Beruflich Dokumente
Kultur Dokumente
'
..
. .
Contents
' .
List ofFigures
List of Tables
Notes on Contributors
...........................................................................
PART I. CONCEPTUAL AND THEORETICAL ISSUES
............................................................................
1. Introduction: Democratic Accountability in Latin America
Scott Mainwaring
. ...
3. Horizontal Accountability: Concepts and.Conflicts .:
1
Charles D. Kenney .•. , .--··
............................................................................
PART II. LEGISLATURES, EXECUTIVES, AND
OVERSIGHT AGENCIES
............................................................................
4. The Accountability Deficit in Latin America
Erika Moreno, Brian F. Crisp anT·-~- .
.!!9:_tf.!!:...'!!!-'Boberg Sh1ffi.ar:!:
5. Legislative Oversight: Interests and Institutions
in the United States and Argentina
Scott Morgenstern. and Luigi Manzetti
Index 333
• • • • • • • • • • • • • • • • • • •• • •••••••••••••••••••••••••••••••••••
1.1. Conceptualizing political accountability 17 Rosangela Batista Cavalcanti is a researcher at the Institute of
A 4.1. Constitutional provisions for separation of Social, Economic and Political Studies (IDESP) in Sao Paulo. A
origin and survival: Supreme Courts in Latin America 119 Ph.D. candidate in Political Science at the Universidade de Campi
A 4.2. Constitutional provisions for separation of nas (UNICAMP), she is the author of the book Cidadania e Acesso
origin and survival: Constitutional tribunals in a Justice (Sumare, 1999) as well as several articles in journals and
Latin America 121 collected volumes.
A 4.3. Constitutional provisions for separation of origin
and survival: Attorneys General in Latin America 122 Brian F. Crisp received his Ph.D. in political science from the Uni
A 4.4. Constitutional provisions for separation of versity of Michigan in 1992 and is currently Associate Professor
origin and survival: Prosecutors General and Director of Graduate Studies in the Department of Political
in Latin America 123 Science at the University of Arizona. His work has appeared in the
A 4.5. Constitutional provisions for separation of American Journal of Political Science International Studies Quar
origin and survival: defenders for human terly, Legislative Studies Quarterly, Studies in Comparative Inter
rights in Latin America 125 national Development, Journal of Interamerican Studies and
A 4.6. Constitutional provisions for separation of Worki Affairs, and Latin American Research Review, as well as
origin and survival: Controllers General several edited volumes. His book Democratic Institutional Design:
in Latin America 126 The Powers and Incentives of Venezuelan Politicians and Interest
6.1. CPIS by Government and Period-1946-64 Groups was published by Stanford University Press {2000).
and 1988-99 Democracies 178
6.2. Rates of Conclusion and Distribution of CPIS Michael Dodson is Professor of Political Science at Texas Christian
Proposed and Concluded in the Lower House University. He has held two National Endowment for the Human
According to the Proponent's Party Affiliation 181 ities fellowships and was a Fulbright Senior Lecturer in Britain.
6.3. Committee's Oversight Activities-1989-99 182 His recent publications include: 'Re-inventing the Rule of Law:
A 6.1. Presidents, coalitions, and parties' seats Human Rights in El Salvador', with Donald W. Jackson, Democra
in the Lower House 193 tization 4, 1 (Winter 1997); 'Human Rights and the Salvadoran
9.1. Constraints on the executive 274 Judiciary: The Competing Values of Independence and Account
9.2. Homicide rates per 100,000 inhabitants 277 ability', with Donald W. Jackson and Laura O'Shaughnessy, The
9.3. Change of constitutional rules regarding size, International Journal of Human Rights l, 4 (Winter 1997); and
appointments, terms and dismissal of the 'Protecting Human Rights: The Legitimacy of Judicial System
Supreme Court 284 Reforms in El Salvador', with Donald W. Jackson, The Bulletin of
9.4. Number of justices appointed by president 288 Latin American Research 18, 4 (Wmter 1999).
9.5. Turnover rates of Justices of the Supreme Court 289
9.6. Number of actions and controversies filed Argelina Cheibub Figueiredo is ProfessorLiure Docente at the Uni
at the Supreme Court 297 versity of Campinas, Brazil, and senior researcher at the Centro
9. 7. In your opinion, which is the principal Brasileiro de Analise e Planejamento (CEBRAP). She received her
problem in Mexico City? 298 PhD from the University of Chicago in 1987. She is the author of
Xll Notes on Contributors Notes on Contributors xiii
Democracia o reformas? Altemativas politicos a crise democratica Science Association meeting. She is currently working on two
(Paz e Te1Ta, 1993) and, with Fernando Limongi, Executivo e legie books, one on the political economy of the Mexican democratiza
latioo na nova ordem constitucional (FGV/FAPESP, 1999) and tion process, and the other on the Mexican Supreme Court.
'Presidential Power, Legislative Organization, and Party Behavior
in Brazil', Comparative Politics (2, 2000). Scott Mainwaring is Eugene Conley Professor of Political Science
at the University of Notre Dame. Among his most recent books are
Donald Jackson has been a faculty member of the Texas Christian Rethinking Party Systems in the Third Wave of Democratization:
University (TCU) since 1975 and is now the Herman Brown Profes The Case of Brazil (Stanford University Press, 1999); Presidential
sor of Political Science. Before joining TCU, he served as a -Iudicial ism and Democracy in Latin America (Cambridge University
Fellow at the Supreme Court of the United States. His research has Press, 1997, co-edited); and Building Democratic Institutions:
been on various aspects of the intersection between law and polit Party Systems in Latin America (Stanford University Press, 1995,
ics. His book, The United Kingdom Confronts the European Con co-edited).
vention on Human. Rights (1997), represents his interest in
international human rights enforcement. Other publications in Luigi ~Jis an Associate Professor in the Political Science
clude Presidential Leadership and Civil Rights Policy (1995), co Department at Southern Methodist University. His scholarly work
edited with James Riddlesperger, Jr., and Even the Children of has dealt with regional integration in the Southern Cone, privat
Strangers: Equality under the US Constitution (1992). Professor ization and deregulation policies in Latin America, and corruption.
Jackson's current research on democratic transitions in Central He is the author of Institutions, Parties and Coalitions in Argen
America, focusing especially on the rule of law, has led to a series of tine Politics (Pittsburgh University Press, 1993) and Privatization
articles co-written with Michael Dodson. South American Style (Oxford University Press, 1999) and the
editor of Regulatory Policy in Latin America: Post-Privatization
Charles D. Kenney received his Ph.D. from the University of Notre Realities (North-South Center Press at the University of Miami,
Dame in 1998 and is an Assistant Professor of Comparative and 2000).
Latin American Politics at the University of Oklahoma. His pri
mary areas of interest include democratization, constitutional Erika Moreno received her Ph.D. in political science from the Uni
design, electoral and party systems, and Peruvian politics. He versity of Arizona in 2001. She is an Assistant Professor of Polit
lived in Peru from 1978-79 and 1984-91 and was a Fulbright ical Science at the University of Iowa where she teaches and
Fellow at the Pontificia Universidad Catolica del Peru from Febru conducts research on democratic political institutions and party
ary to July 2000. He was an international electoral observer for systems across Latin America.
the 2000 and 2001 Peruvian elections and is the chair of the Peru
Section of the Latin American Studies Association. His book, 'Pol Scott Morgenstern is Assistant Professor of Political Science at
itics in Peru: Fujimori's Coup and the Dilemma of Democracy in Duke University. He is the author of Patterns of Legislative Polit
Latin America', will be published by the University of Notre Dame ics: An Exploration of Roll Call Voting in the United States and
Press. Latin America (forthcoming) and coeditor of and contributor to
Legislative Politics in Latin America (Cambridge University Press,
Beatriz Magaloni is Assistant Professor of Political Science at 2002). He has published articles on party politics and elections in
Stanford University. She is co-author, with Alberto Diaz-Cayeros, Party Politics, Comparative Politics, The Journal of Politics, Legis
of 'Party Dominance and the Logic of Electoral Design in the Mex lative Studies Quarterly, and several Latin American journals.
ican Transition to Democracy' (Journal of Theoretical Politics 13, 3; . .
2001). She won the Gabriel Almond Award for the best dissertation Guillermo O'Donnell is former Academic Director of the Kellogg
in comparative politics (1998) and is co-winner, with Alberto Diaz Institute and Helen Kellogg Professor of Government and Inter
Cayeros and Barry Weingast, of the 2001 prize for the best paper national Studies at Notre Dame. He was the first Director of
in comparative politics presented at the 96th American Political CEDES (Buenos Aires), a visiting professor at the Universities of
xiv Notes on Contributors Notes on Contributors xv
Michigan (Ann Arbor), California (Berkeley), and Sao Paulo, and a He has co-authored or co-edited five books on the topic: Seats and
researcher of IUPERJ (Rio de Janeiro) and CEBRAP (Sao Paulo). Votes: The Effects and Determinants of Electoral Systems (Yale,
He is a Fellow of the American Academy of Arts and Sciences. He 1989), Presidents and Assemblies: Constitutional Design and
has published extensively on comparative politics, authoritarian Electoral Dynamics (Cambridge, 1992), Presidenticdiem and Dem
ism, and democratization in several languages. Recent publica ocracy in Latin America (Cambridge, 1997), Executive Decree Au
tions include Counterpoints: Selected Essays on Authoritarianism thority (Cambridge, 1997), and Mixed-Member Electoral Systems:
and Democratization; The (Un)Rule of Law and the Underprivil The Best of Both Worlds? (Oxford, 2001). He has also authored
eged in Latin America, coedited with Juan Mendez and Paulo Ser numerous articles on electoral systems, presidentialism, and Co
gio Pinheiro; and Poverty and Inequality in Latin America, · co lombian politics in journals such as the American Political Science
edited with Victor Tokman (all in the Kellogg Institute series with Review, Comparative Political Studies, Constitutional Political
the University of Notre Dame Press, 1999, 1999, and 1998). Economy, and Electoral Studies.
Enrique Peruzzotti is a professor in the Department of Political Catalina Smulovitz, an Argentine political scientist, is Professor at
Science and Government at the Torcuato Di Tella University in the Torcuato Di Tella University in Buenos Aires and a researcher
Buenos Aires. He has published articles on democratic theory and at the Consejo Nacional de lnvestigaciones Cientificas y Tecnol6gi
democratization in Citizenship Studies, The Journal of Democracy, cas (CONICET). She is the author of Oposici6n y gobierno: Los
Constellations An International Journal of Critical and Demo aiios de Erondizi (Centro Editor de America Latina, 1988) and co
cratic Theory, Thesis Eleven, Revista Mexicana de Sociologia, and editor with Enrique Peruzzotti of volume Contolando la Politica.
the Journal of Latin American Studies. With Catalina Smulovitz Ciudadanos y Medios en las Nueuas Dernocracias Latinoamerica
he is conducting a research project funded by the Ford Founda nos (Temas, 2002). Her articles include 'Societal Accountability in
tion, 'The Emergence of a Politics of Societal Accountability in the Latin America' in The Journal of Democracy (Oct. 2000) with Enri
New Latin American Democracies'. He co-edited (with Catalina que Peruzzotti; 'The Discovery of The Law: Political Consequences
Smulovitz) the volume Contolando la Politica. Ciudadanos y Med in the Argentine Experience', in Garth Brian and Yves Dezalay,
ios en las Nuevas Democracias Latinoamericanas (Temas, 2002). eds., Global Prescriptions (University of Michigan Press, 2002);
His current research analyses the emergence of a new form of and 'Adjusting the Armed Forces to Democracy. Successes, Fail
politicization in Latin American civil societies organized around ures and Ambiguities of the Southern Cone Experiences', (co
demands for accountable government. authored with Carlos Acuna) in Elizabeth Jelin and Eric Hersh
berg, eds., Constructing Democracy: Human Rights, Citizenship
Maria Tereza Sadek is a professor in the Political Science Depart and Society in Latin America (Westview, 1996).
ment at the U niversidade de Sao Paulo and a senior researcher at
the lnstituto de Estudos Economicos, Sociais e Politicos de Sao Christopher Welna is Associate Director of the Kellogg Institute,
Paulo (ID ESP). She has published a number of books and contrib Concurrent Assistant Professor :in the Department of Political
uted articles on the justice system to Brazilian and foreign jour Science and Director of the Latin American Studies Program at
nals and periodicals. Her books include O Judiciario em Debate the University of Notre Dame. He specializes in Latin American
(Sumare, 1995), 0 Ministerio Ptiblico e a Justica no Brasil politics and public policy reform, especially in Mexico, Brazil,
(Sumare, 1997), 0 Ministerio Publico Federal e a Administraciio Colombia and Cuba. He has held fellowships from the Social
da Justica no Brasil (co-authored, Sumare, 1998), Acesso a Justica Science Research Council and the Mellon and Thomas J. Watson
(Konrad Adenauer, 2001), and Reforma do Judicidrio (Konrad Foundations. Before coming to Notre Dame, he taught at Duke
Adenauer, 2002). University and worked for the Ford Foundation, the US Depart
ment of State, and the UN Institute for Training and Research. He
Matthew Soberg Shugart is Professor at the Graduate School of holds a Ph.D. in political science from Duke University.
International Relations and Pacific Studies, University of Califor
nia, San Diego. He is a leading specialist on political institutions.
.
.
"':.
..
. .
?r·
...............................
Introduction: Democratic
Accountability in Latin America*
Scott Mainwaring
This volume on democratic accountability addresses one of the
burning issues on the agenda of policy-makers and citizens in con
temporary Latin America. Collectively, we hope that the volume
enhances understanding of three ~.fil'is§_Uef?~irst, it enriches
understanding of the state of<non-electoral)forms of democratic
accountability in contemporary Latin America. What are some of
the major shortcomings in democratic accountability? How can
they be addressed? What are some of the major innovations in the
ef(orts to enhance democratic accountability?
·~ A.. second _contribution of the volume is conceptual. Accountabil
ity is a key concept in the social sciences, yet its meaning varies
widely from one author to the next. The authors in this volume,
especially in tne first four-chapters,explicitly debate how best to
define and delimit conce.11. Although. _the we-caniiof"ciami"co~en.:
sus in our understanding of this concept, we believe that the direct
confrontation of ideas will advance the debate. . ·. , \JtD,,; c..4·<'.~
Yinally, t!~e, volume ~~-~-~~?!=:X.:-S -~~~!'~~~~.~-9f{gf~e.inter~c~io~
between various mechanisms and institutions of accountability,
Many of the authors address how electorat accountability . (i.e., the/."
accountability of elected officials to voters) interacts with other
forms of accountability in which state agencies oversee and sanction
public ofru:ials."-Tbevolume provides a fairly extensive treatment
of this important but hitherto underexplored interaction.
~~~~1:1?.~~i!!tx:.~~~ ~-~-~~~~c! 8.:5. ~~-~- ?.~.~h:~.~~:r _is.~~~?. ~~~.~st
transition period of Latin American politics. In the l.:Y80s,; the
'" '" ,.,. .,....,,.,. .,.,, ,• ••-•' ' ,•• •· • •• "''' •,• • ,,.,.,,.,.., •••' ,.... ,,,.,., ... .,' '" I <'
',.
~iihfu'i\;&~.;~~~;,.:::l-:~:~(..;,tors can provide{§fiiicaf) ti~count~bili_ty_in_thl§-..Y.<2!~1P.:~- (and beyond). All five involve the
accountability. Firstl~lected •· ...... · • · •• · boundaries of the concept: how broad or narrow should it be? ·
,Ji't,~~z at )east jn _~~sesm whicli r~.:~~cti_~g_is ..~lowe_c! Second7 The most fundamental question is whether all activities that
1_ ·~y ~ta~-~~~~!e~~~or~:1ly ~~arged_wi~ ...?.~~eeing andf._2r involve holding public officials responsible for their discharge of
sanctioning public officials ana bureaucracies; I refer to these rela- duties should be included under the very broad rubric of account-
ability. Ma°:L~c~o!~!:.~.... ~.~~~P.~-~~~1-!:-.. ~.J>.i:Q~<!-.~-~!!_~r~~!l_di~ (Day
-~~ii inlif il~~ttJ[fu;~!i1~:fu~:o~~ti~::~=~:~~l;:1t~~::~~ and Klein 1987; Diamond, Plattner, and Schedler 1999; Fox and
agencies that are created to monitor public officials and bureaucra Brown 1998: 12; Merritt forthcoming; Moncrieffe 1998; Paul 1992).
cies with an eye toward possible wrong doing, the legal system For example, Paul (1992: 1047) states that 'Accountability means
when it investigates the possible misdeeds of public officials and holding individuals and o_maniz~tion~~ !~~IiQ'iiS.!~!~~::ror-.th~i.i.·-P.~~=~
bureaucracies, the congress when it initiates a hearing against any
public official, and the assembly in a parliamentary system when it ~%~~:,· ~~H~~Iiii~~~1Ii~~~sus~~-~-~--·il:~~1i~1a:~~
--- - -···· --·-··-- - ~-·-··· . , . . p _ -....... . ,iY, ,.•-.. -· - - _, " ,
makes a political judgment about whether to remove a minister or a concerned with public services to ensure a desired level and type
cabinet] A third category analyzed by Smulovitz and Peruzzotti in of.. perlormance.' 'bay and Klein (1987: 5)° provide anothe;-broa:~f
this volume, $QC.!~~J_<>.Y~!:~!ght!..f~;Is .o:utside the bounds of my defin definition: 'Toaccount is to answer for the discharge of a dµty or
ition of accountability, yet with important 'interactions. with-the for conduct.1tTs--to-provide reck.on_ing. It is to give a satisfactory
a·
formafnetwork-of'Institutions-ofaccouiitability. Intrastate ·account-. i~asori· f(?r l.!_)~--~~---3-:~~gi~~-cige ~SPo.Jl§!~!9.JY .f9.f.
~~!·t~.~!J>.t~J!l.
ability and itsinte:r'iiclions.with' elech:>ral"a'ccountability and soci one's actions.' D. Dunn (1999), Keohane (2002), and Oakerson
etal oversight .we the primary focus of this book. (1989) stipulate that accountability also necessarily involves the
~.P.a~!~~{.~~~!:~.~ .. ~P.~~~ . ~-~£g9.~~--~E-.P.~~liC.. ~ffi~iajs: Yet their
;b~:::~~rfu~c~rt~~ll~cy?fil~f~Z=··~e~:c1~:a~~ii~~~dl·~Hf: definitions remain broad by virtue of including a wide range of
.. ·--· -Both
-cials. ,_ ..
--·..·-·-···-of-- these broad categories should be accountable
_,.,.~~ -···· . . ..
-· .
···to dif-
···--··-··---
non-institutionalized kinds of answerability and sanctions. For
ferent kinds of actors : · lected public officials are accountable example, Keohane mentions reputational accountability, by which
first and" foremost to th - citizens whose" votes. put. them. into office. an actor informally demand answerability and impose reputa-
can
The ways in which elected politicians. are accouiitabletlirougli tional costs ona public official. . .
the vote have been analysed at length by political scientists (Fere '.fhe ~uthors in th.is volume_ wh~_ex,elici!!Y.:_~~~..~.P9.~~P.~~l~.~§!~
john 1999; Fiorina 1981; Key 1966; Maravall 1999; Powell Jr. issue (O'Donnell, Kenney, and Moreno et al.Fall advocate a conceP-t
2000; Przeworski et al. 1999). Although this relationship between tiiafis'.mo~~ Eouiiaea'."tiian
---.....,...,.,~ -·--·--Lf4'i-!!;E_i. ~_,,_ ....... "·-~·--··---2 ... d~ties.in
suu·i--e-·-ramin·"-one's··--ublic
,_~·--····"·--·
voters and representatives is not the primary subject of this ..~espotis.e ~-~.9-..£~~a_req~e~.~:,;I agree that a more hounded concept
volume, it is central to demo~ratic J?Olitics. is . de~i.r:able . . and '
. . that it is important
. to specify what forms of
In democracies, man~n-elected ?.:ffici~§l!;~~-~s~,,P,!"~~~-f!.:.~!X
accountable to _the executive or the legislature. T es~ non-elec~ ' 'r ....... ·, • , • .. • . .. • • ~
f}iiowever, they delimit the concept for different reasons and in different ,vay~/ • • •
~iJicials ~e in ~p.e :firs~J!lstan,ye ag~nts of th~ p~~pal~- who.~tt. O'Donnell and ~~ilney limit the notion pfh_~rizpE._t:?_t!lSCOl!_.ll~i!!:~.J~!ega!_~~~l
point them. As with the relationship between voters and elected gi-essi<i'ns. Kenney a.n<l Moreno et al. limit 'accountability' lo cases in which th~'
ooli!icians, Iiowever, the relationship between non-elected officials . -~geri~, of a:c_<&~~t.~~IgY..-~~~ _t,l~.c.c.t1~.~E~~'.Y. · t<!..~~~P-~~?.-~~~~i<?!£~Moreno et ·arliroit
'a,ccoi,.ihtabjlitt_fi.?_grinci_P,~1::~.G'e_~:~}~.liti~~ . · ·. ·
···1
ey~_r_y action by Vl(~cp.-~~-~~ __;ct?! requests or. demands -~~W.:.:!.'.1~I~: ing-arenot limite"ci to perceptions about the legality of a public
ity of a public official is considered 'accountability'. Answerability · official's or agency's actions. Public officials and agencies must pro
"combmes·~ many" "different.. agents."and~·actions," "'and .1t-iilakes. the vide political as well as juridical accountings of the discharge of ''
~?~~~P.~~LhQ.l.:!,Ild.arj_es_filurh My demarcation provides for much their public duties. One of the classic forms of intrastate account c;
greater specificity and clarity than the excessively diffuse notions ~!>.i!!tr_!~--<:~b~et_(Laver and Shepsle 1999; Strem 2000) ~'=1:. ;aj~~:. .
