 2017 framework indicators were treated either by win-
Annex 3: JRC Statistical Audit of the GII

sorization or by taking the natural

logarithm (in case of more than five
Step 4. Qualitative review outliers; see Appendix IV Technical
Internal qualitative review (INSEAD, WIPO, Cornell University) Notes in this report for details).
External qualitative review (JRC, international experts)

Step 3: Statistical Coherence

Weights as scaling coefficients
Weights of 0.5 or 1.0 were jointly
decided between the JRC and the GII
Step 3. Statistical coherence team in 2012 to be scaling coefficients
Treatment of highly collinear variables as a single indicator
and not importance coefficients, with
Assessment of grouping indicators into sub-pillars, pillars,
sub-indices, and the GII
the aim of arriving at sub-pillar and
Use of weights as scaling coefficients to ensure statistical coherence pillar scores that were balanced in
Assessment of arithmetic average assumption their underlying components (i.e.,
Assessment of potential redundancy of information in the overall GII that indicators and sub-pillars can
explain a similar amount of variance
in their respective sub-pillars/pillars).
Becker et al. (2017) and Paruolo et
al. (2013) show that, in weighted
Step 2. Data checks arithmetic averages, the ratio of two
Check for data recency (77% of available data refer to 20152016) nominal weights gives the rate of
Availability requirements per country: coverage 66% for the Input and substitutability between two indica-
the Output Sub-Indices separately and at least two sub-pillars per pillar
tors, and hence can be used to reveal
Check for reporting errors (interquartile range)
Outlier treatment (skewness and kurtosis) the relative importance of individual
Direct contact with data providers indicators. This importance can then
be compared with ex-post measures
of variables importance, such as the
non-linear Pearson correlation ratio.
As a result of this analysis, 35 out of 81
Step 1. Conceptual consistency
indicators and two sub-pillars7.2
Compatibility with existing literature on innovation and pillar Creative goods and services and 7.3
definition Online creativitywere assigned
Scaling factors per indicator to represent a fair picture of country half weight while all other indica-
differences (e.g., GDP, population) tors and sub-pillars were assigned a
weight of 1.0. Nevertheless, for seven
Source: European Commission Joint Research Centre, 2017. indicators with Pearson correlation
coefficients less than 0.3 with the
respective sub-pillars, some further
not particularly sensitive to the miss- skewness and kurtosis. In past edi- ref lection is needed because they
ing values (as it was the case for the tions since 2011, values were treated seem to be non-inf luential (i.e., they
Output Sub-Index scores of several if the indicators had absolute skew- behave as noise) at all aggregation
countries in past editions). In prac- ness greater than 2.0 and kurtosis levels in the GII 2017 framework,
tice, data availability for all countries greater than 3.5.3 These criteria were despite the fact that their inclusion
was based on conceptual grounds or

included in the GII 2017 is very good: decided jointly with the JRC back
80% data availability for 84% (107 out in 2011. This year and after having practical experience. This applies to
of 127) of the countries. Potentially analysed data in GII 2011GII 2017, 2.1.2 Government expenditure on
problematic indicators that could bias a less stringent criterion was adopted: education per pupil, secondary; 2.2.2
the overall results were identified on an indicator was treated if the abso- Graduates in science and engineer-
the basis of two measures related lute skewness was greater than 2.25 ing; 3.2.3 Gross capital formation;
to the shape of the distributions: and kurtosis greater than 3.5. These 5.2.3 GERD financed by abroad,

Table 1: Statistical coherence in the GII: Correlations between sub-pillars and pillars

Annex 3: JRC Statistical Audit of the GII

Human capital Market Business and technology Creative
Sub-pillar Institutions and research Infrastructure sophistication sophistication outputs outputs

Political environment 0.94 0.76 0.85 0.69 0.71 0.68 0.77

Regulatory environment 0.93 0.68 0.71 0.60 0.68 0.60 0.67
Business environment 0.89 0.73 0.77 0.69 0.66 0.66 0.71
Education 0.57 0.78 0.56 0.45 0.50 0.51 0.55
Tertiary education 0.67 0.80 0.73 0.58 0.48 0.54 0.60
Research and development (R&D) 0.69 0.88 0.76 0.74 0.83 0.85 0.74
Information and communication technologies (ICTs) 0.80 0.85 0.94 0.75 0.68 0.72 0.82
INPUT General infrastructure 0.57 0.53 0.69 0.47 0.49 0.56 0.47
Ecological sustainability 0.65 0.59 0.77 0.54 0.55 0.53 0.67
Credit 0.63 0.58 0.58 0.87 0.53 0.56 0.60
Investment 0.53 0.47 0.42 0.71 0.52 0.44 0.42
Trade, competition, & market scale 0.48 0.66 0.73 0.71 0.52 0.62 0.62
Knowledge workers 0.69 0.79 0.72 0.67 0.86 0.72 0.67
Innovation linkages 0.52 0.42 0.40 0.38 0.74 0.51 0.45
Knowledge absorption 0.56 0.60 0.58 0.55 0.81 0.77 0.61
Knowledge creation 0.62 0.79 0.64 0.65 0.78 0.89 0.76
Knowledge impact 0.50 0.55 0.61 0.48 0.54 0.76 0.62
Knowledge diffusion 0.59 0.60 0.61 0.58 0.69 0.80 0.59
Intangible assets 0.61 0.63 0.70 0.59 0.55 0.67 0.91
Creative goods and services 0.69 0.66 0.69 0.61 0.68 0.71 0.85
Online creativity 0.82 0.80 0.82 0.71 0.75 0.79 0.88

Source: European Commission Joint Research Centre, 2017..

5.3.4 Foreign direct investment Principal components analysis and correlation coefficients are close to or
net inf lows; 6.2.1 Growth rate of reliability item analysis greater than 0.70. (see Table1).
GDP per person engaged; and 7.2.4 Principal component analysis (PCA) The five input pillars share a single
Printing and publishing output. For was used to assess to what extent statistical dimension that summarizes
two out of the seven indicators listed the conceptual framework is con- 80% of the total variance, and the five
above2.1.2 and 7.2.4this is the firmed by statistical approaches. loadings (correlation coefficients) of
first time that they are found to PCA results confirm the presence these pillars are very similar to each
be non-inf luential at all in the GII of a single latent dimension in each other (0.860.92). This similarity
framework. Instead, the remaining of the seven pillars (one component suggests that the five pillars make
five indicators were found to be non- with an eigenvalue greater than 1.0) roughly equal contributions to the
inf luential also in the GII 2016. On that captures between close to 60% variation of the Innovation Input
the other hand, two indicators that (pillar 4: Market sophistication) up to Sub-Index scores, as envisaged by
were found to be non-inf luential 85% (pillar 1: Institutions) of the total the developing team. The reliability
last year3.3.1 GDP per unit of variance in the three underlying sub- of the Input Sub-Index, measured
energy use and 4.1.3 Microfinance pillars. These results reveal that the by the Cronbach alpha value, is very
institutions gross loan portfolioare modest adjustments made to the 2017 high at 0.94well above the 0.70
instead found to be inf luential in this GII framework have left unaffected threshold for a reliable aggregate.4
years framework. It is suggested that the already good statistical coherence The two output pillars

the GII development team carefully properties of the previous version. Knowledge and technology outputs
assess how these variables behave in Furthermore, results confirm the and Creative outputsare strongly
the coming releases of the index. If expectation that the sub-pillars are correlated to each other (0.81); they
the noisy behaviour persists, these more correlated to their own pillar are also both strongly correlated with
variables could eventually be removed than to any other pillar and that all the Innovation Output Sub-Index
from the GII framework. (0.95). This result suggests that the

Table 2: Distribution of differences between pillar and GII rankings

Annex 3: JRC Statistical Audit of the GII

Innovation Input Sub-Index Innovation Output Sub-Index

Human capital Market Business Knowledge and

Rank differences (positions) Institutions (%) and research (%) Infrastructure (%) sophistication (%) sophistication (%) technology outputs (%) Creative outputs (%)

More than 30 14.8% 9.4% 3.9% 21.9% 17.2% 9.4% 3.1%

2029 15.6% 14.8% 14.1% 10.2% 12.5% 11.7% 8.6%
1019 23.4% 21.9% 28.1% 28.9% 18.8% 26.6% 30.5%

10 or more* 53.9% 46.1% 46.1% 60.9% 48.4% 47.7% 42.2%

59 21.1% 23.4% 25.8% 16.4% 22.7% 23.4% 19.5%

Less than 5 21.9% 26.6% 23.4% 18.8% 25.0% 25.8% 32.0%
Same rank 2.3% 3.1% 3.9% 3.1% 3.1% 2.3% 5.5%

Total 99.2% 99.2% 99.2% 99.2% 99.2% 99.2% 99.2%

Pearson correlation 0.88 0.90 0.89 0.81 0.86 0.92 0.93

coefficient with the GII

Source: European Commission Joint Research Centre, 2017.