'I
sanctions would allow almost any actor to be an agent of account Is accountability Yes (Kenny, this volume; No (Day and Klein 1987;
ability. In contrast, with the exception of Smulovitz and Peruz- ' limited to formal Mainwaring; Moreno et Fox and Brown 1998;
zotti, the authors in this volume focus on a narrower range of (legalized) al., this volume) Keohane 2002; Merritt
authority to oversee forthcoming;Moncrieffe
actors that have formalized responsibilities to oversee public offi and/or sanction 1998)
cials. Within this fundamental agreement about a bounded set of . , .. ··- ,.. -· - ····- .. .
Is direct formally Yes (Kenney, this volume; No (Mainwaring;
actors that can constitute agents of accountability, the authors in ascribed(o-<"4v.,w.>') Moreno et al., this volume) O'Donnell, this volume;
this volume nevertheless have some differences of opinion that sanctioning po,v~ a Schedler 1999; all authors
stem directly from the previous disagreements. Because of their necessary named in the cell
definition of accountability, Moreno et al. limit these actors to component of immediately above)
accountability
voters, to whom elected representatives are accountable; polit -·-· '" . ' .
icians, to whom some bureaucracies are accountable; parliaments, Are accountability Yes (Moreno et al., this No (Kenney, this volume;
relationships volume, Elster 1999; Laver Mainwaring, Merritt
to which cabinets and ministers are responsible in parliamentary limited to and Sbepsle 1999) forthcoming; O'Donnell,
democracies; and other principals in principal-agentrelationships. principal-agent this volume)
They explicitly exclude oversight agencies and institutions that relationships?
--··· ~ . . .' . . ... ,, . ,, . - " . .
are independent from each other, Kenney also excludes oversight Agents of Principals in principal All actors tl1at formally
agencies but includes actors that have sanctioningpower (the pri accountability agent relationships oversee and/or sanction
mary example is the judiciary). O'Donnell includes not only princi (Moreno et al., this volume public officials (Kenney,
pals in principal-agent relationships, but also oversight agencies ·Elster-1999· Laver and this volume; Mainwaring,
'
Shepsle 1999) this volume; O'Donnell,
and the judiciary as agents of horizontal accountability-although
this volume)
all three kinds of actors only when the actor being investigated or
sanctioned has committed a legal infraction. I share his list of
agents of accountability,but as noted above disagree that they are
acting as agents of accountability only in cases of legal transgres are a subset of accountability. Political disagreement per se be
sions. tween the legislature and the executive does not necessarily consti
These arenas of agreement and disagreement about the concept tute a relationship of accountability. For example, if the legislature
of accountability are summarized in Table 1.1. In the final row, refuses to approve an executive-sponsored bill, this does not intrin
'Agents of Accountability', the scholars who do not limit account sically constitute a relationship of accountability because it does
ability to formal authority to oversee and/or sanction public not involve an element of answerability or sanctioning. The distinc
officials have a much less restricted conception than the two possi tion between accountability and checks and balances, however, is
bilities shown in Table 1.1. sometimes a fine one; checks and balances can constitute mechan
Kenney (this volume) raises one other important conceptual isms of accountability. The key issue according to the definition
issue: the relationship between horizontal accountability and proposed here is whether an actor has the constitutionaJJlegal cap
checks and balances. He properly argues that these two concepts acity.. to request an -accounting of a public official's (or agency's)
should not be confoundedor conflated; not all checks and balances discharge of duties or to impose: sanctions on that official. If so, it is
18 Scott Mainwaring· Introduction: Democratic Accountability in Latin America 19
a relationship of accountability. When the legislature creates a O'Donnell (1999c) includes the monitoring by NGOs over public
commission to investigate executive actions, it is a form of over officials as an element of vertical accountability, but NGOs do not
sight and creates a demand for answerability-hence constitutes a vertically control public officials in any way.
mechanism of accountability (see Figueiredo's chapter in this Second, O'Donnell's distinction conflates t,vo c:liff~rent issues
volume). In a similar vein, when the legislature charges a commit t)lat should not be confll!ted~_ij_~.Y~icc;il . metaphor that conve~
tee thereof to oversee an executive agency, a relationship of ac- pn~g~s _gf iI!Q~I!_enden~_(hoti_zont?.litl)_and h:1-e:!ar~hy (verticali~),
countability exists. · and the location of the agent of accountability (state versus soci
A skeptic might wonder whether the concept of accountability etal~~§l-1...l~~P~i:vely).O.;Donrie1f"{f99~c) ·-statecftfiaf vertical
can be rescued in light of the conceptual muddle that is reflected accountability refers to voter~ anq__)?9cie~_~l ._qrgf:lnizations while
in Table 1.1. I believe that the answer is affirmative. Accountabil horizontal accountability refers to state and re~e 8:_g:encies. In
ity is too substantively important a subject in the social sciences his chapter in this volume, Kenney also views the distinction be
and in politics to abandon. Nonetheless, the vast differences that tween horizontal and vertical accountability as resting on whether
are synthesized in Table 1.1 indicate the need for more concerted the agent of accountability is a state or societal actor. Yet the cor
theorizing about how to define and delimit the concept. Hopefully, respondence between horizontal and vertical relationships, on the
this volume will advance thinking toward accomplishing this ob- one hand, and intrastate and electoral accountability (arid societal
jective. oversight), on the other, is not perfect.
Principal-agent relationships can be seen as vertical: the princi
pal commissions and can 'fire' the agent (see Moreno et al., this
Distinguishing Among Types of Accountability volume). The accountability of a cabinet to the assembly in a par
liamentary system is clearly a case of what I call intrastate ac
~O'Donnell 1994, 1999a, 1999b,_tbis ..~~e) has be_en a pioneer in countability, and it is also a principal-agent relationship. By
~011:..~~~a Ji zi!1..B:..di2~P.:~P.:Q.I!S~~2P.:K.ffe.~~r~~~-kinds of accountabil:. suggesting that accountability is horizontal when state actors must
ity: He dev~lope,d~ rtj,g;;tj~n between_v•JE£a! and_horizontaL answer to other state actors, Kenney and O'Donnell would consider
accountab~lity,_l!y. '{~~l accountability:pp9.Il!l.~ll meant the ac:.. this an example of horizontal accountability. Yet in constitutional
co11:Q.!!l.h..~ir.2f sta gents ~-~itiz~ns ~d_!~civil society., Elec- terms, it is a vertical relationship; the cabinet serves at the dispos
important, 'arguably main of vertical ition of the assembly. In a similar vein, relationships between the
_.
tions,. are -···an
____ ··- _..,,,,_,._ _.-, -·-indeed
..--~---~ ·- ·--··· -··--- ~--the
-- ·····-- ·----·--·facet
···--·-·-----· - ..···----·· . ·
accountability' but the notion of vertical accountability also in- executive or the assembly and some bureaucracies are vertical in
-·---·-··----2-----·····-·· -- ·~··--·---..-----·-----···-·-·--·-·-"'•'"-·------·
.C.lY9.:~.[-~S.~Q~§.PX....~iYP._~9.gi~_t.Y. .. ~D.:~.. ~P-:~-~~ili.~ t<?..-~!~os_~-~-~par_ently the sense that the executive or assembly creates the bureaucracy
~.Q.P--gful.~ct.~_.Qf.tl!.~tP.!i.Pl.!~Y-~t_!t.2.!it.!~~~.Jf~~9-~--~-9t. and formally has control over it, yet at the same tirne it is an
O'Donnell's distinction has set the agenda for contemporary de- intrastate relationship. In short, some intrastate relationships are
bates about non-electoral forms of accountability, but it is ques vertical, at least in formal terms. For this reason, it is problematic
tionable that his spatialmetaphors-provide the best way of to equate intrastate accountability with horizontal accountability,
differentiating among different kinds of accountability. His termin and more broadly it is problematic to combine, as Kenney and
ology has two problems. First, the physical metaphor conjured by O'Donnell do, the vertical/horizontal distinction with a distinction
the notion of vertical accountability is misleading except in princi based on the location (in society versus in the state) of the agent of
pal-agent relationships. Vertical accountability expresses 8.J!. accountability.
ima~_o.f asymmetry ofpow~r._that is, a hie;t"archical relationship~
As Moreno et al. argue, the vertical metaphor can reasonably be _la~~~~!~~tc~~::?of~e~~-~~~~~~o:I)~.i~ii~~;iit8~~i~~~
extended to all principal-agent relationships, even to those in ~lass~~h~~~~...?..f-.~~~£.~EP.!~t>~M!.Y..~~-. !!.~e.lYi.~~---~-~~-<?.1:1:.~~~.~li~Y.._~f..~~~~
which the principal.-:is'm-mostrespects vastly weaker than the ~---~--~~!!!~nister~!£..J~~!E!!Y.1 e~t_~__ }.!!.,_l?~Ji ~!!':~nt~.2Y~,Y;~~
agent. For example, they see voters as principals and elected offi According to. his definition, only if the parliament removed a min-
cials as agents. The 'vertical' image, however, is inappropriate tQ ister or cabinet because of legal transgressions would this be an
describe rel~tiqnship~-~hjit d~__no.!_ have a hierar~hical componen_!;_. instance of horizontal accountability Because one state actor (the
.)
..
20 Scott Mainwaring Introduction: Democratic Accountability in Latin America 21
cabinet or minister) must answer to another state actor (the as
sembly), it is not a case of vertical accountability according to his Interactions Between Electoral and
definition. Intrastate Accountability
Although O'Donnell's notions of horizontal and vertical account
ability are problematic, it is useful to differentiate among different A second key contribution of this volume is an effort, thanks above
kinds of accountability. It is more fruitful to conceptualize these all to Moreno et al., to advance the debate about interactions
distinctions strictly according to the agent of accountability.1 For among electoral accountability, intrastate accountability, and soci
the sake of parsimony, I propose· a distinction between two main etal oversight. Rarely have these interactions been discussed in as·
kinds of political accountability: electoral accountability or (syn detailed a manner as they are here-especially by Magaloni, Mor
onymously) accountability to voters and intrastate accountability. eno et al., O'Donnell, and Smulovitz and Peruzzotti. On this point,
This distinction parallels O'Donnell's but without the connotations too, we hope to advance the debate not through consensus but
of hierarchy and independence suggested by his metaphor, and rather through explicit disagreement.
without conflating the agent of accountability (state versus soci Moreno et al. argue that the linkage between voters and elected
etal) with the nature of the relationship (horizontal versus verti representations sets the tone for all other accountability rela
cal). tionships. In their argument, by getting electoral delegation and
Intrastate accountability can be usefully subdivided into three accountability right, politicians will address the most serious defi
different kinds of relationships. First, in principal-agent relation ciencies in accountability. Tu quote from their chapter, 'The deficit
ships, a principal commissions an agent to perform some function of accountability lies in faulty vertical accountability-legislators
and has ultimate control over that agent. Bureaucracies created who do not represent the values and preferences of the broad citi
and ultimately controlled by the executive or legislature are prime zenry. If the accountability of elected officials were working as
examples. Moreno et al. restrict their understanding of 'account intended-such that voters could and would punish misdeeds
ability' to this first subcategory. Second, the legal system and on separate agencies of superintendence would be unnecessary;" Al
occasions some other state actor (example: a legislature that is though none of the other authors take this argument as far, all
trying a president on impeachment hearings) can impose sanctions agree that electoral accountability is important for intrastate ac
on public officials accused of wrongdoing. These are not principal countability as well. But other contributors including O'Donnell
agent relationships, but rather can be thought of as 'sanctioning argue that electoral and intrastate accountability have strong
actors'. The most common sanctioning actors are within the justice mutual effects. By this argument, electoral accountability and
system. The congress becomes a sanctioning actor when it tries the delegation do less to resolve the problems of intrastate account
president on impeachment charges. Kenney and O'Donnell include ability than Moreno et al. argue. Conversely, by this argument, it
sanctioning actors as well as principal-agent relationships in their is possible to make greater gains in intrastate electoral account
rubric of horizontal accountability. Finally, oversight actors have ability than Moreno et al. argue even without touching electoral
responsibility for monitoring the behavior of state officials and rules. For example, Smulovitz and Peruzzotti argue that social
agencies. Kenney excludes oversight actors that do not have legally mobilization, especially in conjunction with media coverage, can
ascribed, direct sanctioning capacity, and O'Donnell includes them trigger improved intrastate accountability. Sadek and Cavalcanti
only when a legal transgression is at play. document the impact of Brazil's Public Prosecution, which has en
hanced intrastate accountability despite the lack of major electoral
and party reform since 1988.
Although the contribution of Moreno et al. is an innovative and
7 Moreno et al. also avoid distinguishing according to both the agent of account
important work, intrastate and electoral accountability are
ability and the verticality or horizontality of the relationship, but they advocate
the opposite way of resolving this tension. Rather than distinguishing between more mutually interactive than they indicate. They overestimate
different types of accountability on the basis of the agent of accountability, they the capacity to improve intrastate accountability by enhancing
distinguish between vertical (principal-agent) relationships and relationships of electoral- accountability pond understate the capacity to enhance
horizontal exchange. intrastate accountabihty without; changing electoral accountability.
22 Scott Mainwaring
It is questionable that the primary key to bett.er intrastate account Introduction: Democratic Accountability in Latin America 23
ability is balance in the continuum from a party centric to an indi argument, a it should be possible to enhance the effectiveness of
vidual centric electoral system for the national congress. They oversight agencies by changing the appointment procedures and
place great optimism in the capacity to effect change in the func lengthening the terms of office-a change that is in principle inde
tioning of intrastate accountability through electoral reform. pendent of changes in elect.oral accountability.
The 'accountability deficiency' can be seen as a product of feeble Moreno et al. are sanguine about the effects of electoral reform
sanctions. Whereas Moreno et al. locate the inefficacy of sanctions on intrastate accountability. Such optimism assumes that the
in electoral systems that set up inadequate mechanisms of ac connection between elected representatives and oversight and
countability between voters and elected representatives, perhaps sanctioning agencies within the state is fairly tight: intrastate ac
even more important are the failures of sanctions that should be countability (superintendence in the lexicon of Moreno et al.) will
imposed by the justice and penal systems for wrongdoings of public improve because elected politicians find it in their interest to make
officials. The justice system works reasonably well at the upper these institutions work better. This argument probably overstates
echelon in some countries (see Magaloni, this volume), but in most the extent to which members of congresses would effect improve
of Latin America it is notoriously bad at the local level, where most ments in the justice system and oversight agencies even if stronger
cases are tried. In much of the region, at the local level, the justice' mechanisms of electoral accountability gave these politicians a
system ('injustice' system frequently more appropriately captures bigger stake in national policy outcomes.
the reality) is still under the sway of powerful elites (Dodson and The argument also may be excessively sanguine about the effect
Jackson, this volume; Magaloni, this volume; Pasara 2002). In sev iveness of electoral accountability between voters and members of
eral countries, at the local level, drug barons with seemingly un national congresses in Latin America. Electoral accountability
limited resources to buy off and intimidate the police, judges, and offers politicians enormous opportunities to shirk (Ferejohn 1999;
witnesses have captured the justice system. When public officials Maravall 1999; Stokes 1999). Elections occur intermittently, usu
are confident that they can commit wrongdoings without facing ally every two to eight years, and nothing assures that elected rep
penal and legal sanctions, electoral reform is not likely to fix the resentatives will behave as voters would prefer between elections.
problem. Greatly compounding this problem is the inefficacy and Tremendous information asym.m.etries between elected officials and
frequent complicity of police forces in much of Latin America; thep; the average voter give the forme:r ample opportunities to behave
too, are frequently captured and paid off by local elites from drug with autonomy vis-a-vis the latter (Przeworski et al. 1999). As
barons to landowners and traditional oligarchs. Schumpeter (1950: 256-64) and Downs (1957) argued long ago,
With an important original argument and a wealth of new em the average voter cannot invest the considerable time needed to
pirical information, Moreno et al. show that there is great variance closely monitor representatives' actions. These information asym
in how independent one would expect oversight agencies to be on metries are larger in Latin America, where most voters have
the basis of institutional arrangements (Figure 4.3 and the Appen 1imited education and little information about politics, than in the
dix of their chapter). They argue that such agencies are likely to advanced industrial democracies.
be more independent from political pressures, and hence more ef Even if voters become disenchanted with a politician or a party
fective as mechanisms of oversight, if (a) they are not named dir in power, they may not believe that other options are better. The
ectly and exclusively by the legislature and (b) they have long 'supply' of politicians or parties may appear to voters to be oligop
terms of office, such that they need not depend on the ongoing olistic or quasi-monopolistic. This feeling that all politicians or
support of politicians for remaining in office. They conclude that Parties are the same diroinishes the extent to which elections
many oversight agencies are insufficiently insulated from political serve as instruments of accountability: if voters do not believe
pressures to be effective and that the only effective way to enhance that the competition is any better, they are less likely to punish
accountability is the vertical (electoral) linkage between voters
and representatives. But this argument overlooks the possibility of 8
trying to boost the effectiveness of oversight agencies by making· In a region where many institutional reforms have produced less than their
supporters hoped for (P:isara 2002), we cannot be sure that such reforms of insti
them more independent of political pressures. By the logic of their tutional arrangements would renlly produce the desired effect of enhancing the
independence of oversight Offoncies.
•
..
24 Scott Mainwaring Introduction: Democratic Accountability in Latin America 25
incumbents by voting them out of office. In sum, as Ferejohn (1999: contemporary El Salvador and Guatemala (see also Pasara 2002).
137) succinctly points out, 'Electoral punishment ... is a fairly These deficiencies make it difficult for these judiciaries to serve
blunt instrument, and incumbent officials will be, at best, only either of the two functions central to O'Donnell's notion of horizon
moderately responsive to public wishes.' (See also J. Dunn 1999: tal accountability: preventing corruption and preventing improper
338-9; Fearon 1999; Manin et al. 1999; Maravall 1999.) state encroachments.
This book does not purport to resolve the important issues re Other authors in this volume, however, point to innovations in
lated to the interactions among electoral accountability, intrastate intrastate accountability in the region. The authors who focus on
accountability, and societal oversight. It does, however, set out a innovations show that in some countries, there is potential to re
new research agenda on this issue. The authors do not converge dress the accountability deficit. Smulovitz and Peruzzotti note
in their understanding of how electoral and intrastate accountabil promising innovations in what they call 'societal accountability',
ity interact, but they advance the debate about this important and I call societal oversight. Civil society has organized to oversee
subject.' public officials and agencies. Civil society organizations denounce,
mobilize, and forge alliances with the independent media to call
attention to problems in the discharge of public officials.
Intrastate Accountability in Latin America Sadek and Cavalcanti (this volume) analyze an old but trans
formed Brazilian institution whose mandate in the 1988 constitu
A third key debate in this book revolves around the quality of tion is broad: the Public Prosecutor's Office (Ministerio Publico).
intrastate accountability in Latin America. In the past decade, ac This institution is charged with undertaking criminal prosecution,
countability became salient in reflections about democracy among defending collective rights and public goods, and defending minor
such organizations as the World Bank and the Interamerican De ity rights. The 1988 constitution devised multiple means to ensure
velopment Bank, as well as a host of NGOs. the autonomy of the Public Prosecutor's Office vis-a-vis politicians.
One of O'Donnell's central claims is that mechanisms of hori Its oversight of and eventually prosecution of state actors (includ
zontal accountability are weak in contemporary Latin America. ing individuals in the executive, legislative, and judicial branches)
Indeed, he sees the weakness of horizontal accountability as one of falls within the broad domain of accountability. Their chapter
the most important differences between the old and wealthy dem underscores the innovative and transformative potential that this
ocracies and the post-1978 democracies of Latin America (and else institution has manifested. Their portrait of the judiciary and the
where). His work over the last decade has addressed differences Public Prosecutor's Office differs sharply from the one that Dodson
between these older democracies and Latin American regimes and and Jackson draw of the Salvadorean and Guatemalan judiciaries.
called attention to deficiencies of elected governments in Latin It suggests that some mechanisms of accountability in the region
America. In a seminal article (1994), he argued that weak horizon are innovative, vibrant, and somewhat powerful. Morgenstern and
tal accountability characterizes most of Latin America's elected Manzetti, while underscoring the limitations of mechanisms of
governments-in particular, those that he labelled 'delegative dem oversight in Argentina, also note some promising developments in
ocracies'. 'The horizontal accountability characteristic of represen recent years.
tative democracy is extremely weak or non-existent in delegative Figueiredo presents a mixed assessment of intrastate account
democracies' (61). Subsequent articles (1999a, 1999b) furthered ability in contemporary Brazil She argues that several potentially
this argument. important mechanisms of oversight-parliamentary investigative
Most scholars accept O'Donnell's viewpoint (e.g., Diamond et al. commissions, the Federal Accounting Tribunal (Tribunal de Con
1999: 1). While diverging from O'Donnell on other points, Moreno tas da Uniao), the Joint Congressional Budgetary Committee
et al. (this volume) agree that there is an accountability deficit in (Comissao Mista de Orcamento), and a wide range of other tools at
the region. In a similar vein, Morgenstern and Manzetti note the the hands of the legislature-have been relatively toothless in
deficiencies in mechanisms of oversight and corruption control in practice. As mechanisms of intrastate accountability, they have
"contemporary Argentina compared to the United States, Dodson not been as effective as their legal mandates allow. On the other
and Jackson underscore the severe deficiencies of the judiciary in hand, she argues that they generate information and stimulate
.'
sanctions would allow almost any actor to be an agent of account Is accountability Yes (Kenny, this volume; No (Day and Klein 1987;
ability. In contrast, with the exception of Smulovitz and Peruz- ' limited to formal Mainwaring; Moreno et Fox and Brown 1998;
zotti, the authors in this volume focus on a narrower range of (legalized) al., this volume) Keohane 2002; Merritt
authority to oversee forthcoming;Moncrieffe
actors that have formalized responsibilities to oversee public offi and/or sanction 1998)
cials. Within this fundamental agreement about a bounded set of . , .. ··- ,.. -· - ····- .. .
Is direct formally Yes (Kenney, this volume; No (Mainwaring;
actors that can constitute agents of accountability, the authors in ascribed(o-<"4v.,w.>') Moreno et al., this volume) O'Donnell, this volume;
this volume nevertheless have some differences of opinion that sanctioning po,v~ a Schedler 1999; all authors
stem directly from the previous disagreements. Because of their necessary named in the cell
definition of accountability, Moreno et al. limit these actors to component of immediately above)
accountability
voters, to whom elected representatives are accountable; polit -·-· '" . ' .
icians, to whom some bureaucracies are accountable; parliaments, Are accountability Yes (Moreno et al., this No (Kenney, this volume;
relationships volume, Elster 1999; Laver Mainwaring, Merritt
to which cabinets and ministers are responsible in parliamentary limited to and Sbepsle 1999) forthcoming; O'Donnell,
democracies; and other principals in principal-agentrelationships. principal-agent this volume)
They explicitly exclude oversight agencies and institutions that relationships?
--··· ~ . . .' . . ... ,, . ,, . - " . .
are independent from each other, Kenney also excludes oversight Agents of Principals in principal All actors tl1at formally
agencies but includes actors that have sanctioningpower (the pri accountability agent relationships oversee and/or sanction
mary example is the judiciary). O'Donnell includes not only princi (Moreno et al., this volume public officials (Kenney,
pals in principal-agent relationships, but also oversight agencies ·Elster-1999· Laver and this volume; Mainwaring,
'
Shepsle 1999) this volume; O'Donnell,
and the judiciary as agents of horizontal accountability-although
this volume)
all three kinds of actors only when the actor being investigated or
sanctioned has committed a legal infraction. I share his list of
agents of accountability,but as noted above disagree that they are
acting as agents of accountability only in cases of legal transgres are a subset of accountability. Political disagreement per se be
sions. tween the legislature and the executive does not necessarily consti
These arenas of agreement and disagreement about the concept tute a relationship of accountability. For example, if the legislature
of accountability are summarized in Table 1.1. In the final row, refuses to approve an executive-sponsored bill, this does not intrin
'Agents of Accountability', the scholars who do not limit account sically constitute a relationship of accountability because it does
ability to formal authority to oversee and/or sanction public not involve an element of answerability or sanctioning. The distinc
officials have a much less restricted conception than the two possi tion between accountability and checks and balances, however, is
bilities shown in Table 1.1. sometimes a fine one; checks and balances can constitute mechan
Kenney (this volume) raises one other important conceptual isms of accountability. The key issue according to the definition
issue: the relationship between horizontal accountability and proposed here is whether an actor has the constitutionaJJlegal cap
checks and balances. He properly argues that these two concepts acity.. to request an -accounting of a public official's (or agency's)
should not be confoundedor conflated; not all checks and balances discharge of duties or to impose: sanctions on that official. If so, it is
38 Guillermo O'Donnell The Legal Institutionalization of Mistrust 39
1f.. . Democracy ') At its origins in Athens, the basic idea of democ :J~iberalis,i} Liberalism is the only one of these great currents.
racy=is-menisfic. It does not postulate a sharp division between that is, eo-alarge but not exclusive extent, a direct manifestation
the public and the private spheres or, at any event, it does not of mistrust of political power. Liberalism, like republicanism, pos
recognize proper entity to the latter, except as endowing some indi tulates a distinction between the public and the private spheres,
viduals, the (male) citizens, with the time and means for devoting but it inverts their valence. It is in the manifold activities of the
themselves to public life. Democracy is also strongly majoritarian; private sphere-beginning, historically, with the freedom of con
ultimately what is decided by the majority is effected without im tract and of religious belief.-that the best potentialities of the
pediment from individual or prepolitical rights. This, of course, is human being may be achieved. The resulting 'private' freedoms
dangerous, as the memory of Socrates has attested since then. On (which I will call civil) are seen, as the consequence of some kind
the other hand, the very short terms of incumbency (many of of contract, or consensus, or natural law, as previous and/or super
them for just a day), the post-facto audit of the conduct of many ior to the public sphere, which is deemed to exist basically for
of these incumbents, ostracism, and the ever-present possibility of protecting, and eventually nurturing, these rights.
suffering severe sanctions under the accusation of having proposed In synthesis, both democracy and republicanism tend toward the
decisions contrary to the basic laws of Athens, established some pole of high decisiveness. Yet, because those who implanted the
controls against the, in this city-state, highly mistrusted incum historically original versions of these currents profoundly mis
bents of P.Q~itions of public power. 5 trusted ~h~~~-~~o.!~~Y-~~~?.~Jc!_with_potent~~}lt invasive p_owers,
deI.9:9~ra~ al!fil~ubli.£~.sn.!estabJ~~~~)~~~~~~?-01:1§1!!l~~an1_~~s_
2,... Republicanism It establishes a sharp distinction between ~m~__o(_~em..!~~!:.-9!!._.~d~:e~~t_~x.~~-<!~:E~.E-~!!19.<?.E~.!!~-~onstitu
0 ..
the ··pn:vate··and public spheres, asserting that it is in the latter tions-intended --------·
to ameliorate this -risk. Most of these mechanisms
--- .. - .. -.•- ,---·--- i::·y·----y: .
that the proper flourishing of the human being (once again, male were horizontal ones, consisting_of agencies_j_!l th_~U.QQUP_l!e~e
citizens) may be achieved. Virtue is dedicating oneself unselfishly aimed at controllh?,,.g_~E..<?!~!~-~~~Z-t?!P~!'-
a_g~n~ies_'?!..~.~divj.d~~!tloc~
to the public good, without allowing private or factional interests ted in the same sphere. This location was consistent with democ
to interfere with this pursuit. 6 As with democracy, republicanism racy's denial ofsubstantive entity to the private sphere and with
may be dangerous, because it easily leads to elitism and oligarchy: republicanism's assertion of the public sphere as the privileged one.
why should those who claim to be virtuous represent, or even Instead, liberalism tends to the pole of high protection, centered
listen to, those who are submerged in the pettiness of private pur in the private sphere. However, the effective protection of the free
suits? The tyranny of Savonarola and the cruelty of Robespierre doms cherished by liberalism entails the construction of a strong
are reminders of this danger. On the other hand, republicanism's political power that, among other things, commands important
insistence on virtuous dedication to the public good and, conse economic resources and establishes a complex system of courts
quently, the jealous controls of incumbents that it established and auxiliary institutions for the adjudication of conflicts over
where this current was stronger, the Italian medieval republics,' these rights, both between 'private' citizens and the state, and
placed some barriers to the pursuit of private or factional interests among the former. 0 As a result, one of the distinctive marks of
from the public sphere. liberalism is its more ambivalent position than democracy and
republicanism with respect to the contradiction I am discussing: in
order to obtain effective protection of its rights, liberalism, in fact
5 Elster (1999) discusses these controls; for details see Hansen (1991).· if not in purpose, cannot but endow the state with great, and con
6 Wood (1992: 104) puts it well: 'According to the republican tradition, man was sequently always threatening, powers."
by nature a political being, a citizen who achieved h.is greatest moral fulfillment
by participating in a self-governing republic.... Liberty was realized when the
citizens were virtuous-that is, willing to sacrifice their private interests for the
8
sake of the community.... Public virtue was the sacrifice of private desires and For pertinent argument about the 'cost of rights' and their ultimate depend
interests for the public interest,' . ence on the political process, see Holmes and Sunstein (1999).
9
7 See Waley (1988) for details, including the fascinating figure of the Podesta, The ambiguitiesof John Locke on this matter are emblematic; see especially
arguably the ultimate expression of mistrust toward political power holders. Scott (2000) and Zuekert (1996).
40 Guillermo O'Donnell The Legal Institutionalization of Mistrust 41
in the processes that lead to incumbency in governmental roles, hibition of de legibus solutus is one of the distinctive characteristics of political
and at least a minimum set of civil rights (especially of expression, democracy in relation to all sorts of authoritarian rule.
14 As Preuss (1996: 24} puts it: 'The quality of constitutions as laws, i.e., as
association, access to information, and movement) without which
generating the legal obligations of the governors, is essential. Only if the bond
such participation would be at best vicarious. The reasonable ef between governors and the ruled obligates the ruler ... and if it cannot be revoked
fectiveness of these rights is a necessary condition for the existence unilaterally by the ruler, is it possible to form a reliable institutional structure of
of democratic political power and of its authority to rule. Conse government in which the governed are recognized as the ultimate source of polit
quently, the exercise of this power should be guided-negatively- ical authority.'.
44 Guillermo O'Donnell The Legal Institutionalization of Mistrust 45
subject to its rules. Contemporary" democratic constitutions, the main fear was of excessive~Eremacy of the legislature; how
following the example of the United States, make these powers ever, along time, with the growth of the state apparatus it turned
relatively autonomous by means of their division and partial inter out that often the more serious threat consisted of encroachments
penetration, and assign to them jurisdictions that are carefully spe and corruption by the executive and by non-elected state bureau
cified by legal, including constitutional, rules. As noted above, crat~...:,~ew of these threats, balance HA suffers severaljlimita
many of the resulting relationships may be conceived as instances, tionsftr~ is that the b8-l~!?-<:e i~stitutions i_~ecutive, legi~ature,.
depending on the case, of mutual control or of exchange. These
relationships are based on the idea: that the jurisdiction of state ~~y?c~;.t~~!~U:~i§-=:~;?o~sc~;s:&e~n~!}~;~
agencies has lawfully defined limits that they are forbidden to t1ons. A ~limitation is that the actions toward horizontal
transgress. This idea holds even for the carefully speci!!_e4,__i:q account3:~=~ ba~an~~--ip.~ti~~9,Qg~_tend=to ~"t2e-veri:"iti-amat1£:.·
stances __ in_which1 __as_ l\1adison qesigned_ them,__tge juri$diction of _a_ These actions may create highly visible and costly conflicts be
_given powe~netrate~--t~~-j~~di-~~~Q~_of.... ~ot~~....Q!.l..!L{see also tween the supreme state institutions. These conflicts further
Manin 1994). From these rules results a lega}!l_deli,mite~.E.1:~-~.f sharpen when, as in presidentialist regimes, they may involve
the juri~diction ~ st!l-te institution:_st ... including_ tJ,ie three ...:ru:~ powers-the executive and legislative-that share electoral legit
eminent ones~_executiy~,_l~~lative,andjudiciary. This is not only imacy. ~j) the actors in these conflicts are often perceived as
~-.§.YSt.~.m._.9.f..m!lil!~-~Q~t1:Q!~: ..!t..2r.gan.iz~§...th.!t..fl9JY...2f..J.?OWer @d.. mainly motivated -oy partisan--ieasons,· tlius adding to the·
8:!!~Q~J.~.Y....~--~h~_g2._y~?;Oine_!!!__?.J.ldJ_he__;:;t3:.~ thus normally produ Sharpening and ffie drtficulties of soMng the- _gfven- conflic~~
cing a division of labor among the relevant public institutions, Finally, because-ortliefr".own"'iiiaui""fuiicti"ons"-ruicf"consequently of
-~-!rich ~re therefore en~]:>1~9- to discharge thej.r respective respOJ?:~t-:·__ the foci of attention of their leaders, the balance institutions are
bilities. u; too blunt an instrument of control for the growing complexity of
· Some of these relationships, however, originate in the view by a state agencies and their policies.
(Qyen agency that another one ~as_unlawfully transgressed its~-~ The realization of these limitations for the effectiveness of HA
jurisdiction, and encroache~ on_t1:te.,h!ris~iction of the former (or a led practically everywhere, albeit with varying timing and char
third agency, in some cases). When a situation of this kind is acti acteristics, to the addition of agencies of mandated horizontal
vated, we face a case of horizontal balance a£COf!,ntabi~J!Y..:_I adopt accou~ili_2':_ . T..~~~-~--~-~. iti~. _varj9~~-. ~g~-~~!~~..(Qm.l?..yg._sm~n,_a.~
this nomenclature because this kind of horizontal accountability is C(!_al~!n~-~~-~~--~o~~!.9 ..l:!~!.~t.EO"f!-S~-~~~-<J..~"!i~!:!~"-·f!:.§.£<J:.l~as,contrq,lorias,
enacted by some of the powers (executive, legislative, or judiciary) ~- Qi.eJik~). th~t are leg_~_Y.._<!~~~_gn~Q_-~ -~-~~~~~~t.p_rev~~--~i. ~..ES.~~-~
that contemporary democratic constitutions attempt to keep age, promote the sanctioning, or sanction presumably unlawful.
roughly in balance." Originally, in the United States and elsewhere act1onsoroniissions·-of-otlier"'
- . "'agenc1es~'"-·nationaf
-----··,,,-: : __ "state,, ,_., _. . . or"'
. "su6na:··,_
tfonal::Tliese agencies, unuxe the older ones, were invented not so
16 I say 'contemporary' because only after the Second World War continental muchhaving in mind overall balances of power but rather specific,
European constitutions, basically by means of the enactment of various kinds of but still quite general, risks of encroachment and/or corruption. In
judicial or constitutional review, cancelled the formal supremacy of the legislature. principle, the mandated agencies have several advantages over
This way the differences that in this respect exist between presidentialist and the balance ones. One is that they can be proactive and continuous
parliamentary regimes have been significantly mitigated; see Guarnieri and Ped in their activity. Another is that, by the same token, they can be
erzoli (1996), Pasquino (1998), Stone Sweet (1992 and 2000), and Vanberg (1998).
After Communism, the Central/Eastern European countries have also adopted effective in preventing or deterring unlawful actions by the agen
various kinds of judicial or constitutional review (see Magalhaes 1999 and cies they oversee. Still another advantage is that they can invoke
Schwartz 1993); this is also true of new democracies elsewhere (see Maddex 1995 for their actions criteria that are professional rather than partisan
and Tate and Valloder 1995). ·. , or 'political'. Finally, these agencies can develop capabilities that
ie On this positive, enabling aspect of the legally-delimited jurisdiction of state allow them to examine complex issues of policy. Having said this, I
agencies see Holmes (1995) and Waldron (1999).
17 As noted, with the arguable exception of the United Kingdom, this is gener
hasten to add that, ideally, mandated agencies should not be con
ally true of both presidentialist and parliamentary regimes, after the adoption by ceived as substitutes for the balance institutions-a reasonably
most of the latter of some form of judicial or constitutional review. working democracy finds in the former a useful complement and
46 Guillermo O'Donnell The Legal Institutionalization of Mistrust 47
reinforcement of the latter. We shall see, on the other hand, that in Now I come back to an argument that needs emphasizing. For
less than smoothly working democracies, mandated agencies can HA to effectively function it is not only necessary that for a given
be important in facilitating and even promoting various kinds of issue a given state agency is legally empowered and willing to act.
vertical accountability. It is also necessary that there exists a whole network of state
Before discussing other topics, here I want to emphasize two agencies, culminating in high courts, committed to preserving and
points. One, the enactment of horizontal accountability by what eventually enforcing horizontal accountability, if necessary against
ever institutional mechanism is entailed by the prohibition of de the highest powers of the state. Otherwise, the investigations of
legibus solutus. Two, this prohibition is in turn entailed by a con auditorias, fiscalias and the like may feed an angry public opinion
ception of democracy that includes the protection of some basic but do not reach legally appropriate resolution. 18 As noted, in a
participatory and civil rights. democracy a legal system is supposed to be such a system, one that
'closes,' in the sense that nobody is exempt from it. This system
consists not only of legal rules but also of institutions committed
A Map to upholding these rules. Surely the acid test of the existence of
this kind of system is if it applies or not in terms of the horizontal
Let me propose an image. It is a space marked by a series of accountability of the highest powers of the state.
boundaries, or limits, the biggest of which lies between a public
and private sphere, as postulated (albeit, as we saw, with different
valences) by republicanism and liberalism. Empirically this is a (Vertical) Societal Accountability
rather fuzzy limit, especially after new patterns of state and social
activity have created numerous mixed public-private zones. But in Recently Catalina Smulovitz and Enrique Peruzzotti, have pro
spite of its imprecision this is still an important boundary, as it posed a concept that I find very useful, a type of vertical but non
separates those who can mobilize the collectively binding and ul electoral accountability that they call 'societal'. They define it as
timately coercive powers of the state from those who hold rights follows:
that these powers are prohibited to void. In another part of the
Societal accountability is a non-electoral, yet vertical, mechanism of con
map, at the level of the state, we see a series of institutions, in trol of political authorities that rests on the actions of a multiple array of
cluding those that concern themselves, whether in a reactive and citizens' associations and movements and on the media, action that aim
intermittent or a proactive and continuous way, with issues of HA. at exposing governmental wrongdoing, bringing new issues to the public
Precisely because, as we saw, the balance institutions are designed agenda, or activating the operation of horizontal agencies. It employs
so as to partially interpenetrate, the idea of legally defined limits both institutional and non-institutional tools. The activation of legal
is extremely important. Encroaching occurs when a state institu actions or claims before oversight agencies are examples of institutional
tion oversteps the limits of its own legally mandated jurisdiction, resources; social mobilizations and exposes illustrates some of the non
and invades those of another; redressing encroachment is to re institutional ones, [the effectiveness of which] is based on symbolic sanc
store those limits and eventually punish the transgressors. In still tions. (2000: 8. See also this volume.)
another part of our map, the level of society, there exist the mani These authors argue persuasively that it would be wrong to limit
fold boundaries entailed by the rights of individuals and various the concept of vertical accountability to elections. There exist mani
kinds of associations. Some of these boundaries are solid, as they fold actions, individual and collective, that take place in society at
consist of rights that are widely respected and, if necessary, upheld any time between elections. A good part of these actions is aimed at
by courts. Other boundaries may be weak, because the respective redressing, stopping and/or sanctioning actions, and sometimes
rights are legally or politically problematic or because, as I have omissions, of elected (national or subnational) individuals, as well
argued is the case in some new democracies (O'Donnell· 1999b), as non-elected state officers, also national or subnational. Especially
rights are selectively upheld across various regions and categories. ~ ·,
of citizens. The strength and actual functioning of all these bound ·,a See, for' ek~if:i11ile; l.lJ.Q impotence of' the Ombudsman in Fujimori's Peru, dis
aries, across cases and time, is a major issue that needs research. cussed by Kenney (this.votume).
48 Guillermo O'Donnell The Legal Institutionalizationof Mistrust 49
in countries where, as in Latin America, electoral vertical account· voting, elections, and the aggregation of preferences, including the
ability functions quite deficiently,19 the societal version of vertical conclusions reached by Bernard Manin, Adam Przeworski, and
accountability becomes extremely important for the workings and, Susan Stokes in a recent careful study of these matters:
eventually,even for the survival of a democratic regime." [T]he conclusion of this analysis must be that citizen's control over polit
This concept advances over and usefully elaborates my original icians is at best highly imperfect in most democracies. Elections are not a
remarks (O'Donnell 1999a) about the importance of the linkages sufficient mechanism to insure that governments will do everything they
between the horizontal and vertical dimensions of accountability. can to maximize citizen's welfare .... Yet even if responsibilities are
On one hand, an alert and reasonably well-organized society, and a clearly assigned, bad governments can be punished and good ones chosen,
media that does not shy away from reporting cases of encroach voters are well informed about the relations between politicians and
ment and corruption, provide crucial information, support, and special interests, and the rent-seeking behavior of politicians is well scru
political incentives for the often uphill battles that agencies of tinized, elections are just not a sufficient instrument of control over polit
horizontal accountability may wage against powerful transgres icians. Governments make thousands of decisions that affect individual
sors. On the other hand, the perceived availability of this kind of welfare; citizens have only one instrument to control these decisions:
the vote. One cannot control a thousand targets with one instrument
horizontal agency may encourage undertaking actions of vertical (1999: 50).
societal accountability. These effects, respectively of stimulation of
horizontal and of inducement of vertical societal accountability, are It is, precisely, the experience of the deficiencies of electoral ac
extremely important for understanding the dynamics of demo countability that has encouraged the development of both vertical
cratic politics, especially in countries where, as noted, vertical and horizontal accountability.
electoral accountability works quite deficiently. Furthermore,
actions of societal accountability-particularly if they are vigorous,
sustained, and gain widespread public attention-may send strong Linkages and Interactions
signals to politicians wishing to be elected or re-elected.