* This column is the sum of the prior three rows.
This column is the sum of all white rows.

Output Sub-Index is also well bal- statistically coherent in the GII 2017 statistical reliability of the five input
anced in its two pillars. Furthermore, framework, and that the GII has a pillarsmay be interpreted by some
building the GII as the simple average balanced structure at e`ach aggrega- as a sign of redundancy of informa-
of the Input Sub-Index and Output tion level. tion in the GII. The tests conducted
Sub-Index is also statistically justifi- The only recommendation for by the JRC confirm that this is not the
able because the Pearson correlation next year relates to a careful ref lec- case. In fact, for more than 42% (up to
coefficient of either sub-index with tion of the seven indicators discussed 61%) of the 127 economies included
the overall GII is 0.97; the two sub- above2.1.2 Government expen- in the GII 2017, the GII ranking and
indices have a correlation of 0.89. diture on education per pupil, sec- any of the seven pillar rankings differ
Finally, an important part of ondary; 2.2.2 Graduates in science by 10 positions or more (see Table2).
the analysis relates to clarifying the and engineering; 3.2.3 Gross capital This is a desired outcome because it
importance of the Input and Output formation; 5.2.3 GERD financed by demonstrates the added value of the
Sub-Indices with respect to the abroad; 5.3.4 Foreign direct invest- GII ranking, which helps to highlight
variation of the GII scores. The GII ment net inf lows; 6.2.1 Growth other aspects of innovation that do
is built as the simple arithmetic aver- rate of GDP per person engaged; not emerge directly by looking into
age of the five Input sub-pillars and and 7.2.4 Printing and publishing the seven pillars separately. At the
the two Output sub-pillars, which outputbecause their information same time, this result points to the
implies that the Input-related pillars content is lost in the aggregation at value of duly taking into account the
have a weight of 5/7 versus a weight the pillar level or higher (sub-index GII pillars, sub-pillars, and individual
of 2/7 for the Output-related pil- and overall GII). For five out of the indicators on their own merit. By
lars. Yet this does not imply that the seven indicators (2.2.2, 3.2.3, 5.2.3, doing so, country-specific strengths
Input aspect is more important than 5.3.4, 6.2.1) this was also the case in and bottlenecks on innovation can be
the Output aspect in determining the last years audit. identified and serve as an input for
variation of the GII scores. In fact, evidence-based policy making.
the Pearson correlation coefficient of Added value of the GII
either sub-index with the overall GII As already discussed, the Input and Step 4: Qualitative Review

is 0.97 (and the two sub-indices have Output Sub-Indices correlate strongly Finally, the GII resultsinclud-
a correlation of 0.89), which suggests with each other and with the overall ing overall country classifications
that the sub-indices are effectively GII. Furthermore, the five pillars in and relative performances in terms
placed on equal footing. the Input Sub-Index have a very high of the Innovation Input or Output
Overall, the tests so far show that statistical reliability. These results Sub-Indiceswere evaluated to
the grouping of variables into sub- the strong correlation between Input verify that the overall results are, to a
pillars, pillars, and an overall index is and Output Sub-Indices and the high great extent, consistent with current

Table 3: Uncertainty parameters: Missing values, aggregation, and weights

Annex 3: JRC Statistical Audit of the GII

Reference Alternative

I. Uncertainty in the treatment of missing values No estimation of missing data Expectation Maximization (EM)

II. Uncertainty in the aggregation formula at pillar level Arithmetic average Geometric average

III. Uncertainty intervals for the GII pillar weights

GII Sub-Index Pillar Reference value for the weight Distribution assigned for robustness analysis
Innovation Input Institutions 0.2 U[0.1, 0.3]
Human capital and research 0.2 U[0.1, 0.3]
Infrastructure 0.2 U[0.1, 0.3]
Market sophistication 0.2 U[0.1, 0.3]
Business sophistication 0.2 U[0.1, 0.3]

Innovation Output Knowledge and technology outputs 0.5 U[0.4, 0.6]

Creative outputs 0.5 U[0.4, 0.6]

Source: European Commission Joint Research Centre, 2017.

evidence, existing research, and The rationale for these choices is three key modelling issues considered
prevailing theory. Notwithstanding manifold. For instance, expert opin- in the assessment of the GII were the
these statistical tests and the positive ion coupled with statistical analysis treatment of missing data, the pillar
outcomes on the statistical coherence is behind the selection of the indi- weights, and the aggregation formula
of the GII structure, the GII model vidual indicators, common practice used at the pillar level.
is and has to remain open for future and ease of interpretation suggests Monte Carlo simulation com-
improvements as better data, more the use of a min-max normaliza- prised 1,000 runs of different sets of
comprehensive surveys and assess- tion approach in the [0100] range, weights for the seven pillars in the
ments, and new relevant research the treatment of outliers is driven by GII. The weights were assigned to
studies become available. statistical analysis, and simplicity and the pillars based on uniform continu-
parsimony criteria seem to advocate ous distributions centred in the refer-
for not imputing missing data. The ence values. The ranges of simulated
The impact of modelling assumptions on unavoidable uncertainty stemming weights were defined by taking into
the GII results from the above-mentioned model- account both the need for a wide
Modelling assumptions with a direct ling choices is accounted for in the enough interval to allow for mean-
impact on the GII scores and rank- robustness assessment carried out by ingful robustness checks and the need
ings relate to: the JRC. More precisely, the meth- to respect the underlying principle of
odology applied herein allows for the GII that the Input and the Output
setting up an underlying struc-
the joint and simultaneous analysis Sub-Indices should be placed on equal
ture for the index based on a
of the impact of such choices on the footings. As a result of these consider-
battery of pillars,
aggregate scores, resulting in error ations, the limit values of uncertainty
choosing the individual variables
estimates and confidence intervals for the five input pillars are 10%30%;
to be used as indicators,
calculated for the GII 2017 individual the limit values for the two output pil-
deciding whether (and how) or
country rankings. lars are 40%60% (see Table3).
not to impute missing data,
As suggested in the relevant lit- The GII developing team, for
deciding whether (and how) or
erature on composite indicators,5 the transparency and replicability, has
not to treat outliers,

robustness assessment was based on always opted not to estimate missing

selecting the normalization
Monte Carlo simulation and multi- data. The no imputation choice,
approach to be applied,
modelling approaches, applied to which is common in similar contexts,
choosing the weights to be
error-free data where potential out- might encourage economies not to
assigned, and
liers and eventual errors and typos report low data values. Yet this is not
deciding on the aggregation rule
have already been corrected in a the case for the GII. After 10 editions
to be implemented.
preliminary stage. In particular, the of the GII, the index-developing

Figure 2a: Robustness analysis (GII rank vs. median rank, 90% confidence intervals)
Annex 3: JRC Statistical Audit of the GII

11 l Median rank
GII 2017 rank
GII 2017: ranks and intervals of simulated ranks



Source: European Commission Joint Research Centre, 2017.

Notes: Median ranks and intervals are calculated over 4,000 simulated scenarios combining random weights, imputed versus missing values, and geometric versus arithmetic average at the pillar level. The Spearman rank correlation between the
median rank and the GII 2017 rank is 0.997.

team has not encountered any inten- argue that the United Kingdom and profiles and motivates economies to
tional no-reporting strategy. The Germany, despite their similar per- improve in the GII pillars in which
consequence of the no imputation formance at the Innovation Output they perform poorly, and not just in
choice in an arithmetic average is Sub-Indexboth close to 53.5 points any GII pillar.8
that it is equivalent to replacing an (rank 6th and 7th respectively)are Four models were tested based
indicators missing value for a given very different if one considers how on the combination of no imputa-
country with the respective sub- these countries perform within the tion versus EM imputation, and
pillar score. Hence, the available data sub-index. Germany ranks 8th in arithmetic versus geometric average,
(indicators) in the incomplete pillar Knowledge and technology outputs combined with 1,000 simulations
may dominate, sometimes biasing and 7th in Creative outputs, while per model (random weights versus
the ranks up or down. To test the the United Kingdom is much more fixed weights), for a total of 4,000
impact of the no imputation choice, diverse: the country ranks 13th posi- simulations for the GII and each
the JRC estimated missing data using tion in Knowledge and technology of the two sub-indices (see Table3
the Expectation Maximization (EM) outputs, but it notably improves for a summary of the uncertainties
algorithm.6 its overall position in the Output considered).
Regarding the aggregation for- Sub-Index thanks to its 4th rank in
mula, decision-theory practitioners Creative outputs. To assess the impact Uncertainty analysis results

challenge the use of simple arith- of this compensability issue, the JRC The main results of the robustness
metic averages because of their fully relaxed the strong perfect substitut- analysis are shown in Figure2 with
compensatory nature, in which a ability assumption inherent in the median ranks and 90% confidence
comparative high advantage on a arithmetic average and considered intervals computed across the 4,000
few indicators can compensate a instead the geometric average, which Monte Carlo simulations for the GII
comparative disadvantage on many is a partially compensatory approach and the two sub-indices. The figure
indicators.7 For example, one may that rewards economies with balanced orders economies from best to worst

Figure 2b: Robustness analysis (Input rank vs. median rank, 90% confidence intervals)

Annex 3: JRC Statistical Audit of the GII

GII 2017: Input ranks and interval of simulated ranks
11 l Median rank
21 GII 2017 Input rank



Source: European Commission Joint Research Centre, 2017.

Notes: Median ranks and intervals are calculated over 4,000 simulated scenarios combining random weights, imputed versus missing values, and geometric versus arithmetic average at the pillar level. The Spearman rank correlation between the
median rank and the Innovation Input 2017 rank is 0.997.

Figure 2c: Robustness analysis (Output rank vs. median rank, 90% confidence intervals)

l Median rank
GII 2017: Output ranks and interval of simulated ranks

GII 2017 Output rank



Source: European Commission Joint Research Centre, 2017.

Notes: Median ranks and intervals are calculated over 4,000 simulated scenarios combining random weights, imputation versus no imputation of missing values, and geometric versus arithmetic average at the pillar level. The Spearman rank
correlation between the median rank and the Innovation Output 2017 rank is 0.995.

Table 4: GII 2017 and Input/Output Sub-Indices: Ranks and 90% confidence intervals
Annex 3: JRC Statistical Audit of the GII

GII 2017 Input Sub-Index Output Sub-Index

Country/Economy Rank Interval Rank Interval Rank Interval
Switzerland 1 [1, 1] 3 [2, 4] 1 [1, 1]
Sweden 2 [2, 3] 2 [1, 4] 3 [3, 4]
Netherlands 3 [2, 3] 9 [8, 13] 2 [2, 2]
United States of America 4 [4, 5] 5 [2, 8] 5 [4, 8]
United Kingdom 5 [4, 5] 7 [4, 7] 6 [5, 10]
Denmark 6 [6, 10] 6 [4, 8] 12 [10, 13]
Singapore 7 [6, 11] 1 [1, 2] 17 [16, 19]
Finland 8 [6, 9] 4 [4, 8] 13 [11, 13]
Germany 9 [6, 9] 17 [14, 18] 7 [4, 7]
Ireland 10 [7, 12] 19 [13, 19] 8 [5, 12]
Korea, Republic of 11 [7, 11] 16 [11, 19] 9 [5, 10]
Luxembourg 12 [11, 13] 24 [23, 27] 4 [3, 6]
Iceland 13 [13, 18] 21 [20, 22] 10 [9, 14]
Japan 14 [13, 15] 11 [9, 11] 20 [17, 21]
France 15 [13, 17] 15 [13, 18] 18 [16, 19]
Hong Kong (China) 16 [13, 21] 8 [4, 10] 25 [23, 25]
Israel 17 [14, 21] 20 [12, 21] 14 [14, 20]
Canada 18 [17, 22] 10 [8, 13] 23 [23, 29]
Norway 19 [18, 21] 14 [12, 19] 22 [22, 23]
Austria 20 [17, 21] 18 [15, 20] 21 [20, 21]
New Zealand 21 [19, 23] 13 [12, 20] 24 [22, 24]
China 22 [16, 23] 31 [24, 33] 11 [8, 11]
Australia 23 [22, 26] 12 [10, 16] 30 [29, 30]
Czech Republic 24 [21, 26] 27 [25, 28] 16 [13, 16]
Estonia 25 [24, 26] 26 [24, 27] 19 [16, 20]
Malta 26 [24, 26] 28 [27, 31] 15 [14, 17]
Belgium 27 [27, 27] 22 [21, 22] 27 [26, 29]
Spain 28 [28, 28] 25 [23, 27] 26 [25, 27]
Italy 29 [29, 30] 29 [27, 32] 29 [26, 29]
Cyprus 30 [29, 31] 32 [29, 33] 28 [26, 31]
Portugal 31 [30, 32] 33 [30, 33] 31 [31, 33]
Slovenia 32 [31, 32] 30 [27, 32] 34 [34, 35]
Latvia 33 [33, 34] 35 [35, 38] 33 [29, 35]
Slovakia 34 [33, 35] 39 [38, 41] 35 [32, 35]
United Arab Emirates 35 [34, 40] 23 [23, 31] 56 [54, 58]
Bulgaria 36 [34, 37] 45 [41, 47] 32 [31, 33]
Malaysia 37 [34, 37] 36 [33, 37] 39 [38, 39]
Poland 38 [38, 39] 37 [36, 39] 41 [40, 41]
Hungary 39 [37, 39] 41 [39, 44] 37 [36, 37]
Lithuania 40 [39, 41] 34 [34, 35] 49 [47, 52]
Croatia 41 [41, 45] 44 [42, 47] 46 [45, 49]
Romania 42 [41, 45] 51 [45, 52] 44 [42, 48]
Turkey 43 [40, 46] 68 [57, 71] 36 [36, 40]
Greece 44 [42, 54] 38 [36, 46] 59 [57, 63]
Russian Federation 45 [41, 46] 43 [36, 48] 51 [48, 53]
Chile 46 [43, 48] 42 [39, 45] 53 [50, 53]
Viet Nam 47 [43, 53] 71 [65, 75] 38 [36, 43]
Montenegro 48 [47, 52] 50 [47, 54] 52 [51, 55]
Qatar 49 [47, 55] 48 [45, 55] 54 [54, 60]
Ukraine 50 [43, 52] 77 [59, 80] 40 [37, 40]
Thailand 51 [46, 51] 65 [55, 67] 43 [42, 44]
Mongolia 52 [46, 55] 67 [61, 73] 48 [40, 49]
Costa Rica 53 [48, 54] 57 [53, 63] 50 [48, 53]
Moldova, Republic of 54 [51, 58] 73 [70, 81] 42 [42, 45]
Saudi Arabia 55 [54, 62] 46 [40, 52] 66 [65, 73]

Kuwait 56 [55, 68] 80 [73, 90] 45 [44, 57]

South Africa 57 [54, 61] 49 [42, 51] 69 [65, 70]
Mexico 58 [55, 59] 54 [49, 58] 60 [58, 62]
Armenia 59 [56, 63] 82 [76, 87] 47 [46, 48]
India 60 [52, 63] 66 [52, 69] 58 [52, 59]
TFYR of Macedonia 61 [59, 65] 53 [53, 66] 63 [62, 67]
Serbia 62 [58, 63] 58 [55, 66] 61 [58, 62]
Panama 63 [55, 66] 74 [67, 82] 55 [48, 57]
Mauritius 64 [56, 74] 47 [43, 62] 82 [68, 87]

Table 4: GII 2017 and Input/Output Sub-Indices: Ranks and 90% confidence intervals (continued)