These various vertical/horizontal interactions are another im It follows from the preceding discussion that all types of account
portant frontier for future research. This frontier would remain ability are important for the workings of a democratic regime.
closed if, as Moreno et al. demonstrate, the vertical dimension of Among these types, vertical electoral accountability has logical pri
accountability is restricted to the electoral one and if, in addition, ority due to the sheer fact, already noted, that without fair elec
horizontal accountability is conceived as a mere 'adjunct' of the tions and, with them, the possibility of unseating the incumbents
electoral. Furthermore) these authors are unduly optimistic about in the next round, simply there would not exist, by definition, a
what, even under ideal circumstances, electoral accountability can democratic regime. But it is mistaken to equate logical priority
accomplish. See, for example, '[I]f the accountability of elected offi with pre-eminent factual importance. In the present stage of our
cials were working as intended-such that voters could and would knowledge, the only point that I believe may be confidently made
punish misdeeds-separate branches of [horizontal, O'D] superin is that the more important effects, and the more intriguing ones as
tendence would be unnecessary' (Moreno et al., this volume). This a frontier for research, are the ones to be found at the interactions
assertion contradicts the findings of a large and solid literature on among various kinds of accountability.21
Of course, too much of a good thing can be bad. For example,
excessive horizontal accountability moves the situation toward
19 For a series of reasons closely related to the party systems as well as to the the pole of deadlock and, consequently, very low decisiveness. But
electoral and party legislation. These reasons are well spelled out, among others, we are in a terrain that may invite simplistic conclusions. Con
in Mainwaring (1999) and Moreno et al. (this volume). . sider what I have called 'delegative democracy' (O'Donnell 1994).
20 This importance is worn out by Smulovitz and Peruzzotti (2000 and this
volume), as well as by the analyses and the case studies contained in their edited
21
volume (2002). Incidentally, the sections on the United States of Morgenstern and For evidence on the importance of these interactions see Manin, Przewroski
Manzetti in this volume provide interesting evidence on the impact of societal and Stokes (1999). For these interactions in relation to the office of the public
accountabilityin the history of the United States. prosecution in Brazil, see Sadek and Cavalcant_i (this volume).
50 Guillermo O'Donnell The Legal Institutionalization of Mistrust 51
Without entering into the discussions that this concept has gener this kind of information, because it is legally required and undue
ated, I note that a delegative conception and practice of political ·secrecy would generate horizontal accountability actions. On their
power is centrally based in the denial of the appropriateness of. part, some mandated agencies dig into what at times are arcane
boundaries among state institutions. For a delegative view, balance amhits of the agencies they oversee. Usually this adds a lot of free
and mandated institutions are a tremendous nuisance, impedi floating public information. This information can be utilized by
ments placed by the 'formalities' of the legal system to the high . those who are exercising vertical accountability, not only voting
decisiveness that delegative rulers claim it is their right and duty but also societal accountability. Little of this· information would
to exercise. In turn, authoritarian 'rulers are more radical than exist were it not for the functioning of HA. Actually, this availabil
delegative ones: intrinsically hostile to all kinds of accountability, ity is a by-product of the functioning of balance and mandated
they void the legal rules that promote it and, in addition, they make institutions (in general, they are not designed with the purpose of
the effectiveness of civil rights contingent on their de legibus solu generating this kind of information), but they do provide links
tus expediency. Yet none of this, I want to emphasize, make au that at least potentially invigorate electoral and societal vertical
thoritarian rulers, and even less so delegative ones, omnipotent. accountability. This increases the 'availability of alternative
Often these rulers find constraints in their power relations with sources of information' that Robert Dahl correctly lists as one of
other segments of their regimes and powerful social actors. Yet, in the central characteristics of polyarchy or political democracy
contrast with horizontal accountability, these constraints do not (1989: 120).
result from the effectiveness of legally enacted rules. The con A final comment on HA pushes us away from its interactions
straints are the result of naked power relationships which may, as and into the mysterious but relevant world of non-events. Presum
in fact they have in not a few cases, lead to severe deadlocks." ably (but, I take it, very plausibly) no small part of the importance
In addition to the mutual relationships of stimulus and induce of HA, insofar as it is effective and perceived to be so, lies in the
ment already noted, another aspect of horizontal accountability transgressions it prevents or deters. We do not have hard data on
provides an important link with its vertical side. AB conceived in this matter, but it stands to reason that the decision function of a
Athens, if the citizens are the source of the authority of the polit would-be transgressor will be heavily influenced by her assess
ical power exercised over them, then they have the right to be ment of the probability of being caught and sanctioned. It may
informed of the decisions of this power. It follows that the decisions well be that the effectiveness of HA, both balance and mandated,
of democratic authorities must be public, in the double sense that is curvilinear. It works well when very little happens in this
the rationale and content of these decisions must be 'made public', matter, because there are few and not serious transgressions; yet
and that the procedures" that lead to these decisions are specified it also works well when some agencies bring to the top of the
in legal rules that are also publicly available. a. Truly, no democracy public agenda, and sanction, serious transgressions perpetrated by
has ever fully complied with this requirement, and even among high state powers.
highly established ones there are important variations in this
matter. But the contrast remains with the secrecy with which au
thoritarian regimes surround many decisions. 26 HA is an import Some Conclusions
ant generator of information that becomes publicly available. The
relationships among balance institutions generate large flows of HA is a legally grounded and legally activated interaction between
state agencies in view of presumed unlawful actions or omissions
22 I insist on this observation because, even though I t.ook care to make it (encroachment or corruption). Ho~ntal accountability,_thus, is a_
explicit in O'Donnell (1994), some readers have interpreted that there I assert s~m_l suJ?.§~!_.Q.f....t.h~.. rµ~.if.o!g... jP.:!~!~~t!9.9:~..... ~h.?.!.... ~~~E~--,-~g_el!~~~
that delegative rulers are in all circumstances close to omnipotent. _:gnde:;:tak.e among the!D,sel'\:'.~S. Its importance lies not only in the
23 And, in Athens, the deliberations themselves.
24 Bobbio (1989) and Garzon Valdes (1993) usefully analyze these and other
actions it triggers but also in the transgressions it prevents or
aspects of the publicness of democracy. deters. Conversely, if responsibilities of horizontal accountability
25 To the point of the absurdity of the 'secret laws' enacted by Stalinist and some exist in the °letter of the law but the respective agencies are
military rulers. · rendered ineffective by superior powers, the consequence will be
52 Guillermo O'Donnell The Legal Institutionalization of Mistrust 53
not only facilitating transgressions of these powers but also the Garzon Valdes, Ernesto. 1993. 'Acerca de los Conceptos de Publicidad,
discrediting of those agencies. Opinion Ptiblica, Opinion de la Mayoria y sus Relaciones Reciprocas'.
!JA:j.oe~..-~o~--~~p.d_.QE!Y!o n~tj.Qnal~!ec~.d officials; its scop~ Doxa 14: 77-95.
includ~st1:~na~~n~~J~ctedofficials ~g..J}Qp.-el~_ct~d__members...9.f_ Guarnieri, Carlo and Patrizia Pederzoli, 1996. La Puiseance de Juger.
the_ state_ bureaucracies,_ natio.naj_and.. i=Jub-natiQn&..Not neglecting Pouuoir Judiciare et Democratie. Paris: Michalon.
the sub-national level of HA seems to me particularly important in Hansen, M. H. 1991. The Athenian Democracy in the Age of Demosthenes.
countries where the legal system operates rather effectively in Oxford: Oxford University Press.
Hirschman, Albert. 1970. Exit, Voice,and Loyalty: Responses to Decline in
some regions but not in others. At all these levels fIA is an import
Firms, Organizations, and States. Cambridge, MA: Harvard University
ant component of a democratic regime, per se and because of its Press.
linkages, actual and potential, with both electoral and societal ver Holmes, Stephen. 1995. 'Constitutionalism', In Seymour Martin Lipset,
tical accountability. These linkages need much research, among ed., The Encyclopedia of Democracy, pp. 299-306. London: Routledge.
other reasons because they may identify strategic opportunities Holmes, Stephen and Cass R. Sunstein. 1999. The Cost of Rights: Why
for improving the presently poor functioning of many democracies. Liberty Depends on Tuxes. New York, NY: W.W. Norton.
On the other hand, I agree that 'fundamental party and electoral Lijphart, Arend. 1984. Democracies: Patterns of Majoritarian and Con
reform is a necessary condition for improving the quality of democ sensus Government in Twenty-One Countries. New Haven, CT: Yale
racy in the region' (Moreno et al., this volume). But the necessary University Press.
efforts in this direction should be complemented by no less persist Maddex, Robert. 1995. Constitutions of the World. Washington DC: Con
gressional Quarterly Inc.
ent efforts to solve the serious deficits in HA existing in many
Magalhass, Pedro. 1999. 'The Politics of Judicial Reform in Eastern
new, and some not so new, democratic regimes. Any view that Europe'. Comparative Politics 32, 1: 43-62.
unilaterally asserts the pre-eminence of a single dimension of Mainwaring, Scott. 1999. Rethinking Party Systems in the Third Wave:
accountabilityis not likely to be helpful for the achievement of the The Case of Brazil. Stanford, CA: Stanford University Press.
goals of democracy-enhancement that the contributors to this Manin, Bernard. 1994. 'Checks, Balances and Boundaries: The Separ
volume share. ation of Powers in the Constitutional Debate of 1787'. In Biancamaria
Fontana, ed., The Invention of the Modern Republic, pp. 27-62. Cam
bridge, UK: Cambridge University Press.
References Manin, Bernard, Adam Przeworski and Susan Stokes 1999. 'Elections
and Representation'. In Bernard Manin, Adam Przeworski and Susan
Bobbio, Norberto. 1989. Democracy and Dictatorship: The Nature and Stokes, eds., Democracy, Accountability, and Represeniation; pp. 29-54.
Limits of State Power. Minneapolis, MN: University of Minnesota New York, NY: Cambridge University Press.
Press. O'Donnell, Guillermo. 1994. 'Delegative Democracy'. Journal of Democ
Dahl, Robert. 1989. Democracy and Its Critics. New Haven, CT: Yale racy 5, 1: 55-69. First published as Kellogg Institute Working Paper
University Press. No. 172, 1992.
Dunn, John. 1999. 'Situating Democratic Political Accountability'. In -- . 1999a. 'Horizontal Accountability and New Polyarchies'. In Andreas
Bernard Manin, Adam Przeworski and Susan Stokes, eds., Democracy, Schedler, Larry Diamond, and Marc F. Plattner, eds., The Self-Restrain
Accountability, and Representation, pp. 329-51. New York, NY: Cam ing State: Power and Account-ability in New Democracies, pp. 29-52.
bridge University Press. Boulder, CO and London: Lynne Rienner. First published as Kellogg
Elster, Jon. 1999. 'Accountability in Athenian Politics'. In Bernard Institute Working Paper No. 254, May 1998.
Manin, Adam Przeworski and Susan Stokes, eds., Democracy, Account --. 1999b. 'Polyarchies and the (Un)Rule of Law in Latin America'. In
ability, and Representation, pp. 253-78. New York, NY: Cambridge Uni- Juan Mendez, Guillermo O'Donnell, and Paulo Sergio Pinheiro, eds.,
versity Press. '. The (Un)Rule of Law and the Underprivileged in Latin America, pp.
The Federalist. 1961. The Federalist Papers: A Collection of Essays Writ 303-37. Notre Dame, IN: University of Notre Dame Press.
ten in Support of the Constitution of the United States, from the Orig -- . 2001. 'Law, Democracy, and Comparative Politics'. Studies in Com
inal 1ext of Alexander Hamilton, James Madison, John Jay. Roy P. parative International Development 36, 1 (Spring). Previously pub
Fairfield, ed. Garden City, NY: Anchor Books. lished as Kellogg Institute Working Paper No. 274 (April).
54 Guillermo O'Donnell
O'Donnell, Guillermo. 2002. 'Acerca de varias accountabilities y sus
<Ill •••••• " •••••••••••••••••••••
interrelaciones'. In Catalina Smulovitz and Enrique Peruzzotti, eds.,
Controlando la Politico. Ciudadanos y Medios en las Nueoas Democra
cias Latinoamericanas, pp. 87-102. Buenos Aires: Temas. 3
Pasquino, Pasquale. 1998. 'Constitutional Adjudication and Democracy. . .
Comparative Perspectives: USA, France, Italy'. Ratio Juris 11, 1:
38-50.
Preuss, Ulrich. 1996. 'The Political Meaning of Constitutionalism'. In Horizontal Accountability:
Richard Bellamy, ed., Constitutionaliem; Democracy, and Souereignty:
American and European Perspectioes, pp. 11-27. Aldershot: Avebury.· Concepts and Conflicts
Rawls, John. 1993. Political Liberalism. New York, NY: Columbia Univer
sity Press.
Schwartz, Herman. 1993. 'The New East European Constitutional
Courts'. In A.E: Dick Howard, ed., Constitution Making in Eastern Charles D. Kenney*
Europe, pp. 163-207. Baltimore, MD: The Johns Hopkins University
Press. Introduction
Scott, John T. 2000. 'The Sovereignless State and Locke's Language of
Obligation'. American Political Science Reoieui 94, 3: 547-61. One of the important contributions Guillermo O'Donnell has made
Smulovitz, Catalina and Enrique Peruzzotti. 2000. 'Societal Accountabil to recent debates about democracy is the conceptual framework of
ity in Latin America'. Journal of Democracy 11, 4: 147-58. vertical and horizontal accountability.1 According to O'Donnell's ini
Stone Sweet, Alec. 1992. The Birth of Judicial Politics in France: The tial presentation· of this framework, accountability was said to run
Constitutional Council in Comparative Perspective. New York, NY:
Oxford University Press. not only vertically, making elected officials answerable to the ballot
--. Governing with Judges. 2000. Constitutional Politics in Europe. box, but also horizontally, across a network of relatively autonomous
London: Oxford University Press. powers (i.e., other institutions) that can call into question, and eventually
Tate, C. Neal and Torbjorn Valloder, eds. 1995. The Global Expansion of
Judicial Power. New York, NY: New York University Press. * The original version of this chapter was written for the conference on Insti
Vanberg, Georg. 1998. 'Abstract Judicial Review, Legislative Bargaining, tutions, Accountability, and Democratic Governance in Latin America held at the
and Policy Compromise'. Journal of Theoretical Politics 10, 3: 299--326. Kellogg Institute for International Studies at the University of Notre Dame in
Vile, M. J. C. 1967. Constitutionalism and the Separation of Powers. May 2000. An earlier version was presented at the Latin American Studies Associ
Oxford: Clarendon Press. ation meeting in March 2000. David Close, Michael Coppedge, Jonathan Hartlyn,
Waldron, Jeremy. 1999. Law and Disagreement. Oxford: Clarendon Press. Scott Mainwaring, Philip Mauceri, Guillermo O'Donnell, Andreas Schedler, and
Waley, Daniel. 1988. The Italian City Republics. London: Longman. Chris Welna offered thoughtful comments that enabled me to improve this chap
Wood, Gordon S. 1992. 'Democracy and the American Revolution'. In ter, although a number of their concerns have not received proper attention here.
1 I refer here to O'Donnell's writings on delegative democracy, first published in
John Dunn, ed., Democracy: The Unfinished Journey, 508 BC to AD
1993, pp. 91-106. Oxford: Oxford University Press. Spanish in 1991 and then in English in 1992 as a Working Paper of the Kellogg
Zuckert, Michael. 1996. The Natural Rights Republic: Studies on the Institute for International Studies and in 1994 in the Journal of Democracy, and
Foundation of the American Political Tradition. Notre Dame, IN: Uni to his article on horizontal accountability published in 1998 in the Journal of
versity of Notre Dame Press. Democracy.For further explorations of the theme of horizontal accountability, see
Schedler, Diamond, and Plattner (1999); for studies on the topic of vertical ac
countability, see Przeworski, Stokes, and Manin (1999). As Robert Pastor (1999)
and Andreas Schedler (1999: 25) point out, there is an important third dimension
of accountability not theorized here, that exercised by the international commu
nity. Much of what passes for international relations falls more in the category of
'naked power re.1a~1onsh,ip~· than those of accountability relationships governed by
some sort" of rule of law. Increasingly, however, governments find them.selves en
gaged to. agr_eeme:ci"l:s and treaties that form a body of international law within
which acc<1untnbili.ty rclationshl ps may develop.
56 Charles D. Kenney Horizontal Accountability: Concepts and Conflicts 57
punish, improper ways of discharging the responsibilities of a given office. to be roughly synonymous in Williams' usage, and today one may
(1999a: 165) say-as Jon Elster does-that 'accountability', 'responsibility', and
Democratization, O'Donnell argued, required not only the streng 'answerability are 'near synonymous terms' (1999: 255).
thening of vertical accountability-which at the time was making If this is the nature of accountability, what precisely is horizon
great strides in Latin America-but also the building of an insti tal. accountability? According to O'Donnell,
tutional network to provide horizontal accountability-something horizontal accountab~~-. is . . the existence . of_ state_ a,ge_!l.cies ... that are
which was famously lacking in Latin America. ' . . "···---~--~~L~~-~~p.9~~r~~L~--f~£~!Y....wiJ:ling__~1!~.~!?_!~1-~ ..~~--
Since O'Donnell's initial work, a fruitful debate ···on
r·-- _ . . __---·-----··· ------·· . the concept
·-···-··fi- -··-···---··---·-··- of.. ~ction§.__ih.~!.1!P.!ID...f!Q!!!... ~~E~.~----2.Y~r~~gJ~.t.J.2
.. .<.:1.i~~!..~!nctiops Q!_im::
l.!9.riZ.Q!!!.?........~~~2~-~_!;-~QJJitY-.}t~e!f. ..h~f>.. . e~~.rg~-~-: In the rst part· of peachment in relation to. actions ..9r .Q_missionsby_other _agents or agencies
this chapter, I will take up this debate and attempt to clarify what q_f_ih.~_§t~Jetha!_rn~~-P.~..9.!1.aji~g41!.~.!!!Y.?..wf)1!:.J!~Jl_~9_:_?.§)
horizontal accountability is and is not. The main issues I discuss
are whether horizontal accountability is best co.rice1vea in terms of Four aspect~_!>f ho:tiz....?.:':1:!~, ..~~~<?~~-~~!!~:Y-are specified in O'Don
nell's definition:
plincipru--agent-relationslups:wlleffieinonzontafaccountaniHtyTs
tlie "same - as"" ch.eeks'-andbalances"-"'whetlier it re-·uires -sanctions···· 1
~ As is the case with vertical accountability, the objects of hori-
··--···--··-··--·-········-····· ---·-----------·---:::J -·- . ·---·--·---·-·-- ~-----·--····-···· ... -·-·..~
and whether it extends to actions not considered unlawful. The ~ zontal accountability are state agents.
second part of this chapter explores a puzzlerelated to democratic 2 Unlike vertical accountability, the subjects of horizontal ac-
legitimacy. In recent years, many observers have been surprised .,'!' countability are also state agents. -·---
---
and often dismayed by the ability of undemocratic leaders like ~ The means of horizontal accountabili~ include oversightt
Fujimori in Peru and Chavez in Venezuela to win critical battles sanctions, and im:~achment. - -- . . -
for domestic democratic legitimacy. Why did democratic legitimacy 4 µhe scope of horizontal accountability is limited to actions or
sometimes accrue to these apparently undemocratic leaders? The - onussions 9...uaIHred·as unlawful linc1udT:og_violations of a
key to understanding this puzzle, I argue below, lies in recognizing counb;:y'~fug~~tJa~~;·I~-~~sti~tion), ---· ---
the ways in which vertical and horizontal accountability may come
That horizontal accountability is about controlling the actions of
into conflict in modern democracies.
s~feage~~-is 'commonlY. acceptecl The -subjects, means, and scope
of horizontal accountability are, in contrast, disputed. Before enter
ing into these disputes, I would like to take up two prior questions,
What Horizontal Accountability Is and Is Not namelii}iow horizontal accountability may be understood as a
relationshi - . .oetween...... rincipals .. andagents,. an····-· ow horizontal
The Concept of Horizontal Accountability
accountaoffi1yrefates··fo·other-·forms-oTgovernm~al seil~control
The concept of accountability has long been present in political balances."":-········-··.· . -
such ailigi~.!~!i--~:~~liec¥:~~~~ ~.. ·:· .).
discourse, but it has generally been used to refer to what O'Don Elster holds that accountability has a 'form.al triadic structure: a.£.
nell calls vertical accountability (citizens holding officials account agent A is accountable to a principal B for en action X' (1999: 255).
able) and less often to what he calls horizontal accountability This is a useful observation, but one that also may prove misleading,
(officials holding one another accountable). Madison and Hamilton because the relationships at the heart of horizontal accountability
used the terms 'accountable' and 'unaccountable' in the sense of often do not align simply with relationships between principals and
vertical accountability in The Federalist Nos. 55 and 70. According their agents. When a popularly elected legislature impeaches and
to the Oxford English Dictionary, the term 'accountability' was removes from office a popularly elected president, both the legisla
first used by Samuel Williams in 1794 when describing the govern ture and the executive may be said to act as agents of the electorate,
ment of the Native Americans living in Vermont. According to their principal. In this case, the impeached· president can be con
Williams, this government had 'no written constitution, or bill of sidered an agent of the legislature only in a tautological sense: if one
rights; no mutual checks, and balances, accountability and respon defines accou~t~~ility relationships as principals holding agents ac
sibility .. .' (1794: 140). 'Responsibility' and 'accountability' appear coup table, as Elster . does, then the president would perforce be the
,,....