Annex 3: JRC Statistical Audit of the GII

GII 2017 Input Sub-Index Output Sub-Index
Country/Economy Rank Interval Rank Interval Rank Interval
Colombia 65 [61, 67] 52 [46, 53] 75 [75, 77]
Bahrain 66 [64, 68] 55 [53, 67] 67 [66, 73]
Uruguay 67 [64, 70] 61 [57, 73] 64 [63, 66]
Georgia 68 [64, 70] 69 [64, 79] 62 [61, 63]
Brazil 69 [68, 73] 60 [51, 67] 80 [78, 83]
Peru 70 [70, 79] 56 [53, 67] 85 [83, 86]
Brunei Darussalam 71 [69, 85] 40 [38, 51] 110 [94, 111]
Morocco 72 [69, 73] 79 [70, 82] 68 [64, 70]
Philippines 73 [67, 75] 83 [72, 84] 65 [60, 69]
Tunisia 74 [71, 79] 81 [71, 83] 71 [71, 82]
Iran, Islamic Republic of 75 [68, 80] 98 [84, 102] 57 [53, 57]
Argentina 76 [72, 78] 72 [57, 79] 81 [80, 82]
Oman 77 [74, 87] 62 [52, 79] 90 [89, 107]
Kazakhstan 78 [76, 84] 64 [59, 68] 93 [88, 94]
Dominican Republic 79 [77, 91] 88 [85, 98] 72 [69, 84]
Kenya 80 [76, 84] 91 [80, 102] 70 [68, 72]
Lebanon 81 [75, 85] 87 [76, 90] 78 [69, 78]
Azerbaijan 82 [80, 89] 78 [72, 88] 89 [87, 90]
Jordan 83 [78, 86] 92 [78, 98] 74 [73, 80]
Jamaica 84 [79, 88] 84 [78, 90] 84 [70, 86]
Paraguay 85 [79, 94] 90 [85, 93] 79 [61, 102]
Bosnia and Herzegovina 86 [82, 92] 75 [68, 86] 96 [91, 97]
Indonesia 87 [77, 90] 99 [89, 101] 73 [71, 74]
Belarus 88 [65, 93] 63 [50, 68] 109 [75, 116]
Botswana 89 [84, 93] 59 [57, 72] 111 [107, 113]
Sri Lanka 90 [81, 91] 94 [90, 101] 77 [74, 79]
Trinidad and Tobago 91 [84, 91] 85 [85, 92] 86 [81, 90]
Ecuador 92 [89, 96] 95 [90, 100] 83 [82, 101]
Albania 93 [92, 107] 70 [67, 86] 115 [115, 122]
Tajikistan 94 [90, 103] 100 [91, 104] 88 [82, 101]
Kyrgyzstan 95 [93, 98] 86 [80, 91] 104 [101, 112]
Tanzania, United Republic of 96 [94, 106] 109 [102, 118] 76 [76, 101]
Namibia 97 [88, 107] 89 [85, 98] 102 [84, 118]
Guatemala 98 [94, 99] 97 [92, 101] 92 [91, 98]
Rwanda 99 [94, 113] 76 [69, 90] 121 [110, 121]
Senegal 100 [94, 102] 102 [92, 103] 98 [93, 98]
Cambodia 101 [98, 106] 104 [103, 120] 87 [84, 88]
Uganda 102 [99, 104] 93 [89, 98] 106 [103, 116]
El Salvador 103 [93, 106] 96 [95, 101] 105 [89, 116]
Honduras 104 [96, 104] 103 [99, 105] 103 [87, 106]
Egypt 105 [97, 106] 106 [102, 109] 97 [94, 97]
Bolivia, Plurinational State of 106 [100, 108] 107 [101, 112] 99 [99, 111]
Mozambique 107 [104, 113] 114 [110, 116] 100 [96, 104]
Algeria 108 [107, 114] 105 [101, 109] 117 [114, 120]
Nepal 109 [108, 114] 108 [105, 120] 114 [102, 114]
Ethiopia 110 [106, 121] 122 [118, 124] 91 [90, 111]
Madagascar 111 [109, 121] 120 [117, 125] 95 [93, 106]
Cte d'Ivoire 112 [107, 114] 121 [113, 124] 94 [89, 100]
Pakistan 113 [107, 114] 116 [107, 120] 101 [98, 108]
Bangladesh 114 [111, 117] 113 [110, 122] 108 [105, 114]
Malawi 115 [114, 124] 112 [111, 122] 112 [109, 124]
Benin 116 [110, 119] 110 [107, 120] 120 [103, 120]
Cameroon 117 [115, 123] 117 [112, 122] 113 [110, 119]
Mali 118 [117, 121] 123 [113, 124] 107 [104, 116]
Nigeria 119 [118, 123] 118 [110, 122] 119 [118, 123]

Burkina Faso 120 [114, 127] 101 [92, 114] 126 [121, 127]
Zimbabwe 121 [117, 124] 124 [112, 124] 116 [113, 122]
Burundi 122 [121, 125] 115 [110, 125] 122 [122, 126]
Niger 123 [102, 124] 111 [101, 112] 123 [100, 125]
Zambia 124 [114, 124] 125 [108, 127] 118 [113, 121]
Togo 125 [113, 126] 119 [114, 121] 127 [106, 127]
Guinea 126 [123, 126] 126 [125, 127] 124 [123, 125]
Yemen 127 [125, 127] 127 [125, 127] 125 [124, 127]
Source: European Commission Joint Research Centre, 2017.

Table 5: Sensitivity analysis: Impact of modelling choices on economies with most sensitive ranks
Annex 3: JRC Statistical Audit of the GII

Number of economies that improve Number of economies that deteriorate

Index or by 20 or more between 10 by 20 or more between 10
Sub-Index Uncertainty tested (pillar level only) positions and 19 positions positions and 19 positions

GII Geometric vs. arithmetic average 0 1 0 3

EM imputation vs. no imputation of missing data 0 3 0 3
Geometric average and EM imputation vs. arithmetic average and missing values 1 (Belarus) 3 0 3
Input Geometric vs. arithmetic average 0 0 0 1
Sub-Index EM imputation vs. no imputation of missing data 0 2 0 2
Geometric average and EM imputation vs. arithmetic average and missing values 0 5 0 7
Output Geometric vs. arithmetic average 0 0 0 3
Sub-Index EM imputation vs. no imputation of missing data 1 (Belarus) 10 1 (Tanzania, U. Rep.) 7
Geometric average and EM imputation vs. arithmetic average and missing values 1 (Belarus) 9 1 (Tanzania, U. Rep.) 7
Source: European Commission Joint Research Centre, 2017.

according to their reference rank data availability within each of the expected rank in the realm of the
(black line), the dot being the median two sub-indices. Some caution is also GIIs unavoidable methodological
rank over the simulations. warranted in the Input Sub-Index for uncertainties, 75% of the economies
All published GII 2017 ranks lay 7 economiesUkraine, Argentina, are found to shift fewer than three
within the simulated 90% confidence Oman, Kenya, Jordan, Rwanda, positions with respect to the median
intervals, and for most economies and Burkina Fasothat have 90% rank in the GII, or in the Input and
these intervals are narrow enough for confidence interval widths over 20 Output Sub-Index.
meaningful inferences to be drawn: (up to 27 for Oman). The Output For full transparency and infor-
there is a shift of fewer than 10 posi- Sub-Index is slightly more sensi- mation, Table 4 reports the GII
tions for 105 of the 127 economies. tive to the methodological choices: 2017 Index and Input and Output
However, it is also true that merely 8 countriesParaguay, Belarus, Sub-Indices economy ranks together
two country ranks vary significantly the United Republic of Tanzania, with the simulated 90% confidence
with changes in weights and aggre- Namibia, El Salvador, Ethiopia, intervals in order to better appreciate
gation formula and because of the Niger, and Togohave 90% confi- the robustness of the results to the
estimation of missing data. These dence interval widths over 20 (up to choice of weights, of the aggregation
two countriesNiger and Belarus 41 for Paraguay and Belarus). This formula and the impact of estimating
have 90% confidence interval widths sensitivity is mostly the consequence missing data (where applicable).
of 22 and 28, respectively; hence of the estimation of missing data
their GII ranks should be interpreted and the fact that there are only two Sensitivity analysis results
cautiously and certainly not taken pillars: this means that changes to Complementary to the uncertainty
at face value. This is a remarkable the imputation method, weights, analysis, sensitivity analysis has been
improvement compared to the GII or aggregation formula have a more used to identify which of the mod-
2015, where conf idence interval notable impact on the country ranks elling assumptions have the highest
widths for 32 economies lay between in the Innovation Output. impact on certain country ranks.
20 and 29, for another 7 economies Although a few economy ranks, Table 5 summarizes the impact
between 30 and 39, and for 2 econo- in the GII 2017 overall or in the of changes of the EM imputation
mies the widths were 40 or greater. two sub-indices, appear to be sensi- method and/or the geometric aggre-

This improvement in the confidence tive to the methodological choices, gation formula, with fixed weights
one can attach to the GII 2017 ranks the published rankings for the vast at their reference values (as in the
is the direct result of the developers majority can be considered as repre- original GII). Similar to last years
choice since 2016 to adopt a more sentative of the plurality of scenarios results, this year neither the GII nor
stringent criterion for an economys simulated herein. Taking the median the Input or Output Sub-Index are
inclusion, which requires at least 62% rank as the yardstick for an economys found to be heavily inf luenced by

Table 6: Pie shares (absolute terms) and efficiency scores for the top 25 economies in the GII 2017

Annex 3: JRC Statistical Audit of the GII

Input pillars Output pillars
Human Knowledge Efficient Difference
capital Market Business and technology Creative frontier Efficiency from GII
Country/Economy Institutions and research Infrastructure sophistication sophistication outputs outputs rank (DEA) GII rank Difference ratio rank rank
Switzerland 0.09 0.17 0.08 0.09 0.19 0.19 0.19 1 1 0 2 1
Sweden 0.20 0.20 0.20 0.10 0.20 0.05 0.05 2 2 0 12 10
Netherlands 0.20 0.05 0.20 0.05 0.20 0.10 0.20 5 3 2 4 1
United States of America 0.20 0.20 0.20 0.20 0.10 0.05 0.05 5 4 1 21 17
United Kingdom 0.20 0.20 0.20 0.20 0.05 0.05 0.10 4 5 1 20 15
Denmark 0.20 0.20 0.20 0.20 0.05 0.05 0.10 5 6 1 34 28
Singapore 0.20 0.20 0.20 0.20 0.10 0.05 0.05 2 7 5 63 56
Finland 0.20 0.20 0.20 0.10 0.20 0.05 0.05 5 8 3 37 29
Germany 0.20 0.20 0.20 0.10 0.05 0.05 0.20 9 9 0 7 2
Ireland 0.20 0.20 0.20 0.10 0.20 0.05 0.05 14 10 4 6 4
Korea, Republic of 0.20 0.20 0.20 0.20 0.10 0.05 0.05 11 11 0 14 3
Luxembourg 0.20 0.05 0.20 0.20 0.10 0.05 0.20 16 12 -4 1 11
Iceland 0.20 0.10 0.20 0.05 0.20 0.05 0.20 19 13 -6 5 8
Japan 0.20 0.20 0.20 0.20 0.10 0.05 0.05 9 14 5 49 35
France 0.20 0.20 0.20 0.20 0.05 0.05 0.10 14 15 1 35 20
Hong Kong (China) 0.20 0.10 0.20 0.20 0.20 0.05 0.05 11 16 5 73 57
Israel 0.10 0.20 0.20 0.20 0.20 0.05 0.05 19 17 2 23 6
Canada 0.20 0.20 0.20 0.20 0.10 0.05 0.05 11 18 7 59 41
Norway 0.20 0.20 0.20 0.20 0.10 0.05 0.05 19 19 0 51 32
Austria 0.20 0.20 0.20 0.10 0.20 0.05 0.05 19 20 1 41 21
New Zealand 0.20 0.20 0.20 0.20 0.05 0.05 0.10 16 21 5 56 35
China 0.05 0.10 0.20 0.20 0.20 0.20 0.05 23 22 1 3 19
Australia 0.20 0.20 0.20 0.20 0.05 0.05 0.10 16 23 7 76 53
Czech Republic 0.20 0.20 0.20 0.10 0.05 0.05 0.20 28 24 4 13 11
Estonia 0.20 0.05 0.20 0.20 0.10 0.05 0.20 23 25 2 19 6

Source: European Commission Joint Research Centre, 2017.

Notes: Pie shares are in absolute terms, bounded by 0.05 and 0.20. In the GII 2017, however, the five input pillars each have a fixed weight of 0.10; the two output pillars each have a fixed weight of 0.25.

the imputation of missing data or the in order to better appreciate to what Efficiency frontier in the GII by Data
aggregation formula. Depending on degree a countrys rank depends on Envelopment Analysis
the combination of the choices made, the modelling choices. Since 2016, Is there a way to benchmark countries
only Belarus or the United Republic following the JRC recommendation multi-dimensional performance on inno-
of Tanzania can change rank by 20 in past GII audits, the developers vation without imposing a fixed and com-
positions or more. choice to apply the 66% indicator mon set of weights that may not be fair to
All in all, the published GII 2017 coverage threshold separately to the a particular country?
ranks are reliable and for the vast Input and Output Sub-Indices in the Several innovation-related policy
majority of countries the simulated GII 2017 has led to a net increase issues at the national level entail an
90% confidence intervals are narrow in the reliability of country ranks intricate balance between global

enough for meaningful inferences to for the GII and the two sub-indices. priorities and country-specific strat-
be drawn. Nevertheless, the readers Furthermore, the adoption of less egies. Comparing the multi-dimen-
of the GII 2017 report should con- stringent criterion for the skewness sional performance on innovation
sider country ranks in the GII 2017 and kurtosis (greater than 2.25 in by subjecting countries to a fixed
and in the Input and Output Sub- absolute value and greater than 3.5, and common set of weights may
Indices not only at face value but also respectively) has not introduced any prevent acceptance of an innova-
within the 90% confidence intervals bias in the estimates. tion index on grounds that a given

Figure 3: GII 2017 scores and DEA distance to the efficient frontier scores
Annex 3: JRC Statistical Audit of the GII


DEA efficiency
0.8 GII 2017 score (rescaled)
GII 2017 (rescaled) and DEA efficiency






Source: European Commission Joint Research Centre, 2017.

Note: For comparison purposes, we have rescaled the GII scores by dividing them with the best performer in the overall GII 2017.

weighting scheme might not be fair preclude the possibility of a country upper and lower bounds. The pie
to a particular country. An appeal- achieving a perfect score by assigning shares are quite diverse, ref lecting
ing feature of the Data Envelopment a zero weight to weak pillars: for each the different national innovation
Analysis (DEA) literature applied in country, the share of each pillar score strategies. These pie shares can also
real decision-making settings is to (i.e., the pillar score multiplied by the be seen to ref lect countries compara-
determine endogenous weights that DEA weight over the total score) has tive advantage in certain GII pillars
maximize the overall score of each upper and lower bounds of 5% and vis--vis all other countries and all
decision-making unit given a set of 20% respectively. The DEA score is pillars. For example, Switzerland
other observations. then measured as the weighted aver- is the only country this year that
In this section, the assumption of age of all seven pillar scores, where obtains a perfect DEA score of 1 by
fixed pillar weights common to all the weights are the country-specific assigning 19% of its DEA score to
countries is relaxed once more; this DEA weights, compared to the best Business sophistication, Knowledge
time country-specific weights that performance among all other coun- and technology outputs, and Creative
maximize a countrys score are deter- tries with those same weights. The outputs, while merely 8% to 9% of its
mined endogenously by DEA.9 In DEA score can be interpreted as a DEA score comes from Institutions,
theory, each country is free to decide measure of the distance to the effi- Infrastructure, and Market sophisti-
on the relative contribution of each cient frontier. cation. Instead, countries including
pillar to its score, so as to achieve the Table6 presents the pie shares and the United States of America, the

best possible score in a computation DEA scores for the top 25 countries United Kingdom, Denmark, and
that ref lects its innovation strategy. in the GII 2017, next to the GII 2017 Singapore would assign 20% of their
In practice, the DEA method assigns ranks and efficiency ratio ranks. All DEA scores to Market sophistication.
a higher (lower) contribution to those pie shares are in accordance with the Only Switzerland reaches a perfect
pillars in which a country is relatively starting point of granting leeway to DEA score of 1, closely followed
strong (weak). Reasonable con- each country when assigning shares, by Sweden, the Netherlands, the
straints on the weights are applied to while not violating the (relative) United States of America, the United

Kingdom, Denmark, Singapore, and undesirable impact on some country weights (fixed vs. simulated), and

Annex 3: JRC Statistical Audit of the GII

Finland, which score between 0.96 scores, with the additional negative the aggregation formula (arithmetic
(Finland) and 0.99 (Sweden) in terms side-effect that it may encourage vs. geometric average) at the pil-
of efficiency. Figure3 shows how countries not to report low data lar level. For the vast majority of
close the DEA scores and the GII values. The adoption, since 2016, by countries these intervals are narrow
2017 scores are for all 127 economies the GII team of a more stringent data enough for meaningful inferences
(correlation of 0.99). 10 Note that by coverage threshold (at least 66% for to be drawn: the intervals comprise
construction, the version of DEA the input- and output-related indica- fewer than 10 positions for 83% (105
used herein is closer to the GII than tors, separately) has notably improved out of 127) of the economies. Some
to the efficiency ratio calculated as the confidence in the country ranks caution is needed mainly for two
the Output Sub-Index score divided for the GII and the two sub-indices. countriesBelarus and Nigerwith
by the Input Sub-Index score (with a Additionally, the choice of the GII ranks that are highly sensitive to the
correlation of 0.63). team, which was made in 2012, to use methodological choices. The Input
weights as scaling coefficients during and the Output Sub-Indices have the
the development of the index consti- same modest degree of sensitivity to
Conclusion tutes a significant departure from the the methodological choices related to
The JRC analysis suggests that the traditional, yet erroneous, vision of the imputation method, weights, or
conceptualized multi-level structure weights as a ref lection of indicators aggregation formula. Country ranks,
of the GII 2017with its 81 indica- importance in a weighted average. either in the GII 2017 or in the two
tors, 21 sub-pillars, 7 pillars, 2 sub- It is hoped that such a consideration sub-indices, can be considered repre-
indices, up to an overall indexis will be made also by other develop- sentative of the many possible scenar-
statistically sound and balanced: that ers of composite indicators to avoid ios: 75% of the countries shift fewer
is, each sub-pillar makes a similar situations where bias sneaks in when than three positions with respect to
contribution to the variation of its least expected. the median rank in the GII or either
respective pillar. Nevertheless, a The strong correlations between of the Input and Output Sub-Indices.
careful ref lection by the GII team the GII components are proven All things considered, the present
is needed for seven out of the 81 not to be a sign of redundancy of JRC audit findings confirm that the
indicators because their capacity to information in the GII. For more GII 2017 meets international qual-
distinguish countries performance than 42.2% (up to 60.9%) of the 127 ity standards for statistical soundness,
is lost in the aggregation at the pillar economies included in the GII 2017, which indicates that the GII index
level or higher. Five indicators related the GII ranking and the rankings of is a reliable benchmarking tool for
to the inputs of innovation 2.1.2 any of the seven pillars differ by 10 innovation practices at the country
Government expenditure on edu- positions or more. This demonstrates level around the world.
cation per pupil, secondary; 2.2.2 the added value of the GII ranking, Finally, the distance to the effi-
Graduates in science and engineer- which helps to highlight other com- cient frontier measure calculated
ing; 3.2.3 Gross capital formation; ponents of innovation that do not with Data Envelopment Analysis
5.2.3 GERD financed by abroad; emerge directly by looking into the could complement the Innovation
5.3.4 Foreign direct investment net seven pillars separately. At the same Efficiency Ratio as a measure of
inf lowsand two indicators related time, this finding points to the value efficiency, even if it is conceptually
to the outputs of innovation, 6.2.1 of duly taking into account the GII closer to the GII score than to the
Growth rate of GDP per person pillars, sub-pillars, and individual efficiency ratio.
engaged and 7.2.4 Printing and pub- indicators on their own merit. By The GII should not be seen as
lishing output, need to be reviewed doing so, country-specific strengths the ultimate and definitive ranking
because their statistical relevance to and bottlenecks in innovation can be of countries with respect to inno-
the GII framework is very weak, identified and serve as an input for vation. On the contrary, the GII