I~
not clearly distinguish horizontal accountability from the concept of checks and r. In addition to the points discussed here, see also the comments offered by
balances. O'Donnell's definition is itself clear enough, but often his readers-· Richard Sklar (19g9) and Marc Plattner (1999), as well as the response by O'Don-
myself included-are insufficiently attentive to the parameters he sets out. nell (1999c). . · · •,.
Charles D. Kenney Horizontal Accountability: Concepts and Conflicts 63
62
Schedler argues that the concept of accountability contains two
irreducible dimensions, those of enforcement and answerability
Executive (Schedler 1999: 14). Noting that accountability is defined by some
in terms of answerability,he holds that answerability implies the
requirement that agents both inform about and explain their
I I · 1 Others, such as actions, and that accountabili~ implie~_something more__th8}1 ju~~
I
I Legislature I Ombudsman
inform.atio~_and ~~12lana~~~~:it cC?_~otes the abili~ to p!]llis~_un.:.
sat}sfact~..9-:~h~vio~. This would seem to place the capacity~
sanction at the heart of horizontal accou11tability. Schedler goes on
Judiciary State to argue, however, that in some cases<accolintabilitymay be di
vorced from sanctions ... without necessarily creating "diminished
subtypes" as a result' (18). ·
In contrast to Schedler, I would argue that the capacity for
sanction is essential to the concept of accountability, and that
accountability without sanctions is indeed a diminished form of
accountability. Schedler distinguishes between enforcement and
answerability, presenting the latter as if it were limited to infor
mation and explanations. I think this distinction is inaccurate.
Answerability is not just a matter of information and explanation,
but itself connotes the capacity to punish. If I am answerable to
Society someone, that person has some power with respect to me that can
0
be exercised in the form of sanction under.certain/" '
circumstances.
",
.... ·.
Answerability_and accounta~~Y.: are 'n~r-~YE-..9.IlXJ.Q~~_to ~~~£:
Horizontal accountability • countable/answerablemeans that someone else has the power to
Vertical accountability - - - - ~ _sanctjgn..you. Even if answerability were conceived_as.simply.the
FIG. 3.1. Horizontal and vertical accountability relationships
obligation to provide information and explanations, this too would
require the capacity for sanction. The obligation of A to provide
information and explanations to B may be distinguished analytic
ally from the capacity of B to sanction A, but the fact is that the
relations might be mistaken for examples of horizontal account former depends on the latter-without sanctions, there is no obli
ability under O'Donnell's definition since they take place within gation.
the state, not between society and the state. However, horizontal.. So understo9_c!i.J!J.e
__capt:1ci,ty_t,o...§.~£tiq!!__i~--~-·~s~~P.-Ji?l.. di!:n.&:g.
accountability focuses primarily on accountabili!I_relationships
-- •••i
---- .... . -----·--·-F••,1•• --··-·------ ... -- .. --.. -- .... - •• §iO!L~f.ilccountabiliIB_~i ~QUt ~hl£h..~~f.Q~t<!,.bi.lityexist.s onlY. .!!1
oetween agencies, not within them, and according to O'Donnell's a limited, truncated form. Schedler is ambivalent on this point. He
denQftio!)._s~~Ii intra-~ency relationsfil2s do not fali-!i'i14~:i- _f4~.
concept of horizontal accountability. 6
According to the Oxford English Dictionary the primary meaning of 'answer
able', one that dates from the sixteenth century, is 'liable to answer to a charge'
Horizontal Accountability and Sanctions and 'liable to be called to account; under legal or moral obligation; responsible,
Does horizontal accountability require the capacity to sanction? accountable'. The original sense of 'answer' is 'to make a statement in reply to a
legal charge; to meet a charge of any kind; to be liable to do so, or to suffer the
This is a second disputed aspect of O'Donnell's definition of hori consequences, to atone, pay the penalty'. The word 'answerability' is not found in
zontal accountability. Andreas Schedler (1999) has raised ques the Oxford English Dictionary, but it seems clear from the above that answerabil
tions about the centrality of sanctions in the concept of horizontal ity should be und~rs,to.od. as directly tied to enforcement, rather than as a separate
accountability, and in the concept of accountabilitymore generally. matter of information and explanation. · ·
Horizontal Accountability: Concepts and Conflicts 65
6~ Charles D. Kenney
the time of the bribe, no bribery of a public official had occurred.
recognizes that 'academic writers are often quite emphatic in
Case dismissed. Subsequent events undermined Montesinos' im
stating that the capacity to punish forms an integral part of polit punity, but the point is clear: publicity cannot substitute for sanc
ical accountability' (Schedler 1999: 16), but he argues that account
tions.
ability without sanctions-such as can be found in central banks
The. question about the role of sanctions also points to the Im
and in the experiences of truth commissions in Chile and South
portance of O'Donnell's (1999a: 165) insistence on the need for a
Africa-is not a 'diminished subtype' (17-18). Yet these are clearly
'network of relatively autonomous powers' within the state for the
instances of limited, diminished forms of accountability. Chile's exercise ofhorizontal_account.~ility: -- ---
truth commission was limited to describing what happened to the
victims of the Pinochet regime, without identifying or punishing An important but seldom noticed point is that if these agencies are
those responsible, while in South Africa those responsible for to be effective, they veD!'.~~Y_c.an.operate in isolation. They can shake
human rights violations were identified and their actions publi public opinion with their proceedings, but normally their ultimate effect
cized, but they.were not otherwise punished for actions they con i,.y-eness depends 011 deci_sio11_s by courts (or eventually by legislatures will
fessed. (This is not an argument about the political and moral ing to consider impeachment), especially in cases that are salient and/or
involve highly placed officials. ~ffective horizontal account_?.biliD7 is not
character of these commissions, but about the degree of account
the product of isolated ag~ncies bu~of networks.of agenci~hat ~clud~ at
ability exercised.) That these are limited, diminished forms of ac
countability seems apparent from what Schedler himself writes: the_~-~.<?E_.~~~aus~~hat is_~here __~ co~1~t~t!onaJ_~~~te~-~-9_s~~--·~Y
means of ultimate decisions--courts (inc u · g the highest ones, commit-
'exercises of accountability that expose misdeeds but do not impose ted to such accountab@ti-fo'i'.5onneii.-i999b: 39) ..
material consequences will usually appear as weak. toothless, "di
minished" forms of accountability' (15-16). In this sense, I think Horizontal accountability depends on a network of state agencies,
that he is closest to the mark when he writes 'unless - ,._..,
·- there is some but this does not mean that all such agencies must be located along
punishment.for .demonstratecl_abuses_of authotj.t_yt..t4~i::e is ,!lO ru~ .. a vertical chain of command. When children play 'rock, scissors,
paper', with each element beating the next in an intransitive chain
gfla'Y( an<!.P--~.Jl.E.~~UQ.~]l_i1tty'(17). of relationships, what is essential is not hierarchy but a network of
For O'Donnell
----·---·-······ the
·· ···-···'---·· ·· - sanctions - , __
__ , - -- .. , . inherent in horizontal
·····--· --·--·····---"""" accountabil-,.
ity include t!_ie.~~P.~~ity toJ,~~ye_so.m~~ne ~o~_Qffl~e_and~9..._a:v.ply accountability .relationships that leaves no agency unaccountable.
civil and crimin8:l penalges. Publicity can contribute powerfully to· Agencies that lack the power to apply sanctions directly may yet
accountability, especially where information about misdeeds has play key roles in providing horizontal accountability within such a
been concealed, but accountability itself requires penalties beyond network. When the network breaks down, however, these agencies
publicity-at the very least the capacity to remove an office-holder will be able to exercise only limited, diminished forms of horizontal
from office. Without political or legal sanctions, publicity can re accountability.
main at the level of mere allegations that the powerful often find This point is well illustrated by the experience of Peru's ombuds
man's office (Defensoria del Pueblo) during President Fujimori's
ways to manage. second term in office. Created by the 1993 Constitution, the om
A case in point is that of Vladimiro Montesinos, Fujimori's
powerful de facto intelligence chief for the last decade in Peru. budsman's office began to function only in 1996 and quickly became
Many of Montesinos' alleged misdeeds were widely publicized almost the only source of horizontal accountability within the Peru
during the 1990s, but while he remained in control of the agencies vian State. This was true because the executive controlled all of the
that were designed to provide horizontal accountability, his impun other institutions that should have provided horizontal account
ity was guaranteed and his power was scarcely affected. Even after ability within the State. By means of its legislative majority, the
apparently irrefutable evidence of corruption emerged with the executive came to control the judiciary, the constitutional tribunal,
airing of a video in September 2000 that showed him bribing a the attorney general, and the comptroller. The ombudsman's office
congressman-elect, the attorney general's office found that no escaped executive control via the legislature largely because a
crime had been committed. According to the prosecutor, since Mon super-majority· of two-thirds of· congress was required both to
tesinos was using his own money, no misuse of public funds had name and to remove the ombudsman from office. In addition to its
L-1--- -lnno Gnrl ~\TI~P. the Congressman had not yet taken office at
66 Charles D. Kenney
relative immunity from executive control, the ombudsman's will Horizontal Accountability: Conceptsand Conflicts 67.
ingness and ability to exercise a degree of horizontal accountability reasons, we are left with a conceptual gap. 7 In these cases, ~ents of
was due both to the quality of its leadership and to the support the_state are empowered to p9.li_tj_(?al!J: s<:!nction other agents of the
both in terms of material resources and public backing-given by a state for actions or omissions that are not ggalified as unlawful..
· number of foreign governments, including the United States. The =---:---:-:-------.. ------·--~-··-- - --
Clearly, these are accountability relati?ns_hips and_ not 12ower- .
ombudsman's office became a vigorous and widely respected de ~haring checlcs and balances. Clearly too, tliey are hor,=--- . _ ~' •
fender of citizens, rights and a thorn in the side of the Fujimori tionships within the state, not vertical relationships between soci
administration. At the same time, Peru's ombudsman's office con ety and state. In other words, they meet all the requirements for
stituted but an island in a truncated network of horizontal account O'Donnell's definition of horizontal accountability, save one--they
ability, and-without underestimating the political significance of need not be responses to allegedly unlawful acts and omissions.
its actions-the accountability it exercised was limited. The om We are faced, then, with a choice: either we broaden .O'Donnell's
budsman could ·investigate and publicize alleged violations of citi definition of horizontal accountability to include some relation
zen's rights, but it was the attorney general's office and the ships in which one state agent holds another state agent account-
judiciary that were in the end responsible for further investigating abl~!£r acts th~t~~. ~~-~~~!!~ge<f1>_J?!_,~nla~ur,or-welioidonio
and acting on the allegations presented by the ombudsman. In OTionnell's unlawfulness requirement and are left with a set of
other words, the accountability exercised by the ombudsman's accountability relationships that fall outside both the categories
office was a diminished, limited form of horizontal accountability, of vertical and of horizontal accountability. Given the importance of
because the other powers on which accountability depended often lawfulness in O'DonneU's theoretical work on democracy and the
failed to act.
rule ofla,v, the stipulation that horizontal accountability be always
a response to improper and unlawful behavior cannot be set aside
Horizontal Accountability and Unlawfulness lightly. In parliamentary regimes, however, it makes sense to relax
this requirement and include parliaments holding ministers polit
A third disputed aspect of O'Donnell's definition is the scope of ically accountable within the concept of horizontal accountability:
horizontal accountability. ~ argues-th-af""liorizontal ac This move allows us to restate the unlawfulness requirement in
countability should be construed so as to 'hold rulers accountable terms of lawfulness: horizontal accountability involves state actors
_ furtfiepolit1cal··~~~?tlustle~_so~_q~e~s~f their behavior or agencies willing and able to sanction" other state actors and
in office'_ (1999: 60). Under what circumstances might state agen a enc1es-for t1ieir actsa.ilif"oiiuss1oiis'"In-accordance with the la.;
cies exercise horizontal accountability for actions that are not g - ----~------ ...... ---.----·----··--·-·-··----·
aiicttlie- constitution. For presidential
..-1,--regimes, this definition de-
-~
u_aj.!_~).,_~!,lt_9.p.J_y_~li~ Certainl~~gal _ anctions would..-.....
inal?.E!.P..Priate i~such cases. Although it is-
be_,
n difficult to disen
notes the same set of cases as O'Donnell's definition, since allega
tions of unlawfulness are usually required in such regimes for
tangle personal and political motives from arguments over the the application of sanctions. For parliamentary regimes, however,
lawfulness of acts and omissions, ]!!?,e im,Eosition of_legal sanctions. this definition would also denote cases in which state agents
\.vithou£ e~ence ofp.legality would its~e illegal. are held accountable for actions that are politically objectionable,
The question of politicaT sanctions is less clear. O'Donnell's but not illegal, as long as the authority to apply sanctions in
unlawfulness requirement works well within the framework of pre such
tution.cases is recognized as conlorming to the law and the consti
sidential regimes, in which the survival of office-holders in each
branch of government is politically independent of the other
branches, and impeachment requires allegations ofunlawful behav 1
ior. When we consider parliamentary regimes, however, or regimes This problem also arises where legislatures can remove presidents from office
such as Peru's in which ministers can be constitutionally censured for mentaJ, moral, or physical incapacity, without alleging unlawful actions. As
noted above, the Ecuadorian Legislature used such a measure to remove President
and removed from office by the legislature for purely political Bucaram from office i.n 1997, and similar provisions exist in a number of Latin
American constitutions.
Charles D. Kenney Horizontal Accountability: Concepts and Conflicts 69
68
To summarize, then, horizontal accountability shares with other Democratic · Republican Liberal
forms of accountability-such as vertical accountability-the Rule Rule I Rule
idea that some actors may sanction other actors for their acts by the people , in ~he public interest · limited by rights
and omissions. Horizontal accountability shares with other kinds
of governmental self-controls-such as legislative checks and Nature -in practice, the -emphasizes the -some rights are
balances-the idea that these are intrastate and interagency of rule majority difference seen. as
between public inalienable
relationships designed to constrain the exercise of power. For and private
O'Donnell, horizontal accountability's specific difference is that it interests
takes place only with respect to behavior regarded as unlawful,
but I argue that this need apply only to purely presidential
-today, through 1-institutional design -tJ1e purpose of
regimes. In regimes that legally empower some state agents to representatives and moral government is
sanction other·state agents for actions that are not unlawful, such education used to to protect
sanctions would also form part of the exercise of horizontal ac- control corruption rights.
of power
countability.
Vertical Vertical Vertical
Role of accountability of accountability is not accountability is
Accountability, Democracy's Multiple Origins, and vertical officals to the strictly speaking not strictly
the Battle for Democratic Legitimacy accountability citizens is essential necessary speaking
necessary
dent is taken to be the embodiment of the nation and the main custodian and
definer of its interests. The policies of his government need bear no resemblance
9
'lb what extent vertical accountability is actually exercised is a separate ques lo the promises of his campaign-has not the president been authorized to govern
tion. On this point, see Przeworski, Stokes, and Manin (1999). as he (or she) thinks best? In this view, other institutions-courts and legislatures,
10 One might amend O'Donnell's description of delegative democracy to specify
for instance-are nuisances ... a mere impediment to the full authority that the
that such regimes lack not only horizontal accountability, but they are also lacking president has been delegated to exercise' (O'Donnell 1999a: 164). There remains
other forms of governmental self-control In O'Donnell's view, delegative democ the empirical question ofjust which regimes qualify as delegative, a question that
racy is characterized not only by certain institutional relationships, but also by a hi ogcs in large part on the operationalization of the concept of horizontal account
style of leadership linked to those relationships. In his words, 'Delegative democ ability, It should be recalled that the concept of delegative democracy does not
racies rest on the premise that whoever wins election to the presidency is thereby entail the existence of an all-powerful executive, but of an executive that experi
entitled to govern as he or she sees fit, constrained only by the hard facts of <H\CO(l relafivcly lit~le
~q~str1xint (rom ~he agencies of horizontal accountability and
existing power relations and by a constitutionally limited term of office. The presi- <:onstitutionnlly prescribed checks and balances. .
'
72 Charles D. Kenney Horizontal Accountability: Concepts and Conflicts 73
Fujimori in Peru and Chavez in Venezuela have not only been ingly, their actions are also presented and perceived by many as
applauded as caudillos, but have also been able to present them consistent with the republican and liberal dimensions of democ
selves-and be perceived by many-as more democratic than their racy. The minority's cry that the new caudillo is a threat to the
predecessors. For those who see these leaders as the antithesis of institutions of democracy is overwhelmed in the public forum by
democracy-as elected dictators dedicated to subverting the insti the caudillos' argument that the primary threat to democracy
tutional basis for democracy-it may be difficult to understand comes from the fact that its public institutions have fallen under
why many citizens have perceived them as democratizing leaders the control of private interests. These private interests vary, but in
whose first order of business was the dismantling of corrupt popu the case of the legislature they include the interests of politicians
list or partyarchic oligarchies. who pursue ends of a purely private nature (corruption) and ends
The high level of democratic legitimacy accorded these leaders that favour the political 'class' generally, and their own party spe
by their citizens is critical to understanding their success. Un cifically. In some cases, the only interests politicians appear to
doubtedly, this democratic legitimacy mixes in complex ways with serve beyond their own are the private interests of business elites,
more authoritarian-charismatic sources of legitimacy. But what both national and foreign. Judges and other judicial personnel are
interests me here is that in the very act of subverting those insti widely perceived to serve a combination of their own private inter
tutions-legislature,judiciary, constitution-that some see as the ests (corruption), the private interests of those who corrupt them
sine qua non of democracy, these leaders are often seen as more (usually the economicallypowerful), and in some cases the particu
democratic than the institutions they have undermined. One way lar interests of those political parties with whom they are aligned
of understanding this is to see that leaders like Fujimori and or th.at guarantee them impunity. Despite much rhetoric to the
Chavez are claiming legitimacy as democrats in the narrow, major contrary, what is absent in each case is the public's perception that
itarian, sense with less concern for democracy's republican and these institutions serve the public interest in any meaningful way.
liberal dimensions. They can frequently demonstrate that their This perception is especially strong in highly unequal societies in
anti-institutional actions have the support of a large majority of which the immense majority of citizens feel that their interests are
the population, and argue that this makes their actions democratic rarely, if ever, represented or defended. Where private interests
in the primary sense of 'rule by the many'. It is as if the legitimacy are perceived to have hijacked these public institutions, democracy
derived from vertical accountability-with its roots deep in the is fatally undermined, and the way is open for leaders such as
democratic tradition-were at war with institutions of horizontal Fujimori and Chavez to 'save' democracy by radically reforming its
accountability whose roots are more republican and liberal than institutions.
democratic. One dimension of democracy overcomes another,all in The new caudillos-who resist any exercise of horizontal ac
the name of democracy. This is one reason why those who defend countability tending to limit their freedom of action-nonetheless
institutions of horizontal accountability often have such difficulty. present themselves as agents of accountability confronting the cor
The institutions they hold to be essential for modern representa ruption of unaccountable legislative and judicial institutions and
tive democracy are overwhelmed by an executive armed with a political parties. In addition, these presidents present the at
discourse that is more democratic, in the majoritarian sense of the tempts of the legislatures and judiciaries to hold them accountable
word, than the republican and liberal arguments on which these as nothing more than the actions of private, often party, interests
institutions rest. These leaders call to mind Giovanni Sartori's ob attempting to resist those who act to defend public interests. In
servation that 'democracy still has foes; but it is now best evaded doing so they claim not only the moral higher ground, but also the
in its own name and by means of its own name' (1987: 4). democratic legitimacy accruing to those who both enjoy majority
support and who promise to rescue the institutions of modern rep
· The Republican Argument resentative democracy from private interests and return them to
the service of the public interest. ·
The new caudillos' majoritarian support constitutes their strongest
The democratic legitimacy thus gained by these caudillos may
card in the battle for democratic legitimacy,but it is not the only
not endure, but it is critical to their initial success in overwhelming
one. Another reason for the success of such leaders is that, surpris-
the institutions of accountability. For the defenders of horizontal
74 Charles D. Kenney Horizontal Accountability: Concepts and Conflicts 75
accountability to win the battle for democratic legitimacy, it is and because those who raise the republican banner of horizontal
first necessary that they recognize the sources of their illegitimacy accountability have themselves often lost even the republican
in the eyes of many citizens, so that they might act to gain the bases for legitimacy,
legitimacy necessary to sustain themselves in a conflict with the
executive. One critical aspect (that calls us back to the conceptual
discussion earlier in this chapter) is the distinction between polit References
ically objectionable actions and those that are unlawful. To the
extent that agents of horizontal accountability attempt to check Dahl, R. A. 1989. Democracy and its Critics. New· Haven, CT: Yale Uni
lawful actions by presenting them as illicit or unconstitutional, versity Press.
they end up undermining the legitimacy that flows from the proper Elster, J. 1999. 'Accountability in Athenian Politics'. In A. Przeworski,
exercise of their duties. In presidential regimes, at least, the more S. Stokes and B. Manin, eds., Democracy, Accountability, and Represen
the exercise of horizontal accountability appears to be nothing but tation, pp. 253-78. Cambridge: Cambridge University Press.
The Federalist. 1990. The Federalist: A Collection of Essays Written in
a reflection of personal or partisan interests, the less legitimate it
Favor of the New Constitution As Agreed Upon By the Federal Conven
will appear. tion, September 17, 1787, by Alexander Hamilton, John Jay, James
Finally, democratic legitimacy accrues to caudillos like Fujimori Madison. George W. Carey and James McClellan, eds. Dubuque, IA:
and Chavez to the extent that they are able to avoid the appear Kendall/Hunt Publishing Company.
ance of being unambiguously repressive of liberal rights to prop Kenney, C. D. 1998. 'Institutionalized Instability? Questions for Democ
erty, free expression, and assembly. Never paragons of liberal rights, racy from Peru, 1980-1992'. Ph.D. dissertation, University of Notre
they nonetheless must avoid the wholesale repression common to Daine.
most dictatorships if they are not to forfeit an important piece of Lakoff, S. 1996. Democracy: History, Theory, Practice. Boulder, CO: West
this valuable democratic legitimacy. If these freedoms do become view Press.
significantly restricted and if this becomes widely known, this O'Donnell, G. 1999a. 'Delegative Democracy'. In Counterpoints: Selected
Essays on Authoritarianism and Democratization, pp. 159-74. Notre
source of democratic legitimacy will be eroded and eventually
Dame: University of Notre Dame Press.
lost. --. 1999b. 'Horizontal Accountability in New Democracies'. In A. Sche
dler, L. Diamond and M. F. Plattner, eds., The Self-Restraining State:
Conclusions Power and Accountability in New Democracies, pp. 29-51. Boulder, CO:
I have made two points in these reflections on horizontal account Lynne Rienner Publishers.
--. 1999c. 'A Response to my Commentators'. In A Schedler, L. Dia
ability in Latin America. I :first argued for the appropriateness of a
mond and M. F. Plattner, eds., The Self-Restraining State: Power and
definition of horizontal accounta61lity.In whiclitneagents of ac Accountability in New Democracies, pp. 68-71. Boulder, CO: Lynne
c~untability-are-fi:mited-to those within. tlie "state -(and not aclQri_ Rienner Publishers.
Pastor, R. A. 1999. 'The Third Dimension of Accountability: The Inter
ffi!s~de 1~~1i6~~~~~cTI~~?c!:d-~fjii~€~~-J6~~),!liJ1~!1~l::;.
__ pE_! .._ - - -·--·-----··-······----·---··········· - , .. national Community in National Elections'. In A. Schedler, L. Diamond
the
-·--·--li
scope ._of -- accountability
_ - is generally
" ,,. limited
,. ,, -. -.,to actions
- ._ . --·--· o~s-
or --·- - and M. F. Plattner, eds., The Self-Restraining State: Power and Ac
sioi:J.s _!_~-~-.13-:E.~ ~~-~~b._!?~! ~~~-~~-..~· ·~-~P.~ ~Y.~~~~---1:!!aY_!~.clude countability in New Democracies, pp. 123-42. Boulder, CO: Lynne
actions that are not unlawful. Rienner Publishers.
I then explored the ways in which an attention to the demo Plattner, M. 1999. 'Traditions of Accountability'. In A. Schedler, L. Dia
cratic, republican, and liberal dimensions of modern representa mond and M. F. Plattner, eds., The Self-Restraining State: Power and
tive democracy helps us understand the initial success of leaders Accountability in Neu: Democracies, pp. 63-7. Boulder, CO: Lynne
Rienner Publishers.
like Fujimori in Peru and Chavez in Venezuela. These contempor
Przeworski, A., S. Stokes, and B. Manin, eds. 1999. Democracy, Account
ary caudillos often win the battle for democratic legitimacy at crit ability, and Representation. Cambridge: Cambridge University Press.
ical moments because their evidently massive public support Sartori, G. 1987. The Theory of' Democracy Revisited. Chatham: Chatham
constitutes a democratic trump of their more republican opposition House Publishere.
76 Charles D. Kenney
Schedler, A. 1999. 'Conceptualizing Accountability'. In A Schedler, L.
Diamond and M. F. Plattner, eds., The Self-Restraining State: Power
and Accountability in New Democracies, pp. 13-28. Boulder, CO: Lynne
Rienner Publishers.
Schedler, A, L. Diamond and M. F. Plattner, eds. 1999. The Self-Restrain
ing State: Power and Accountability in New Democracies. Boulder, CO:
.............
Lynne Rienner Publishers. PART II
Schmitter, P. 1999. 'The Limits of Horizontal Accountability'. In A. Sche ..............
dler, L. Diamond and M. F. Plattner, eds., The Self-Restraining State:
Power and Accountability in New Democracies, pp. 59-o2. Boulder, CO:
Lynne Rienner Publishers.
Sklar, R. 1999. 'Democracy and Constitutionalism'. In A. Schedler, L. Dia Legislatures, Executives, and
mond and M. F. Plattner, eds., The Self-Restraining State: Power and
Accountability in New Democracies, pp. 53-8. Boulder, CO: Lynne Rien Oversight Agencies
ner Publishers.
Smulovitz, C. and E. Peruzzotti. 2000. 'Societal Accountability in Latin
America'. Journal of Democracy 11, 4: 147-58.
Williams, S. 1794. The Natural and Civil History of ~rmont. Walpole,
Newhampshire: Isaiah Thomas and David Carlisle.
Zakaria, F. 1997. 'The Rise of Illiberal Democracy'. Foreign Affairs 76, 6:
22-43.
~
< •
176 Argelina Cheibub Figueiredo Lessons front the Brazilian Experience 177
sixth CPI is possible but requires a majority vote of both houses; .inY.estigation and application of civil or ~nal san~ns if neces
The establishment of this ceiling reflects the fear of the house lead13 sary. 11ie ~osition of sanctions, however, depends on the deci-
ership that too many investigations could paralyze the legislature. ~!!~ of the l~i!s~a.!~~·~:]f!L~::~e..11~fil.IfJn~ g2i.w.w.~ii~.
The approval and formation of a CPI does not imply that the For instance, the report of the joint CPI that investigated corrup
investigation will be completed. In fact, most CPis never complete tion during President Collor's administration (1990-92) concluded
their work. Some of them are not even installed, that is, the com that the president's behavior was 'incompatible with the dignity,
mittee members are not appointed. Often, CPI proponents may not the honor and the decorum of the position of chief of state' {Rodri
even intend to undertake an investigation. A CPI may be used as gues 1999: 228). This conclusion gave support to the lower house's
currency for political exchange and, thus, its approval alone may be decision to initiate the impeachment of the president. The sanc
sufficient to produce the intended political consequences. A CPI tions that came later depended on the decisions of the Senate and
may also be proposed as part of an individual politician's strategy the Judiciary. Changes in public policy originating from investiga
to establish a record on a particular issue for an upcoming election. tions follow a similar process. A recent example was a CPI's impact
With a ceiling on the number of CPis functioning simultaneously, a on the Ministry of Health's policy concerning the production of
CPI can also be proposed in order to prevent the formation of an generic drugs. In sum, the conclusion of an investigation is an
other one. For instance, if the members of the government coalition appropriate indicator of success since it is the necessary first step
anticipate that the opposition is likely to gather support for a CPI to the imposition of sanctions or corrections in public policy. The
they find undesirable, they can form another one first, just to beat distribution of CPis proposed and concluded by administration is
out the opposition. shown in Table 6.1.
Finally, a CPI proposed by a member of the opposition may be Although the two periods present similar rates of proposition, on
approved but prevented from forming by government parties if average 0.77 and 0.66 CPis per month, they differ greatly in two
they do not appoint members to it. Constitutionally, the compos respects. First, proposal rates vary from one administration to an
ition of a CPI must be proportional to the share of seats held by other, much more in the first period than in the second. This fact, I
each party in the house and, according to internal rules, party believe, is probably due to circumstantial factors that led to the
leaders are responsible for the appointment of its members. This intensification of political conflict and social mobilization. Second,
means that to come into being a CPI is dependent on party leaders, errs were considerably more effective in the first period than in the
especially those belonging to the larger parties, and the governing accond: 57 per cent of the CPis in the former against only 17 per
majority can prevent undesirable investigations by simply not cent in the latter.
taking action. In the two democratic periods, the Brazilian political system has
In sum, the whole process of formation and operation of CPis is possessed the same basic institutional features: a presidential form
pervaded by political conflict and strategic maneuvering by both· of government, open-list proportional representation, and federal
individual members of Congress and party leaders. Various insti". lam. This institutional framework provides multiple entries into
tutional and political factors account for the initiation and even the political system and the appropriate incentives to motivate
tual success or failure of different investigations. In the analysis· mutual checks. In both periods, most governments have been
that follows, I will use the conclusion of an investigation as an multiparty coalitions, as the table with the composition of the
indicator of its success. The conclusion of a CPI implies the ap-. coalitions in the period show (see pp. 193-4). Only the first two
proval by the committee of a report with recommendations of the· democratic governments formed two-party coalitions. Dutra's ad
actions to follow. The 1988 ·constitytion maq.dates ~hat a cotD.IDit:. ministration (1946-50) was the only one in either period in which
for further·
t~~.finaj._r~P._?_1::!b~_fory~.w..d-to._the..Min.ister::ia..Eu.blico the president's party (PSD) held the majority of the seats. In 1988,
' SRrney's party, the. PMDB (Partido do Mouimento Democrdtico
13The concern with the 'slow pace' and other 'deficiencies' of the legislative lJr(J.sileiro), held the second largest share of seats. In both periods,
work dominated the debates (published in the Diorio do Congresso Nacional) that iiho share of seats. held by the allied parties in the lower chamber
precededthe approval of the standing orders in 1989. (I.e., coalition patties ~\-u1.t· support the government) tended to be
Argelina Cheibub Figueiredo Lessons from the Brazilian Experience . 179
178
greater th.an the president's party. There is variation within each
""d "@ ,....._ period, but on average the allied parties held similar shares of
~..., ~ seats (48.5 and 47.5 per cent, respectively). The average represen
::s .s 00
...... <+< 0 tation of the president's party in Congress was larger in the first
() 0 p.
A o r-< C'I:> (C) t- 00
<O 00 c-1 t- t- 00 t- C(I C'1 .-i .-I -qt
o~ ....
0 ...__, p.. l.O -.:!' '<l' 1.0 (C) (C) lO period (27.3 per cent) than in the second (11.7 per cent). Conse
quently, coalitions supporting the president in Congress were
'"O larger in the first period with an average of 75 per cent of the
(I.) >,,.....
'"O ~ (I) seats compared to 59 per cent in the second.
::s ..., ~ C'1 ..-1 00 M .-1 I• During these two periods the governing coalitions were both
'o A $..< lO C':l 00 ~ -.;j< co """
A o C1.l C'1 rl O r-i c<':l
,-f io lO ,-1 co """
o a ;:. . eo.
ooo
. o....loo 00000 made up of parties with ideological differences. The main governing
0 ._,aj
and electoral alliances were composed of parties with differentcon
stituencies: rural and urban-based parties, such as the PSD (Par
~ ,.g "<ii'
(/l ~ bl) tido Social Democr<itico) and the PTB (Partido Trabalhista
O I=: Cd CO CJ> <D CO O .-i ~ t 00 ..-! 00 00 tO C'1
g, :a 0 <N <O <NCC O> t- C'1 t- ~ ~ '<I! ~ tt! ~ Brasileiro) in the first democratic period, and the PFL (Partido da
oo...-100....l....lo o.-cooo.-i Frente Liberal) and the PSDB (Partido da Social Democracia Brasi
.....~ &:: .§, ~
~
leira) today. These two democraticperiods, however, differ radically
'1:1 ,.-., with respect to the legislativepowers of the executive and the power
~ ~ al of congressional party leaders. These institutional mechanisms
cu
Q -5-s .-I 00 t0 <D I t- et:> l.O
lO ,-f reduce the effects of the system of government, the form of state
0) A ::S m ;; oo I co <N a, ............
,-f
o A () organization (federalism),and electoral and party legislation on the
0 ...__, o ....
......
:.:::l functioning of the political system. They endow governments with
i ,g
....... "O ,:; greater capacity to increase cohesion in their supporting coalitions
<I) (l)
and to undertake concerted action. A coalition government possess
-.;j<
00 -.;j<
§ ~g, 's::s
":Ii .... ,:::
1
so
CJ)
so m m
.-I <N .-1
rl -.;j<
l.O
c..:i
,.-;
oo a,
(1'j <.O
oo
!g
O'l
,-f
.-, CQ
oo
r-<
c-.:i m
("';) 00
co
lO
<N
ing institutional mechanisms to overcome internal dissent from
p... <:» r-1 r-< members of the coalition has greater capacity to avoid or to control
l
'<;II
CJ')
cop.gressional oversightaction.a...J~SpJ~ciallyJ.nvestigativeactiyities.
'""" ~
<I)
i
-c:,m
Am
In sum, the two democratic regimes analyzed here .had th~ same
o~ electoralpnd party system liuL<flffered.griatJ.Y-·;&..r~g~rdsthe deci
§'m .... "
10• Cl)
a, ""
Q) P<
al.
.0
lO... ID• si_on-makiog pr~~: If we consider only political variables, such
s; th l,i,
i~
li1> ~0 O C'1 lO Ct;) 0 t- r-<
00
00
.-i
i:-ot-o~<N
..-< «) C'-1 CO 00 ID ~ a ..a,
O
a,
.D.
Q)
Cl)'
:
as thewstribution of seats and the ideological correspondence be
(.() ~ ...... c.o .-i Ct)
c.!:l
o ._, a .g g 9; tween the executive and the legislature, we would expect, first,
§ ""' ~A ~:
O d ;,;; that the parties outside the government would be the most motiv
r-1 .;::i :0
1! ...... (I) ~
ID C.O O Cll ...... ated to request investigations and, second, that within thP- govern
LO • CJ)
~ r-< '<di ~ Cl)
"<f( • ~ (0 (0 0 "'<14 0) ·a ment coalition, the members of the president's party would be the
Cl) ~ Q)
.-I l.O '7 me-Im. ~o ~
fcu . m ~ ....
(I)
$..< least motivated to do so. The behavior of the members of the
~ lO • z6 ~ <l:, ~ ~ ~ . o A c. . ill•
0 governing coalition would also depend on the ideological distance
o ~ !J § ~ i~ (I)
~ .S·rn .gal_'.;
.-1 ~ ~ .... QJ among the parties in the coalition and/or their policy disagree
£ c.o (0 ...> 0... ··~
(J} 0) rl - 0 . ll>;.
~
(C)
1 ~ ~ l .g'":
~
l.O
0)
•
p..
l.O
0) !%< rl ,-i
l!~~
oo
m m
~ m
r-< v, m
~
,..... ~ A•
a,
0) Cl)
"CS.; •
ments. The greater the ideological distance among the parties in
v• .-i (1) r-< ,..0 m>·8', the governing coalition, the more difficult it would be to maintain
r-< • en . ..!<:I (I) P.. ...; • ...; aj
~
t)~()~
d ..ci ..a O ~ ~ l°L! (1) its cohesion~ We would also expect the size of the coalition to be
Q) Cl).
,-i 0~00~
': ~-'~j:q'
~~~A,£~~] >, 8(/J~ "d "d '.:.' ;.,.;
(C) important, de'pending partly on policy and ideological differences.
<I) ~ 0 ..........
.s "@ ::I ::s ~..,
ro~l%<~~,a@.S u .... !.:'.:
~ d A· ':::1. 'l'he cost of mar:i.aging t~ large coalition composed of heterogeneous
I ai3...a ~ .... 0
Cll
h:;:l b.ta O~!:a :;:l ::I O :j
~
!; ~~ !%<~]-§~
Cf.l O 0 Cf.l ~ a .o Cll parties is greate~··;th.t1JiJ:\ sr:(tsl.ller, m.ore homogeneous one.
E-< o A?-0 ~G'Clrn \ .
Argelina Cheibub Figueiredo Lessons from the Brazilian Experience 181
180
Q)
rn
However, institutional variables can reinforce or counteract the, -s • r.Q
Cll Q)
effects of the distribution of preferencesin the legislature. In order s 0 .....
the centralization of the decision-making process in Congress fu::: This is in contrast to other countries, such as Chile, for instance, where legal
restrictions
Alcala) imposed
(1997). -. '.., : by
· ,the
. . mil~_ry-~
· n~~in_ate_d (Siavelis, 1999; and
•• 1
' •q
#:
•··:.• (". • ·1.-
184 Argelina Cheibub Figueiredo Lessons from the Brazilian Experience 185
and oversight roles. The information system serving Congress was The constitution also increased considerably the TCU's oversig_h.t
created during the military regime, and can be seen as a positive, · -~~pacities, as well as the scope of its action. This bureau became
although unintended, consequence of the bureaucratization and responsible not only for the accounting and ·financial supervision
modernization of military rule in Brazil This system has been of. t~~~vernment but also 'for exami,ning...theJegaL9ng__~~onomic
continually improved and today it carries out thorough coverage of aspects of revenue application. This expansion of the TCU's func
congressional activities. The improvements in the organizational tion, as· Pessanna(I998: 21) points out, allows it to overcome the
structure include the growth in the. number and specialization of strict accounting conception of oversight and to move in the direc
the personnel hired for technical support in the consulting bur tion of more ~e.c..tive.£.9.n.tr.o.l.Qygr.... th~. .J!~y~!QRI!le.nt._of_goy_~~.n
eaus of both legislative houses. Besides that, today, the organiza tal pr.oj.e.c..t.s.
tional support provided in both houses is predominantly connected AB far as sanctions were concerned, the constitution was also
to the work of committees contrary to the previous emphasis on .innovative in conferring on the TCU the ability to determine_fi!?&§.
assistance to individual members of Congress." propcrtional to the amount of the damag,(il_J:;.1~J.1sed.tQ ..pu..'bli.r,... fun.d$.
However, I will show that despite the legal, informational, and The.. TCU'saecisioiisregardini.fhese·fii:i~§s.5!.n....be.Jm.m~.diately...elb
organizational apparatus available..1. direct and routin.e ovei:sight is for~~d.(F~the constitution institutionalized broader partici
not carried out by Congress. The success ~~-9_f overID.ght_initi& pation in-oversight by ruling that 'any ~itizen, political p~,
''fives association or union may !eEQtimatel.y_..2ut fo.!}!ar<;!._denunciations
--·----=::-
is much lower than 1he rate -at...._which they are initiated.
-- .
Q!J:lJL~fulor 1mdue action before the TQJJ' (article 74, 2). This
Central Control of Government Accounts measure is obviously not self-enforced since it depends on various
factors external to the agency, but the TCU has taken some initia
The Tribunal de Contas da Unuio (TCU) is responsible for routine
tive to make it easier for individuals or groups to file complaints.
oversight of government accounts. This bureau has nine ministers
Notwithstanding these favorable institutional regulations, ef
and is staffed by public career personnel. Although formally belong
[ective change in the supervision of federal accounts has been
ing to the legislative branch, this agency has from the outset been
slow. Reports on governme~t accounts are in general favorable
strongly associate.cL.mth_the president, who, according to previous
:,,yi.th specific criticism. The TCU's reports are presented on time,
regulation, nominated all its members." Senate approval of presi
but congressional approval sometimes occurs long after the report
dential nominations to the TCU w~s-r<;g!.!.ir_¢ J~..Y~»-_tbmJghthe.Ee l}_as beeJ?. publicized, and alyv.Jl_y~..b.Y...unanimous...YQre, Approval of
.!!!'-.§..1!~~:tJ?.~e.!l_~_c~~...2!.-'f.~jecti~n.
The 1988 CQ;t1$.tjtJ,ttJ9J}....dimiP., the government's accounting after the end of the president's term
-~~h:~.~_th~_.'!'Q!!.~~-~~.P..~~4~~I.09,_t~~._ex~:ggve. Today Congress has has not been unusual (Pessanha 1998). A positive change is the
the prerogative of nominating two thirds of the TCU's ministers ..
fact that the TCU's_~.Yit]&s_se.em-inde.ed..tQJ).e... mar.e.-in.tegrated..