unlike their strong conceptual rel- evidence-based policy making. best represents an ongoing attempt
evance. The no-imputation choice All published GII 2017 ranks lie by Cornell University, the business
for not treating missing values, com- within the simulated 90% confidence school INSEAD, and the World
mon in relevant contexts and justi- intervals that take into account the Intellectual Property Organization
fied on grounds of transparency and unavoidable uncertainties in the to find metrics and approaches
replicability, can at times have an estimation of missing data, the that better capture the richness of

innovation, continuously adapting 9 A question that arises from the GII approach References and related reading
Annex 3: JRC Statistical Audit of the GII

is whether there is a way to benchmark

the GII framework to ref lect the countries multi-dimensional performance
Barbosa, N. and A. P. Faria. 2011. Innovation across
Europe: How important are institutional
improved availability of statistics on innovation without imposing a fixed
differences. Research Policy 40: 115769.
and common set of weights that may not
and the theoretical advances in the be fair to a particular country. The original Becker, W., M. Saisana, P. Paruolo, and I.
field. In any case, the GII should be question in the DEA literature was how to Vandecasteele. 2017. Weights and
regarded as a sound attempt to pave measure each units relative efficiency in Importance in Composite Indicators: Closing
production compared to a sample of peers, the Gap. Ecological Indicators 80: 1222.
the way for better and more informed given observations on input and output
innovation policies worldwide. quantities and, often, no reliable information Charnes, A. and W. W. Cooper. 1985. Preface to
on prices (Charnes and Cooper, 1985). A Topics in Data Envelopment Analysis. Annals
notable difference between the original of Operations Research 2: 5994.
DEA question and the one applied here is
Cherchye, L., W. Moesen, N. Rogge, T. Van
that no differentiation between inputs and
Notes outputs is made (Cherchye et al., 2008; Melyn
Puyenbroeck, M. Saisana, A. Saltelli, R. Liska,
1 OECD/EC JRC, 2008, p. 26. and S. Tarantola. 2008. Creating Composite
and Moesen, 1991). To estimate DEA-based
Indicators with DEA and Robustness Analysis:
distance to the efficient frontier scores, we
2 The JRC analysis was based on the The Case of the Technology Achievement
consider the m = 7 pillars in the GII 2017 for
recommendations of the OECD/EC JRC Index. Journal of Operational Research Society
n = 127 countries, with yij the value of pillar
(2008) Handbook on Composite Indicators 59: 23951.
j in country i. The objective is to combine
and on more recent research from the JRC.
the pillar scores per country into a single Groeneveld, R. A. and G. Meeden. 1984. Measuring
The JRC audits on composite indicators
number, calculated as the weighted average Skewness and Kurtosis. The Statistician 33:
are conducted upon request of the index
of the m pillars, where wi represents the 39199.
developers and are available at https://
weight of the i-th pillar. In absence of reliable Little, R. J. A. and D. B. Rubin. 2002. Statistical Analysis
information about the true weights, the
3 Groeneveld and Meeden (1984) set the weights that maximize the DEA-based scores with Missing Data. 2nd edition. Hoboken, NJ:
criteria for absolute skewness above 1 and are endogenously determined. This gives the John Wiley & Sons, Inc.
kurtosis above 3.5. The skewness criterion following linear programming problem for
Melyn, W. and W. Moesen. 1991. Towards a
was relaxed to account for the small sample each country j:
7 Synthetic Indicator of Macroeconomic
(127 economies).  yij wij Performance: Unequal Weighting when
4 Nunnally, 1978.
Yi  max j=1
Limited Information Is Available. Public
wij 7 constraint)
max  ycj wij Economics Research Paper 17. Leuven: Centre
yc {dataset} j=1
5 Saisana et al., 2005; Saisana et al., 2011. for Economic Studies.

6 The Expectation-Maximization (EM) algorithm subject to Munda, G. 2008. Social Multi-Criteria Evaluation for
(Little and Rubin, 2002; Schneider, 2001) is an (non-negativity a Sustainable Economy. Berlin Heidelberg:
wij  0,
iterative procedure that finds the maximum constraint) Springer-Verlag.
likelihood estimates of the parameter vector
where Nunally, J. 1978. Psychometric Theory. New York:
by repeating two steps: (1) The expectation
E-step: Given a set of parameter estimates, j = 1, , 7, McGraw-Hill.
such as a mean vector and covariance matrix i = 1, , 127
OECD/EC JRC (Organisation for Economic
for a multivariate normal distribution, the Co-operation and Development/European
In this basic programming problem, the
E-step calculates the conditional expectation Commission, Joint Research Centre). 2008.
weights are non-negative and a countrys
of the complete-data log likelihood given the Handbook on Constructing Composite
score is between 0 (worst) and 1 (best).
observed data and the parameter estimates. Indicators: Methodology and User Guide. Paris:
(2) The maximization M-step: Given a 10 Instead, only Switzerland achieved a 1.0 score OECD.
complete-data log likelihood, the M-step in the Innovation Efficiency Ratio, calculated
finds the parameter estimates to maximize as the ratio of the Output Sub-Index over Paruolo, P., M. Saisana, and A. Saltelli. 2013. Ratings
the complete-data log likelihood from the the Input Sub-Index. The Efficiency Ratio and Rankings: Voodoo or Science? Journal of
E-step. The two steps are iterated until the and the DEA score embed very different the Royal Statistical Society A 176 (3): 60934.
iterations converge. concepts of efficiency, leading to completely
Saisana, M., B. DHombres, and A. Saltelli. 2011.
different results and insights. A high score in
7 Munda, 2008. Rickety Numbers: Volatility of University
the Innovation Efficiency Ratio is obtained
Rankings and Policy Implications. Research
8 In the geometric average, pillars are by scoring more on the Output Sub-Index
Policy 40: 16577.
multiplied as opposed to summed in the than on the Input Sub-Index, irrespective of
arithmetic average. Pillar weights appear the actual scores in these two sub-indices. Saisana, M., A. Saltelli, and S. Tarantola. 2005.
as exponents in the multiplication. All pillar Instead, a high score in the DEA score can be Uncertainty and Sensitivity Analysis
scores were greater than zero, hence there obtained by having comparative advantages Techniques as Tools for the Analysis and
was no reason to rescale them to avoid zero on several GII pillars (irrespective of these Validation of Composite Indicators. Journal of
values that would have led to zero geometric being input or output pillars). The DEA scores the Royal Statistical Society A 168 (2): 30723.
averages. are therefore closer to the GII scores than to
the Innovation Efficiency Ratio. Saltelli, A., M. Ratto, T. Andres, F. Campolongo,
J. Cariboni, D. Gatelli, M. Saisana, and S.

Tarantola. 2008. Global Sensitivity Analysis:

The Primer. Chichester, England: John Wiley
& Sons.

Schneider, T. 2001. Analysis of incomplete

climate data: Estimation of mean values
and covariance matrices and imputation
of missing values. Journal of Climate 14,
ANNEX 4 73

Measuring Innovation in Agriculture and Food Systems

Annex 4: Measuring Innovation in Agriculture and Food Systems

Agriculture and food innovation the most part.1 Although the agricul- measuring innovation in specific
systems are complex and constantly ture sector is likely to be included in systems and sectors, thereby laying
evolving. Today robotics and bio- future revisions of the Oslo Manual, it the foundations for interesting future
technological and digital technolo- is currently unclear whether the cov- work.
gies are applied in agriculture and erage of the business sector alone will
food systems. New actors enter the satisfy the innovation data require-
systems and traditional actors, such ments of the agriculture sector. Potential indicators to benchmark
as farmers and food companies, grow Third, in developing countries, innovation in agriculture and food
into commercial farmers, bio/organic agricultural activities and related systems
producers, and so on. innovations often take place at the Table1 shows how the GII frame-
Agriculture and food systems farm or household level (especially in work could be used to measure the
also vary greatly across countries, case of subsistence farming), not in characteristics of agriculture and
ref lecting each countrys level of private-sector firms as captured by food innovation systems. The table
development as well as the role that most data collections. Statistically, includes only the indicators that are
agriculture and food sectors play. however, capturing activity in the relevant to measuring innovation in
Measuring agricultural innova- informal sector or at the grassroots agriculture and food systems and that
tion is challenging for several reasons: level is challenging.2 are available for a large number of
First, agriculture and food sys- Clearly, the work of the African economies.5 The next sections look
tems span many different sectors, UnionNew Partnership for Africas into some of these indicators and
products, and service groups that Development (AU-NEPAD) on the provide snapshots of top performing
are not easily grasped and that go far African Innovation Outlook,3 and economies in each selected indicator.
beyond the agriculture sector or agri- application of innovation surveys, for
cultural farms alone. As Chapter 1 example, is ongoing.4 Yet the focus is
and the other substantive chapters of currently not on the informal or the Human capital and research
this report show, innovations occur agriculture sector. Education and research and develop-
along the value chain and involve (1) As a result of the complexities ment (R&D) investment are key to
agricultural inputs such as fertilizers outlined above, and because of a lack boosting productivity; they are also
and seeds, at times coming from the of robust metrics (see Chapter 2), key for advancing the agriculture and
chemical or the biotechnology sector; measuring innovation in agriculture food sector.6,7 Various studies dem-
(2) product innovations coming from and food systems is a difficult endeav- onstrate that better-educated farm-
the capital goods sector; and (3) pro- our. This annex maps agriculture ers have the skills to run their farms
cess or organizational innovations in and food systems based on the GII more efficiently and are more prone
the fields of payments, logistics, and framework. to embracing innovation.8 Education