On the other hand, the new constitution mandated increased tech
nical capability (expertise in finance, accounting, economics, and. ~~t~-~~e__ wo~...Qf_.fug_!!!_and.i.Qg_c2mmiti&f,'l~,... as•.iadicated,...for..in
public administration) and proven experience (over 10 years. o( st~~~~..?.....n.x.JJi~_,ip._c:r:e.as..e.i.J,1,_,e.o.J.n!llj~~~-~re.9.~~!~--fQI.J~QI!l?.W.i~tiQP,.
I-Iowever, the improvements in the TCU's technical expertise and
professional activity in those areas) in candidates for a ministerial
capacity for auditing government accounts have outpaced its cap
post at the TCU. Two of the ministers nominated by the president
ucity to impose policy changes and sanctions against the misuse of
must be chosen from the office's career personnel, a condition that
public resources. The recommendations contained in the reports
has restricted the president's choice even more. Further legislation·
prepared by the TCU's technical personnel are usually not followed
and internal regulations corrected previous operational deficieti{
hy its board of ministers for political reasons. One example is a
cies and strengthened the TCU's links to Congress, and especially.
recent scandal related to the construction of a building for the
to its committees. - Labor Courts in Sao Paulo, which was disclosed by a CPI formed by
18
the Senate. The chief justice of the tribunal was proved to have
See Diniz (1993) for a study oftheA.ssessoria Legislatiua.
10
The following paragraphs on the TCU rely heavily on Pessanha (1998). Besides embezzled more than 200 million Reais (over US$50 million) from
an account of the institutional changes in the TCU, Pessanha provides a detailed : the construction of tho building. Irregularities in the expenditures
analysis of its composition, showing its increasing independence from the executive.· for tho construction .had been detected and the TCU was notified
Lessons from the Brazilian Experience 187
186 Argelina Cheibub Figueiredo
ing offices, i.e., the technical staff belonging to the organizational
in 1992, but only in 1998 did the TCU's board of ministers decide
structure of each legislative house. Their work, however, is
to include that construction in a list of illicit public works. In directed primarily to aiding individual members of Congress. The
the course of the CPI, a PMDB senator was also proved to be
opposition, especially the lea-wing parties, has had an important
involved with the construction of the building and, as a result of
role in supervising the implementation of the budget and bringing
public outrage, he was the first senator in history to lose his man-
it before the public for discussion. In addition, as the PRODASEN
date. system can be accessed on the Internet, associations, unions, inter
est groups and NGOs have increasingly participated in budget
Committee Oversight oversight.
Committee oversight takes place under different institutional for
mats. The supervision of budget implementationis carried out by Standing Committee Oversight Activities The standing commit
a large and centralized congressional committee, the Comissiio tees undertake oversight activities through the following instru
Mista de Orcamento (CMO), composed of members of both houses. ments: public hearings tAudiencias Publicas-AP); proposals for
The standing committee system carries out other kinds of over- oversight and control (Proposta de Fiscalizaciio e Controle-PFC);
sight activities, as well as legislative activities. convocation of ministers to provide information regarding policies
in committee meetings (Conoocactio de Ministros-CM), which can
The Comiseao Mista de Orqamento (CMO) and Budget Over take the form of a written requirement or an informalrequest; and
sight Control over budget execution is crucial because budget ap requests for information from ministries and state agencies
propriations approved by Congress are not mandatory and (Requerimento de Informaqiio-RI). These oversight mechanisms,
regulations for the execution of the budget allow the government · previously regulated by the houses' standing rules, acquired con
plenty of leeway in reallocating approved budget items. Congress stitutional status in 1988. An individual member of Congress may
can also pass legislation that modifies the approved budget bill request any of these oversight instruments but initiatives have to
through extraordinary transfers of resources from one budget item be approved by the appropriate committee.
to another. Consequently, the CMO must function continuously .. Table 6.3 shows the number of these oversight activities on an
This committee has its own support staff. Moreover, the Senate annual basis from 1989-99. There is one common feature in the
and the Chamber of Deputies have separate consulting offices, distribution of all these oversight instruments: the usage increases
with technically trained personnel who are responsible for moni considerably in the first year of each legislative session in the
toring the release of resources by the executive. Budget monitoring · period-1991, 1995, and 1999. This same pattern occurs with re
by the CMO relies on a data base kept by PRODASEN (Centro de spect to the rate at which legislation is proposed by members of
Informatica e Processamento de Dadas do Sevado Federal), the Congress. And as we will see below,just as legislation proposed by
Senate data processing system. Its data base contains complete.' members of Congress (rather than the executive) is not likely to
information about each one of the thousands of budget items. lit pass quickly, the success rate of oversight activities by Congress is
addition to information about the nature of each program, activity,·'. much lower than the rate at which they are initiated.21
place of application, function, and ministerial jurisdiction, it also Public hearings (Audi.encias Ptiblicas) combine oversight and le
contains the specific amounts that were in the original executive gislative functions. They are special committee meetings in which
proposal, the amendments introduced by members of Congress; experts or persons with connections to the issue under discussion
the amounts approved, the reallocations mandated by furthei speak about the subject matter. These meetings can occur either
budgetary laws, and the monthly disbursement for each item." ·. · during the formulation of new policy or the review of existing
The CMO does not undertake routine budget supervision. Moii~/ policy. The same· meeting may serve both legislative and oversight
toring of the budget implementation is undertaken by the consult-, purposes. Since the records do not break down the nature of
the hearings, it is not possible to estimate the amount of time
w This information is available on the World Wide Web for budgets from 1995 ~ :.. , O~ the. lQgialntion see Figueiredo and Limongi (2000).
the present..
'
.,..
188 Argelina Cheibub Figueiredo Lessons from the Brazilian Experience 189
TADLE 6.3. Committee's OversightActiuities-1989-99 the left wing parties; they initiated 48 per cent of the Rls ap
proved. The remainder were requested in similar proportions by
Year Public Request for Convocation Proposal for the members of the center and right wing parties (Almeida 1999:
hearings information of ministers oversight and 26). Most requests for information had to do with the management
(PA) (RI} (CM) control (PFC) of state agencies (54 per cent); 30 per cent concerned information
on social policies, and 16 per cent on the infra-structure (Almeida
1989 124 328 9 1999: 26).
1990 15 285 8 Table 6.3 shows that 106 convocations of ministers were re
1991 164 1108 20 37 · quested in the lower house. However, only four were approved by
1992 84 889 3 35 the committees and actually took place. The great majority (70 per
cent) have not even been considered by the committees. On the
1993 105 868 4 19
other hand, informal requests for information were more fre
1994 47 525 3 11 quent. 23 This form of communication is in fact a means of preempt
1995 205 1319 21 40 ing formal convocation and is preceded by an agreement between
1996 96 864 6 17 party leaders, usually of the parties that support the government,
1997 144 953 10 30 and the minister. Formal convocation of ministries is basically a
1998 58 945 8 13 strategy of the left wing parties. Although they have never held
more than 20 per cent of the legislative seats, they are responsible
1999 279 1495 14" 19
for more than 50 per cent of the requests. This is particularly true
'lbtal 1321 9579 106 221 of the Worker's Party (PT), which has been responsible for 30 per
cent of all convocations. The members of the president's party
" Thirteen are still in progress. rarely use this strategy. Since, throughout the 1988-99 period a
Source; Banco de Dados Legislativos, Cebrap. coalition of right or center right parties have dominated the gov
ernment, it is not surprising that communication between Con
gress and the cabinet has been more informal.
dedicated to either one of these activities. However, unlike the
other instruments of oversight, we can presume that the sheer Proposals for oversight and control (PFC) have had an even
occurrence of these hearings be taken as an indicator of their sue tower rate of approval and execution than the other instruments
of oversight: only five out of 221 proposals have been approved and
cess."
The standing committees have approved a great number of re carried out. This is not a low cost initiative, since the PFC must
quests for information from ministries, but it is difficult to assess include a justification, a plan of execution, and a methodology for
evaluation. Proposals for oversight and control have also been pre
the consequences. In general, whenever Congress initiates a re,~.'
quest for information, it gets some response, either formal or infori dominantly used by left wing parties which have been responsible
for 52 per cent of the total requests.
mal. But when too many requests are referred to the same
ministry they are usually ignored (Soares 1999). This happens.
despite the formal requirement that the information be provided;
lnstruments for Punishing Executive Authorities and
within 30 days; if not, criminal charges can be brought agafinit.'
Suspending Legislative Acts of the Executive
the Minister responsible. There is, however, no reported case o.( The main punishment Congress can impose on authorities in the
charges as a result of non-compliance. Most demands for informa-' executive branch of the government is their removal from office.
tion from the ministries and state agencies come from members ·.~t Constitutionally the lower chamber is responsible for authorizing
·.
cri minal proceedings against the president, the vice-president and
:.z.:i The data available does not allow a distinction between the two. For this i« Although r de> not huvo (>rt1ehic ligurcs, information in the press suggests that
distinction in the United State see Aberbach (1990: 132-40). tlt!t'l often happens." · · · ·
190 Argelina Cheibub Figueiredo
ministers, for crimes of responsibility; the Senate prosecutes and Lessons from the Brazilian Experience 191
judges them for these crimes. The process can result in the im ,l.&46 ruirnocrac,t anclthe..p.o.~!~~~ei:~2-<l.d.~.mQ.~trates that parti
peachment of these authorities, but even when the whole legal ~ con~!--1:!Pd!ariation in gov~~e~~.§~££2.rt.do g9J_l!,~_ym_J..o.r
process is not completed the political removal of ministers is pos differences in the effectiveness of congr~.$..$._~Q)J~.L~ti.Ql.1.23 The mul
sible as a result of congressional action. tiple enine's-as"suredby the political system, a common feature of
In 1992, President Collor was impeached in a process growing both democratic periods, provide the motive for launching oversight
out of a CPL Formal removal of a minister has not yet happened as actions, but today's institutional structure is not conducive to their
a direct result of one of these oversight activities, but political re success. Legislators ~an be motivates!_~.Q..~~ight bu~__they_d~.B_~
moval has occurred as the result the withdrawal of legislative of ~on~. the means to exert eff~ti.Y..~_g_l!§ga on the_exe..c;:uti_ye_.20
support, In some cases, public testimony before congressional com The analysis presented here calls attention to institutional vari
mittees have played an important role in making it clear that a ables usually neglected by studies in comparative politics. A
minister lacked legislative support. One example is the dismissal of centralized decision-making process based <m.J;l!_e institutioµal_pre
the Minister of Communications, Luiz Carlos Mendonca de Barros, ]2.~g-~~§.mpreduces the role of.
rogatives of the ttrelti_C!~.P1.!~£!~P~Y
in 1998 after his testimony before a Senate committee regarding _Qong1:essas an agency of hoE_i~n_~I accountabJ¥~ 27 The Brazilian
his conduct during the privatization of the federal telecommunica case shows that the effect or this variable is not reduced by the
tions agency.
nature of the electoral and party systems. It suggests that ~ll. .~1-~.
Besides the removal of public officers, the 1988 constitution allows Qf..cQroJlP..tati@_bg_t.weenagfillda~tt~r~~.m~j~_th~fl~ifilQ.:g.-m.~l<:mg.
Congress to suspend legislative acts of the executive that exceed its system ~~~~t_~mpe~BQ.n.-il~!:i.Ying_fm.m.~tb.e_ru,~~r~.L~~~
power of regulation or the limits of legislative delegation. It also and party fragmentation.
allowsjudicial review of both congressional and presidential legisla Following the prevailing view in comparatj.ve politi~s, Moreno,
tive acts through an 'action of direct unconstitutionality' (ADIN) Crisp, and Shugart sustain, in this volume, that the proper function~
that, according to legislative rules, can be initiated by the speaker, ing of horizontal accountability depends on 'g,e.ttmg vertic~ac.c.o..un..t
the committees or by individual members of Congress. ..ability right.' The conditions that produce or enhance vertical
Congress has never used its power to suspend legislative acts of ~countability in turn depend on the electoral and party systems. 28
the executive. Most ADINs that argue the unconstitutionality of · I.would like sugges) that congressional failure to countervail ex
to
the executive's legislative acts have been initiated by left-wing ecutive action might also affect.v.ei:.t:i.c.alJl~Qun.tabili.ty,Since mech
parties; they account for 93 per cent of the 141 ADINs proposed by anisms of horizontal accountability compel the executive to justify
political parties. 2-4 The success of these actions depends on the deci and defend its actions before other government branches, they also
sions of the Supremo Tribunal Federal (STF, Federal Supreme serve to inform citizens (Przeworski 1996: 32). Concentration of in-
Court), responsible for judging them. Until 1998, only seven §titutj.onal authoricy_r~Jl.the.~ibi.J.ity_of 11qblic ~-...; ..........
--.... . I =-1.. . -..__..
deciajo-!!§..~d
ADINs proposed by the leftist parties had been accepted for con ~eprive~_citizens of_!he_opp2._tl!yrityto O.Qtain ioforroation_abQJJ.!.P.2!:.
sideration by the court. 1
~' theie6y redu~l!_g-_~E:!!. ~3:J?~Cl~ZiQ.£.cmk~rutY.~r~~fil.actiws.
Concluding remarks is In this volume, Manzetti and Morgenstern emphasize the role of partisan
differences
26
between the executive and the legislature.
Perejoh.n(1999) proposes a model in which authority increases with greater
The 1988 constitution empowered the legislative branch with., accountability, but he is concerned with the agents' interest in making themselves
greater oversight ability than it had had before. However, there i$,. ~· accountable in order to get the principal to trust them with more resources.
difference between the number of oversight actions proposed· arid "7 In his analysis of congressional oversight in the Unit:ed States, Aberbach
their rates of conclus1oa"Tliecomparison··of"iii.vestigatrons (990: 53-73) relates increase in oversight to decentralization of legislative organ.
(ir.:ition.
', ...__ .,._ .. __ .. ,.... ;.•.._ -·- , , ·~ ••• ,,..\,.),,r••.•,..,,,,....,,-. .....
,..,,,,lil"\",'I'" ......... ,.,,.,,.,,\•·••""'!,,,..,,"',~'""'~ .....-
in ~ ,,,:,.,-
24 ~" The authors criticize the use of the concept of horizontal accountability to
Of the 336 ADINs proposed by political parties from 1988 to 1998, 141 r~ refer to non-ttiera.rchicnl rclutions between agencies. For this reason they adopt
ferred to the federal executive. Vianna et al. (1999: 98~9): :. v •,
. Lbe term 'horizontal exch.aogo,' but this is defined quite similarly to O'Donnell's
definition of hodion.tril uccountt.bility, (1.997).
.I,.
192 Argelina Cheibub Figueiredo Lessons from the Brazilian Experience 193
This, however, does not imply that the oversight activity of Con (fl
of Congress to 'fire alarms' (McCubbins and Schwartz 1987) and IO 00 ro CX) ..-4 C'I (N ...,. (N c lO
8. :e OC! ~
0. tll (J') ~ 0 .... ,-.! (0 -tj< -.:!' t- lO 0 .....
0 0. C"J ..... IN ..., ..... ..-4 C,:) er;, rl lN '<I<
allows the flow of information to groups and individual citizens,
....ro
who then activate other mechanisms of accountability. Thus, the ,,,
Q)
organizational and informational support structures established in
Congress can enhance the indirect role of Congress in fostering 1 !lc,
government accountability. -rj '11
(I) Q)
c ...,. 0) -qt 0 00 cq lO rl 0 ......
:::! ,,, """. tq
Congress can also play a role in establishing the conditions for t-= 0) <.O .... ..... (0 .....i ..-4 0 '<!' . 0:, CX) 0
~~ (N lN t- t-- t- C,:) eo Ol (.D l.Q 00 -.:!'
the triggering of 'fire alarms' in a wider network of institutions. 25
One final case in point is the establishment by Congress of a system .s
~Cll ~
of rules and procedures regulating public civil action by the Minis ~ (fl
,ii·t
'¢ l:O O .-f CO '<!< LO (0 1 ..-i ~ (N (N <N -o:i
-.:f< LO t::: LO ,_. u:, :::: lO \0 \0 c.., <O ...... «:) t0 .....t0 «:)
i\.~ '° 1:i' m cii i:i
a1 ::l ui al
......as .....;o ::::
as ~ ......as a,
A as :S::l a;ci o;;,:
j\~I
t6 OJ
h O !:I ~ hr
:::10) ~.-,
:::iJ ::;iJ
~1 ~co
~1 o'<I'~ Jo '1'° :x: '"?dJ ~li1~l~~if
,..... :3 ,..... ::I lN ·C\I
~r: i::
a, Q 'St' Q '<I< Q l.O >O
;S l..O :§o >n LO 10 ~
J <:.0 0 <:o O (.D C:0
::IO ro (D
0:, 0) 0) 00) OOl 'M(j) Ol • 0) '..JI Ol rk Ol rk (j) 0) rk a:,
irJ •,:i 2 .-I 8 .-, 3 .-i ;~ .-I :a~ ,-i r,:. .-1 ::l o-< ID ..-4 0.-I '-'..-4 '-'-.-l .-I U.-1
' • ' . ,a, . Q) • <) •
-~· 0 • 0 0 • 0 •
li{ (1) I-< CJ UJ
'§ ..0' .§ .;,,. -~. ·~ Po, ,., ..Cl • §• 'o-i p. .0 .a ,a:s p. <a.I r:-1 •<ll p. •<XI d
11(
·-'" <ii O·"' · uJ ru w (!} (I) (I) < (l) o<ll o;:, O(I) oro
,. d:: µ'.l µ,, . l'.:il z 0:1 . . i:!, i:t.,·. 0 '°";) . 0 (I) z r:::i .; IL, >-:, P't ~ Cf) t-:, t-:, >-:, U) t-:, t-:,
'
.
194 Argelina Cheibub Figueiredo Lessons from the Brazilian Experience 195
~ References
Q)
A en
0 .,s,
:0 <>~
'$ ~
Aberbach, Joel. 1990. Keeping a Watchful Eye: The Politics of Congres
o]
Ac ' 00 (N
~ .... co ,-(
~ e- t- O'l 00
sional Oversight. Washington, DC: The Brookings Institution.
0. .... ...; C") CN 0) 10 0 '
O p. C") (N -,j, tQ ~ ~ co ~ C'1 ~ ~
Alcala, Humberto N. 1997. 'El control parlamentario en Chile'. Contribui
M ciones No. 4. Santiago: Fundaci6n Konrad Adenauer/CIEDLA
"" .a
(l)
C) Almeida, Francisco Inacio. 1999. Ideias para o desenvolvimento do man
~
..... rt} dato parl,arnentar. Brasilia, DF: Instituto de Estudos Socioeconomicos
0
ti > (INESC).
Q.~ :t:J
"O .,, Cl1
Q) Q)
IO 0) t- 0) (N ...... <O lO
Amorim Neto, Octavio. 1995. 'Cabinet Formation and Party Politics in
~ Ul O'l O'l v.:'Sc
O') 0) t c,:i ...; <O co t Brazil'. Paper delivered at the meeting of the Latin American Studies
;a~ co ~ rl "<I' ~ lQ °"s "<!' lQ .....
'° .s Association, Atlanta, GA.
.g. --. 1998. 'Of Presidents, Parties, and Ministers'. Ph.D. Dissertation,
as University of California, San Diego.
-'° i Cl
i:i to Amorim Neto, Octavio and Fabiano Santos. 1997. 'The Executive Connec
~
~l:l'i! tion: Explaining the Puzzles of Party Cohesion in Brazil'. Paper pre
co 00 0 ,-j ,.... IO t-
""". oci ,..... sented at the XX Congress of the Latin American Studies Association,
t- <O CX) 00
'it
d:: Ac
00
(N C'1 (\') ~ ,-I ,-( ....
!
i:Q, Guadalajara.
~ Arantes, Rogerio. 1999. 'Direito e Politica: o Ministerio Publico e a defesa
0)
. 03·
gJ p.. re ......
._, · · · dos direitos co1etivos'. Revista Brasileira de Ciencias Socia.is 14, 39.
C)
:.!:! 'l'pq Pi ~ ;:l Cain, Bruce and W. T. Jones. 1989. 'Madison's Theory of Representation'.
g g pq pq l e;s·
.g 0 0
l§ In Bernard Grofrnan and Donald Wittman, eds., The Federalist Papers
0. ~ rn ~ ~ 8·
.,...
-
gJ 0 and the New Institutionalism. New York, NY: Agathon Press.
s ~ ....:i Pi '1 '1 Z.
:0 pq p:i
..... 00 .......
3 '1pq pq ........
·~ Carey, John M. and Matthew S. Shugart. 1998. 'Executive Decree Author
00
6 !
Cl o·
al
0
~ Ol O> ity: Calling Out the Tanks, or Just Filling Out the Forms?' In John
Q .... 0) .
u ~ ~ ~ ~
..:I ~
'1 '1
; ,-( .
.....,
Cl)
Carey and Matthew S. Shugart, eds., Executive Decree Authority:
§ 'l' ~ l l l .s·. Calling out the Tanks or Just Filling Out the Forms? Cambridge: Cam
I ~ ~ E-<
pq ~ ~- bridge University Press.