distribution services coming from the Although incomplete, this map- has also proven to spill over, affecting
banking, transport, and retail sectors. ping illustrates the above challenge the productivity of family members
Second, key innovation data and provides guidance to research- and neighbours.9
sources such as the innovation sur- ers and policy makers interested in In spite of its role in agriculture
veys based on the Oslo Manual focus benchmarking their agriculture and and food systems, data on farmers
on the manufacturing and services food systems. It also shows how the education are limited. This has led
sectors, thus excluding agriculture for GII framework could be adapted to researchers to use other proxies, such

Table 1: Adapting the GII framework to agriculture and food systems

Annex 4: Measuring Innovation in Agriculture and Food Systems

GII pillar GII indicator Are indicators available for agri-food? Corresponding indicator in agri-food Additional indicators

Expenditure on education For only a few economies

Tertiary enrolment Yes Tertiary students in agriculture
Graduates in science & engineering Yes ODA for agricultural education/training
Human capital
Researchers Yes Agricultural researchers
and research
Gross expenditures on R&D Yes Agricultural R&D expenditures ODA for agricultural
Global R&D companies, average expenditure No
QS university rankings No
Ease of getting credit For only a few economies
Domestic credit to private sector Yes Credit to agriculture
Microfinance gross loans For only a few economies
sophistication Venture capital deals No
Applied tariff rate Yes Applied tariff rate for agriculture
and food items
Intensity of local competition No
Knowledge-intensive employment
Firms offering formal training Yes Firms offering formal training in food-
GERD performed by business For only a few economies
GERD financed by business No
Females employed w/ advanced degrees No
University/industry research collaborations No
State of cluster development No
GERD financed by abroad No
JV-strategic alliance deals No
Patent families in 2+ offices Yes Agri-food patent families in 2+ offices
IP payments No
High-tech imports Yes High-tech imports for agri-food sector Use of fertilizers;
Machinery in use
FDI net inflows Yes Agri-food FDI inflows
Patents by origin Yes Agri-food patents by origin Plant varieties
PCT patent applications Yes Agri-food PCT patent applications
Utility models by origin Yes Agri-food utility models by origin
Scientific and technical articles Yes Scientific and technical articles in agri-food
and Citable documents H index Yes Citable documents in agri-food
technology Growth rate of PPP$ GDP/worker Yes Agriculture labour productivity growth
outputs New businesses No
ISO 9001 quality certificates No
IP receipts No
High-tech exports Yes Agri-food exports

FDI net outflows Yes Agri-food FDI outflows

Trademarks Yes Agri-food trademarks Geographic
indications registered
Creative Industrial designs Yes Agri-food industrial designs
ICTs & business model creation No
ICTs & organizational model creation No

Notes: The GII pillars Institutions and Infrastructure are not included in this table because the metrics in those pillars already capture the role of institutions and infrastructure in agriculture and food systems. ODA = official development
assistance; = data currently under review.

Table 2: Official development assistance Table 3: Agricultural R&D expenditures: Top five economies

Annex 4: Measuring Innovation in Agriculture and Food Systems

for education and training: Top five Economy US$, thousands Economy Share of agriculture value added
India 3,857 Singapore 1.48
Economy ODA in US$, millions Korea, Rep. 1,521 Qatar 0.11
Afghanistan 8.2 China 1,149 Netherlands 0.10
Ethiopia 4.6 Netherlands 1,145 Trinidad and Tobago 0.10
China 4.3 Australia 842 Denmark 0.06
Indonesia 4.1
Data source: UNESCO-UIS Science & Technology Data Center, February 2017. Available at
Uganda 3.4
Notes: Where data are not available, data from previous years are used. R&D expenditures are in 2005 PPP$. Data are available for 73 economies. Many
Organisation for Economic Co-operation and Development (OECD) economies, including the United States of America (USA), as well as other large economies
Data source: FAOstats, February 2017. Available at such as Argentina and Brazil, are excluded because of a lack of data.
Note: Data refer to total disbursements from bilateral and multilateral
donors for 2014.

as official development assistance delay the adoption of innovation.13 of its agriculture sector (as measured
(ODA) for education and training Indeed, research demonstrates that by value added), investing roughly
(see Table2). According to available developing countries that invested 150% of its output in R&D. Qatar,
data, Afghanistan, Ethiopia, China, the most in R&D while simultane- the Netherlands, and Trinidad and
Indonesia, and Uganda receive the ously investing in extension have had Tobago follow, with roughly 10%
highest amounts of aid in agricultural the strongest productivity leap.14 of their agricultural output spent in
education and training. Other top Overall, the top agricultural R&D. Denmark spends 6% of its
recipients include Malawi, Myanmar, R&D spenders are India, the agricultural output in R&D.
and Sierra Leone. Republic of Korea (Korea), China, Another way to look at R&D is
Lagging R&D expenditures in the Netherlands, and Australia (see through ODA disbursements to agri-
high-, middle-, and low-income Table3), with India spending more cultural research. Nigeria, Argentina,
economies affect productivity than double than Korea. Singapore India, Uganda, and Ethiopia are the
growth and innovation in agriculture. spends the most in relation to the size largest recipients of ODA in this
According to the data available, only
about 6% of the worlds R&D invest-
ments and researchers are devoted to
agricultural sciences (see Figure1).10 Figure 1: Researchers and R&D expenditure in agriculture sciences
Although advanced economies have
historically been the leaders in agri- 20 2.0

cultural R&D, research capacity has

also reached high standards in several
Researchers (%)/R&D expenditure (%)

emerging economiessuch as China,

R&D expenditure (% GDP)

15 1.5
in agricultural sciences

India, Brazil, Argentina, and South

Africa.11 In agriculture, R&D affects
output with a long lag, but the impact
10 1.0
lasts for a long time.12 R&D spillovers
tend to be geographically bounded
because innovations produced in one
part of the world require adaptations 5 0.5

to work well in local soil and climate


conditions. This makes indigenous

R&D efforts essential. Developing 0 0.0
countries, especially in Sub-Saharan Africa Asia Europe Latin America World
and the Caribbean
Africa, have traditionally underspent
n Researchers (% ag sciences) n R&D expenditures (% ag sciences) l R&D expenditures (% GDP)
in agricultural R&D (see Chapter 2).
When they undertake R&D, poor (or Data source: UNESCO-UIS Science & Technology Data Center, February 2017. Available at
lacking) extension services generally Notes: Data on researchers in agricultural sciences are based on headcount (HC) measurement. Because of a lack of data, Northern America is missing. Data refer
to 2014.

Table 4: ODA to agricultural research: Table 5: Tertiary students in agricultural Business sophistication
Annex 4: Measuring Innovation in Agriculture and Food Systems

Top five economies studies: Top five economies The adoption of synthetic fertiliz-
Economy ODA (US$, millions) Economy Share of tertiary students (%) ers, together with high-yield crop
Nigeria 30.3 Ethiopia 8.0
varieties, has been at the basis of the
Argentina 28.2 Uzbekistan 7.5
green revolution. Today, despite the
India 24.0 Cambodia 6.8
growing demand for organic food,
Uganda 16.9 Viet Nam 6.4
less than 1% of agricultural land is
farmed using organic methods.16
Ethiopia 16.9 Albania 6.3
Although organic farming has a
Data source: FAOstats, February 2017. Available at Data source: UNESCO-UIS Science & Technology Data Center, February 2017.
faostat/en/. Available at
number of advantages, synthetic
Note: Data refer to total disbursements from bilateral and multilateral Notes: When data for 2014 were not available, data points up to 2008 fertilizers are still widely used.17
donors for 2014. were used. Limited access to high-quality
fertilizers is still an issue in many
countries, most notably in Sub-
Table 6: Agricultural credit markets: Top five economies Saharan Africa (see for example
the case of Uganda, described in
Economy US$, millions Economy Share of total credit (%) Chapter 11). Estimates indicate that,
United States of America 74,951 New Zealand 26 from 2009 to 2015, global demand
Germany 57,983 Uruguay 17 for fertilizers grew by roughly 15%,
Australia 54,968 Kyrgyzstan 12 and will grow at least 1.6% annually
France 54,812 Tajikistan 12 from 2015 to 2020. Sub-Saharan
New Zealand 44,903 Bolivia, Plurinational St 11 Africa will be responsible for most
Data source: FAOstats, February 2017. Available at
of this growth, reaching an average
Note: Data for 2014, available for 69 economies. annual growth rate of 4.4%.18
Data on current fertilizer con-
sumption show that global con-
sumption is highly concentrated,
area. Argentina is the only top ODA Market sophistication with one single economyChina
recipient among upper-middle- Financial markets are important consuming 31% of total world fertil-
income economies, while the others components of any innovation sys- izers (see Table7). Although the gap
are mostly low- and lower-middle- tem. In agriculture, credit is essential in fertilizer consumption between
income economies. Among the top to modernize farms and access high- China and other economies is con-
10 recipients are Kenya, the United quality inputs such as seeds and fer- siderable, according to available
Republic of Tanzania, and Indonesia, tilizers. Given the size and nature of data, other top fertilizer consum-
which each received more than US$10 most farms, credit constraints can be ers are India, the USA, Brazil, and
million. Finally, ODA to agricultural often severe.15 According to available Indonesia. By contrast, Sub-Saharan
research reaches much higher values data, the countries with the largest African countries together account
than ODA to agricultural education credit markets for agriculture are the for only 3% of total world con-
and training (see Table4). USA, Germany, Australia, France, sumption. Considered in relation to
Data on the share of tertiary stu- and New Zealand (see Table6). It arable land, Qatar, Malaysia, Hong
dents enrolled in agricultural studies is worth recalling that these econo- Kong (China), New Zealand, and
indicate that agricultural studies are mies have very large credit markets. Bahrain are the five top consumers;
particularly relevant in the develop- Indeed, in the GII, New Zealand, the other important consumers include
ing world. The top five highest shares USA, and Australia rank among the Singapore, Costa Rica, the United

of agricultural students in tertiary top five economies in the Credit sub- Arab Emirates, and Colombia.
students are in Ethiopia, Uzbekistan, pillar. Still, New Zealand is the coun- Mechanization of agriculture
Cambodia, Viet Nam, and Albania try that allocated the highest portion has also contributed greatly to
(see Table5). Other countries with of its credit to agriculture (26%). productivity growth in agriculture.
high shares of agricultural students Uruguay, Kyrgyzstan, Tajikistan, and Estimates indicate that the econo-
in total graduates include Malawi, the Plurinational State of Bolivia are mies with the highest number of
Sierra Leone, Eritrea, and Kenya. the other top economies. machines in their agricultural lands

Table 7: Fertilizer consumption: Top five economies Table 8: Machinery in use:

Annex 4: Measuring Innovation in Agriculture and Food Systems

Top five economies
Tonnes of nutrients
Economy Share of world consumption (%) Economy per hectare of arable land Economy Machinery in use (number)

China 30.9 Qatar 12,111 China 10,802,121

India 13.4 Malaysia 2,064 India 5,960,636
United States of America 11.0 Hong Kong (China) 1,966 United States of America 4,351,616
Brazil 7.3 New Zealand 1,491 Japan 2,112,822
Indonesia 2.6 Bahrain 1,319 Poland 1,539,059

Data source: FAOstats, February 2017. Available at Data source: U.S. Department of Agriculture (USDA), International
Notes: Data refer to 2014. Fertilizers include nitrogen, phosphate, and potash. Agricultural Productivity Data, February 2017. Available at https://www.ers.

Table 9: Agriculture and food FDI net inflows: Top five economies

Economy Agri-food FDI (US$, millions) Economy Agriculture FDI (US$, millions) Economy Food FDI (US$, millions)

United Kingdom 19,186.1 China 1,112.1 United Kingdom 19,093.4

Italy 5,728.7 Brazil 426.7 Italy 5,746.7
Brazil 3,211.4 Ghana 348.8 Brazil 2,784.7
China 2,371.0 Argentina 259.4 Sweden 1,962.9
Sweden 1,962.9 Russian Federation 215.8 Turkey 1,700.5

Data source: FAOstats, February 2017. Available at

Notes: Agriculture includes agriculture, forestry, and fishing. Food includes food, beverages, and tobacco. Data refer to 2012; where data are missing, they refer to 2011, 2010, or 2009. FDI values are expressed in US$, 2005 prices.

are China, India, the USA, Japan, and weed removal. They are more Knowledge and technology outputs
and Poland, with China and India accurate and efficient than previous This section looks at agricultural
respectively accounting for 25% technologies such as satellite imagery labour productivity growth, agricul-
and 14% of all world agricultural and traditional tractors, allowing for ture and food exports, and patents in
machinery in use (see Table 8). productivity gains and cost savings. technological fields related to agri-
Italy, Thailand, France, Turkey, and The last indicator on business culture and food.20
Brazil also stand out in the use of sophistication reviewed in this annex The top five economies in terms
machinery in agriculture. is foreign direct investment (FDI) net of agricultural labour productiv-
Although these statistics are inf lows. Some agricultural and food ity growth are Slovenia, Bahrain,
extremely interesting, in the future, innovation systems prove to be well Luxembourg, Armenia, and Belgium
metrics on the use of drones and integrated in international knowl- (see Table10). Others that stand out
other autonomous vehicles might edge networks, receiving consider- include Bosnia and Herzegovina,
also be useful in assessing the inno- able FDI. The United Kingdom, Senegal, and Morocco.
vativeness of agriculture and food Italy, Brazil, China, and Sweden are Data on agricultural exports are
innovation systems. According to the top five recipients of FDI inf lows widely available through the UN
recent estimates, the market for in food and agriculture, driven by Comtrade database, which covers

drone-powered solutions in agri- FDI in food processing (except for almost all economies in the world
culture is US$32.4 billion25% of China). Ghana, Argentina, and the and allows for a highly disaggregated
the total drone application market.19 Russian Federation are among the analysis. According to these data, a
Drones and robots can be inte- top five FDI recipients in the agri- mix of high- and middle-income
grated at every stage of the produc- culture sector, while Turkey is the economies are among the top five
tion cycle: they can be used for soil fifth FDI recipient in food process- exporters of agricultural and food
analysis, seed planting, spraying, ing (see Table9). products. The USA leads this ranking,

Table 10: Agricultural labour productivity Table 11: Agriculture and food exports: development of new plant varieties
Annex 4: Measuring Innovation in Agriculture and Food Systems

growth: Top five economies Top five economies is often tacit and difficult to appro-
Growth rate of agriculture value Share of agriculture and food
priate.21 Innovation in plant varieties
Economy added per worker Economy exports (%) tends to be cumulative, meaning that
Slovenia 34.6 United States of America 10.2 prior knowledge is needed to come up
Bahrain 29.2 Netherlands 6.4 with new innovations. The adoption
Luxembourg 19.9 Germany 5.8 of new plant varieties also depends
Armenia 16.6 Brazil 5.4 on the efforts to adapt innovation
Belgium 15.8 China 5.0 developed elsewhere to breed locally
Data source: World Banks World Development Indicators, February 2017, Data source: UN Comtrade Database, February 2017. Available at https://
suitable varieties.22 These character-
available at istics may make intellectual property
Note: Data refer to 2-digit commodities codes, and include commodities protection of new plant varieties a
Note: Data refer to agriculture value added per worker (constant 2010 US$). from 01 to 24.
critical issue. The legislation on plant
variety protection is increasingly
being adopted in low-, middle-, and
Table 12: PCT applications in agriculture Table 13: Plant variety applications:
high-income countries.23 WIPO data
and food: Top five economies Top five economies
show that, since the early 2000s,
Economy Total applications Economy Total applications
plant variety application filings grew
United States of America 4,821 Netherlands 2,720 rapidly, with middle-income econo-
Japan 2,142 China 2,100 miesespecially in Asiacontribut-
China 1,418 United States of America 2,027 ing to global figures more and more
Germany 948 France 1,038 frequently.24
Korea, Rep. 798 Germany 942 Top filers of plant variety appli-
Data source: WIPO Statistics Database, May 2017. Data source: WIPO, 2016. cations are the Netherlands, China,
Note: Data refer to 2016. and the USA, followed by France
and Germany (see Table13). Other
important applicants are Japan, Korea,
the Russian Federation, Ukraine, and
accounting for 10% of total world developing economies to import Australia.
agri-food exports. The Netherlands, cheaper grains and grow high-yield
Germany, Brazil, and China fol- seed varieties, which were responsive
low with shares of between 6% and to fertilizers and resistant to diseases Creative outputs