00
·::0 a~ ~'l'
la I>
l l ,....... Diniz, Simone. 1993. 'A Assessoria Legislativa na Camara dos Deputa
~ ~ ·-
c, §i ~ ~ p.. p.. ~ ~ ~ lO;:
~ ~ oo., dos'. Sao Paulo: Documento de Trabalho, Projeto: Terra lncognita: fun
mI··
d?,r. cionament-0 e perspectivas do Congresso Nacional, Cebrap.
JJ! Ferejohn, John. 1999. 'Accountability and Authority: Toward a Theory of
;.:l\
t o•· Political Accountability'. In Adam Przeworski, Susan Stokes and Ber
Ac
gj· nard Manin, eds., Democracy, Accountability, and Representation. Cam
"' . 53·.: ' bridge: Cambridge University Press.
"O pq pq i::l. ~
i:Q
fil..... ! i:: l('.)·. Figueiredo, Argelina C. 2000. 'Government Performance in a Multiparty
en c, cu - ...
Q) -4,,> -"·,
~ ~ ~ C'),
p.. p.. A. "C ....... Presidential System: The Experiences of Brazil'. Paper prepared for
d:: ~ ~ z A '-"
2 .fr the XVIII World Congress of Political Science, IPSA, Quebec, 1-5
........ Q) (l)· August.
~ 0) Cll z,
;1; ,..... C\I 1:1 c,:i 00 '<I' 00 m
~ 1-, Ol 1-, 0) C') ::::: C1>~ 0) ,.... ' A(l) ~-
:
(•'igueiredo, Argelina C. and Fernando Limongi 1999. Executivo e Legisla
0 Oo·o .. ,....,,......, ...... '
,_. CJ) :::::: M ;:::l p. b.o H , H O H II) tivo na nova ordem constitucional. Rio de Janeiro: Funda,;:ao Getu.lio
s ::> ~ ,.....,oioQ):::, 8 ()Q)O~oG>oQ ~ oi
...... ~. 1iJ O I 0~ 1 ~'T .94l ~'T ~ch Vargas..
...., ~~ s:l~ oo oe!t c,:i 1-ic,;i
"',6 "dco "Ca, :a ~
Q). '
'3 <O 1ij I 't:! ~ "Cl O'l ,..... a, ~ Ol ~ O'l !ij a, ~ 0) :>.. (<) ·-. 2000. 'Pre.s.idential Power, Legislative Organization, and Party Be
~ om
~ o ,-( oo~ ~.-1 ~~ 8~ 1ij~ O~O~O'; .... cu havior in Brazil.' Co1nparative Politics 32, 2: 151-70.
<1) ~
~
!
.....
rn 'Q) ~ i.. ~ ..;
fl ,_; fl A P. ::r:l ' ::r:l >, ~ § ] .!:l ::i
.ti ,g §] <nuQ)Cll<l)o..,..uti!G> s:l Cll ~o .Clrunilton, Alexande1·,Jnmcs Madison, and John Jay. 1961. The Federalist
~ 0
~ J: ~ ... ~ ~o ~::S ~< ~o ~rn r,:;~ J:r.i~ r:r,; .. e:,:;. • Cl.l Papers. New Yqrk, NY: NQW Ame.dean Library.
196 Argelina Cheibub Figueiredo Lessons from the Brazilian Experience 197
Hippolito, Lucia. 1985. De raposas e reformistas: o PSD e a experiencia Siavelis, Pet.er. 1999. 'Disconnected Fire-Alarms and Ineffective Police
democratica brasileira (1945-1964). Rio de Janeiro: Paz e Terra. Patrols: Legislative Oversight in Post-Authoritarian Chile'. Journal, of
Huber, John D. 1996. Rationalizing Parliament. Cambridge: Cambridge International Studies and World Affairs 42, 1: 71-9.
University Press. Soares, Rosinethe Monteiro. 1999. 'Instrumentos reguladores da a{!ao fis
Lauvauz, Philippe. 1988. Parlementarisme rationalise et stabilite du pou calizadora do Congresso Nacional e do Tribunal de Contas da Uniao.
uoir executif quelques aspects de la reforme de l'Etat confrontes aux Unpublished paper.
experience etrangeres. Brussels: L'Universite Libre de Bruxelles. Vianna, Luiz Werneck, M. A. Carvalho, M. P. Melo, and M. B. Burgos.
Mainwaring, Scott and Matthew S. Shugart, eds. 1997. Presidetuialism. 1999. A judicializaeiio da politica e das relacses eociais no Brasil. Rio
and Democracy in Latin America. Cambridge: Cambridge University de Janeiro: Revan.
Press.
Manin, Bernard. 1997. The Principles of Representative Government.
Cambridge: Cambridge University Press.
McCubbins, Mathew and Thomas Schwartz. 1987. 'Congressional Over
sight Overlooked: Police Patrol versus Fire Alarms'. In Mathew McCub
bins and Terry Sullivan, eds., Congress: Structure and Policy.
Cambridge: Cambridge University Press.
Nicolau, Jairo. 1998. Dados eleitorais do Brasil (1982-1996). Rio de
Janeiro: Revan/IUPERJ-UCM.
O'Donnell, Guillermo. 1997. 'Horizontal Accountability and New Polyar
chies'. Paper prepared for the conference Institutionalizing Horizontal
Accountability. Institute for Advanced Studies of Vienna and Forum for
Democratic Studies, Vienna.
--'-·. 2000. 'Further thoughts on horizontal accountability. Paper pre
pared for the workshop Institutions, Accountability, and Democratic
Governance in Latin America. Kellogg Institute for International Stud
ies, University of Notre Daine, May 8-9.
Pessanha, Charles. 1998. 'O controle externo do Legislative sobre o
Executive no Brasil: 1946-1998.' Paper presented at the Annual meet
ing of ANPOCS, Caxambu.
Przeworski, Adam, Susan Stokes, and Bernard Manin, eds. 1999. 'Intro
duction' in Democracy, Accountability and Representation. Cambridge:
Cambridge University Press.
Przeworski, Adam. 1996. 'A Refonna do Estado. Responsabilidade Politica
e Intervencao Economica.' Reoista Brasileira de Ciencias Sociais Ano
11, No. 32: 18-40.
Rodrigues, Alberto Tosi. 1999. A Democracia que nos coube: ,nobiUzcu;iiq
social e regimepolitico no Brasil (1989-1994). Ph.D. dissertation, Cient
cias Sociais, Unicamp. :·
Santos, Fabiano. 1997. 'Patronagem e Poder de Agenda na Politica Brasi
leira'. Dados =Revista de Ciencias Sociais Vol. 40.
Santos, Wanderley G., ed. 1990. Que Brasil e este? Manual de Indicadores
Politicos e Socials. Rio de Janeiro: Vertice. .
Shugart, Matthew and John M. Carey. 1992. Presidents and Assembliee:
Constitutional Design and Electoral Dynamics. Cambridge: Cambridge
University Press.
....... :
•••••••••••••••••••••••••••••
7
...............................
204 Maria Tereza Sadek and Rosiingela Batista Cavalcanti The New Brazilian Public Prosecution 205
order, the democratic regime and inalienable social and individual inter entire period prior to 1988, members of the Public Prosecution were
ests. (Federal Constitution of Brazil, Art. 127) : . held in lesser regard and prestige in comparison to fellow mem
As the above citation demonstrates, the constitutionalattributes bers in the judicial branch, despite the fact that both worked
of the Public Prosecution are quite extensive and ambiguous. By together to enforce law and justice.
virtue of being assigned the role of becoming a defender of Most academic papers and press releases published prior to
the 'juridical order' and 'democratic regime', the 1988 Constitution 1988 associated the Public Prosecution only with questions of
effectively allows the Public Prosecution to intervene within a criminal law. The members of the Public Prosecution were seen
wide array of issues, and public policy arenas. Such a degree of only as figures responsible for :filing criminal charges.
discretion is further made evident through listing a few of its insti In addition to its traditional role within the criminal sphere-to
tutional responsibilities, which are also detailed in the Constitu investigate penal infractions and ascertain the nature of the of
tion: fence and the offenders-previousconstitutions defined the Public
Prosecution as the institution chosen to legally defend the inter
• to ensure effective respect by the Public Authorities and by ests of the State and its respective officeholders. Today the Public
the services of public relevance for the rights guaranteed in Prosecution's role has changed radically. For starters, the
this Constitution, taking the action required to guarantee Public Prosecution has achieved independence from other bran
such rights; ches of government-.it is not subordinate to either the Executive
• to institute civil investigation and public civil suit to protect or the Judiciary. Such an institutional transformation, which cul
public and social property, the environment and other diffuse minated in the 1988 Federal Constitution, was accompanied by a
and collective interests; large increase in the Public Prosecution'sjurisdiction. In addition
• to institute action of unconstitutionality or representation for to maintaining its ability to prosecute in the criminal arena, the
purposes of intervention by the Union or by the states, in the institution was granted the responsibility of defending collective
cases established in this Constitution; and diffuse rights (e.g., those of the environment, the consumer, and
• to issue notifications in administrative procedures within its properties of artistic, aesthetic, historical, touristic, and scenic
competence, requesting information and documents to sup value) and protecting minority rights (e.g., those of children and
port them, under the terms of the respective supplementary adolescents, the disabled, the elderly).
law (Federal Constitution of Brazil, Art. 129, Items II, III, rv, Due to this tremendous increase in jurisdiction, the Public Pros
and VJ). ecution has the ability, for example, to file lawsuits demanding
Such responsibilities stand in sharp contrast to the Public Pros that the State provide education, health services, or shelter for
ecution that existed prior to the 1988 Constitution. Previous to homeless children. In the role of overseeing the public administra
1988, the Public Prosecution was almost exclusively confined td tion, it can :file lawsuits of unconstitutionality against existing
criminal law, yet subsequent to 1934 it did come to assume an laws or against normative acts promulgated at the federal, state,
auditing or oversight role that went beyond merely criminal or municipal level.
matters (custus legis). It became its duty to see that the laws were The Public Prosecution wields a powerful juridical instrument to
faithfully executed and enforced. Its intervention in the judicial guarantee diffuse and collective rights:" the A~aoCivil Ptiblica, or
process aimed to protect what were considered indispensable class action suit, created by Law 7,347 in 1985, constitutes
values and social interests, such as legal disputes concerning the primary instrument used by the Public Prosecution to defend
family, marriage, birth and marriage certificates, and the defense such rights.' Despite the fact other public institutions and civil
of the helpless. These responsibilities defined the Public Prosecu
tor almost as a judicial expert rendering opinions on legal issues, :o As formulated in the 1988 Constitution, however, diffuse and collective rights
During the period of military rule (1964-85) the Public Pro are vague enough to encompass almost any collective or group interest.
secution was still not autonomous, but became dependent on the ., The class action suit is a juridic instrument which allows groups within civil
Executive rather than subordinated to the Judiciary. During the society, or the very Public Prosecution, to defend their collective rights, which
'• .. encompass rights established in civil law, According to the Class Action Suit Law
206 Maria Tereza Sadek and Rosiingeia Batista Cavalcanti The New Brazilian Public Prosecution 207
associations can also make use of this legal instrument, the Public From an institutional point of view, changes in the powers and
Prosecutor is the institution that has made most use ofit.6 role of the Public Prosecution constitute the most significant
The Constitution stipulated the Public Prosecution's unity and reform embodied in the Federal Constitution of 1988. No other
indivisibility in addition to guaranteeing it administrative and func institution underwent such a profound reform or expansion of re
tional autonomy. It granted the Public Prosecution the same sponsibilities. The Public Prosecution was no longer known for
prerogatives granted to the Judiciary: lifelong tenure for its mem its classic parquet, or oversight role, and even less for being an
bers, a guarantee prosecutors won't be transferred to other juris advocate or defender of the Executive's interests. Since 1988, the
dictions against their will, and a constitutional guarantee of due institution has been known primarily as a defender of society's
benefits. interests. Hugo Mazzilli (1993), among many others, came to the
All of these powers potentially enable the Public Prosecution to conclusion that the Constituent Assembly effectively transformed
defend the so-called 'Social Right',6 be it through controlling abuses the Public Prosecution into a 'fourth power' autonomous from the
of power within the public or private sphere, or through defending Legislature, the Executive, and the Judiciary.7 These types of an
social rights. In other words, these new institutional powers trans alysis have also provided fodder to arguments that oppose the
form the Public Prosecution into a mechanism that can potentially new-found role, and power, exerted by the Public Prosecution. Des
ensure what Guillermo O'Donnell calls 'horizontal accountability', pite such arguments, it is clear that today the Public Prosecution
-'the existence of state agencies who have the right, legal power; has become a primary actor in Brazilian politics and, according to
and disposition to use their attributes, which range from routine the Constitution, is entitled to defend the democratic regime, the
supervision to legal sanctions Md. impeachment, against acts or juridical order, and the interests of society.
omissions of other state agencies who can 'be' 'characterized as de.::. The Public Prosecution operates at both the federal and state
linquent' (O'Donnell 1998b: 40). levels. The Public Prosecution of the Union, which presides at the
federal level, is in turn subdivided between the Federal Public Pro
secution, Military Public Prosecution, Labor Public Prosecution,
(Lei da A~ao Civil Publica) the Public Prosecution was given the right not only ~i: and the Public Prosecution of the Federal District and its Territor
be a part of the suit, but also to act as a custos legis, which means it is responsible·
for an initial evaluation over its legal consistency. As a result, it has become active..
ies. Their respective jurisdictions are bound by the facets of law
in defending any collective or diffuse rights which might have been infringed. In: they are dedicated toward: Federal Justice, Federal Military Just
addition to this ample attribute, the Public Prosecution also possesses the respon-', ice, Labor Justice, and Justice of the Federal District and its Terri
sibility of conducting the preparatory investigative work necessary for a class'. tories. The Federal Public Prosecution, however, is the only
action suit. Such investigative work is referred to as a civil investigation, which ~ division that can litigate in the Superior Justice Tribunal (STJ),
analogous to a police investigation, but contains characteristics specific not only_4t
its administrative procedures, but also to its investigative work, which seeks uf
which levies verdicts on infra-constitutional law, and in the
•!•',..
collect sufficient evidence to justify the suit in question. ·,; Supreme Federal Tribunal (S'rF)-the highest law court in the
6 According to Macedo, 'the Public Prosecutor is today responsible for 96 pei.', country-which is the last resort for interpretation of const
cent of all environmental class action suits in the country's courts. Despite tn,~i.;.i itutional questions. The distinction between the various subdiv
lack of data in other issue areas, one can presume that such a de facto roonopoiy:} isions within the Public Prosecution refers only to differences of
(since the Public Prosecutor doesn't have a juridical monopoly over the filing ~Ji
such suits) also extends to civil actions in defense of collective and diffuse rigJ..it~r;
(Macedo 1996: 42). . ·,. ,:i.; 7
6 According to Macedo, 'Contemporary Law, a characteristic of the Welfai:~" Other notable legal scholars, such as Manoel Martins da Costa Cruz, draw
State and also known as Social Right, is structured according to a new type ·of similar conclusions. Cruz (in Mazzilli 1993) states, 'the Public Prosecution pre·
juridical rationality. The Social Right increasingly takes into account group in~~;; sents itself as a veritable State Power. If Montesquieu had written Spirit of Law
eats and social inequalities in order to seek, through distributive justice, a mor.i( today, the division of powers would be characterized not by three, but four div
just and egalitarian social order. Such a logic stands in contrast to a more liberl(lJ; isions.' While. this is not .the place to have a discussion over what constitutes a
juridical order. A juridical order based on the Social Right therefore provi~fiji separate 'power within the State, nor over the relevance of the traditional separ
protection to specific groups (and, therefore breaks with the liberal paradigm. <it; ntion of powers theory, we are merely drawing attention to the considerable conse
equality under the same juridical order) such as, for example, consumers.. tlie1 quences associated, · .wilh,·. ciHtF.ltitul:ional reforms enacted over the Public
elderly, minors, the physically deficient ... ' (Macedo 1996: 46). 'f, Prosecution, ·
208 Maria TerezaSadek and Rostingela Batista Caoalcanti The New Brazilian Public Prosecution 209
jurisdiction, not types of juridical instruments at its disposal, tionalization, all give the Public Prosecution a great deal of poten
guarantees, or operating principles. tial and capacity to promote citizen rights. The new legal param
eters and 'political will' embodied in the Constitution have, in fact,
translated into effective action. The Public Prosecution's perform
From Potential to Practice ance subsequent to the 1988 Constitution stands in stark contrast
to its previous record. Both the Public Prosecution of the Union
The noteworthy judicial reforms enacted by the 1988 Constitution and the Public Prosecution of the states have adopted their newly
and the set of laws deriving from them have created a propitious granted oversight and control responsibilities. Members of the in
environment to generate a more effective system of justice: Due to stitution have been utilizing legal instruments at their disposal to
the 1988 Constitution, Brazil's justice system has increased its intervene in the most diverse spheres within public administration
ability to oversee public and private institutions and protect indi and collective life. The Public Prosecution in Brazil consists of
vidual and .social rights. As we have already noted, however, the 10,000 members-of which 9,662 are public prosecutors (state
constitutional attributes delegated to the Public Prosecution pre level) and 338 are prosecutors of the Republic (federal level)-and
sume the existence of an institution with a high capacity for inter rarely a day passes in which the media does not report an action
vention, whether it be to control abuses of power in the public and taken by prosecutors to uphold citizen rights.
private spheres or to defend social rights. In order to help guaran At the federal level, the members of the Public Prosecution have
tee that such responsibilities are carried out, the Constitution fur transformed their institution into a veritable political force. Fed
ther provided for functional and administrative autonomy of the eral prosecutors have actively participated in political disputes
Public Prosecution. with the Executive, Legislature, or groups in civil society. Indeed,
The re-definition of the Public Prosecution's institutional profile prosecutors of the Republic have brought legal action against gov
and the expansion of its powers, has brought about a reconstruct ernment ministers for the misuse of public funds and/or property,"
ing of its identity. Legal texts, which include both the Constitution have questioned social and economic policies, and have exposed
and infra-constitutional legislation, provide the broad parameters corruption in the financial system. Not coincidentally, federal pros
and potential for such a reconstruction. The speed and sense of ecutors are viewed with suspicion by members of the Executive,
that reconstruction, however, fundamentally depend on its own, senators, federal deputies, directors of state-owned enterprises,
members. Individual members of the Public Prosecution are par and monetary authorities.
ticularly important, given the institution's structure, where each Opposition to the power of the Public Prosecution ranges from
member has the ability to act with a high degree of autonomy, specific criticisms to more broad misgivings pertaining to the insti
without having to strictly adhere to precedents or directives estab tution's structure and methods. The reaction of the government
lished by the institution's leadership. In other words, there is no and that of its sympathizers to a search conducted by the federal
strict organizational hierarchy. Public Prosecution and the federal police in the home of the ex
The potentially active role played by the Public Prosecution iii.· president of the Central Bank, Franciso Lopes, is an example of
enhanced by its organizational design and lack of institutionaliza such opposition. The warrant that allowed the Public Prosecution
tion. It is decentralized, lacking a functional hierarchy and there: to search the home of the ex-president of the Central Bank was
fore giving substantial autonomy to each prosecutor. Within any approved by the Judiciary and was intended to find evidence to
given Public Prosecution individual prosecutors vary considerably; support allegations that the Central Bank had unduly favoured
from those who take a very conservative approach to those who. two private banking institutions-Marka and FonteCindam. On
espouse the most liberal interpretation of the institution's respon=. that occasion in April 1999, despite the fact that the majority of
sibilities. This lack of institutionalization allows the Public Prose-.
cution's identity to be shaped by the individual characteristics ~fi " Since 1999, for instance, the Federal Public Prosecutor has brought
its members. · . ·. legal action against 12 ministe rs for. using the Brazilian Air Force's (Forfa Aeria
The combination of constitutional attributes and guarantees, or-: IJNu;ileira) planes. fol• privnt:o beuofit, However, even the Attorney General of the
ganizational structure, and paradoxically, its low level of institu- Republic facos ~~n1il1.1r· c..i)(~~·g<Jl:I, ·
210 Maria Tereza Sadek and Rosiingela Batista Cavalcanti · ' The New Brazilian Public Prosecution au
legal scholars considered the search legal, President Fernando Public Prosecution can only have access to private banking or tax
Henrique Cardoso termed the action an 'invasion' and a 'return to records through authorization from the Judiciary, which has given
arbitrary rule'. Others claimed the action reflected pure exhib its approval to only 60 per cent of such petitions (CONAMP, July
itionism, was based on unfounded accusations, and reflected a lack 2000).
of responsible authority. Despite the obstacles, according to a prosecutor of the Republic,
Criticisms such as these, in the majority of cases, are not re 'we are convinced we are extending the law to areas that were
stricted to critiques over possible excesses committed by a few previously untouched, punishing privileged groups, and increasing
individual prosecutors. Generally, they are accompanied by vehe the access to justice of excluded groups' (Folha de S. Paulo, 28
ment calls for a change in the role and jurisdiction of the .Public July 2000). Along similar lines, a federal prosecutor in Rio de
Prosecution as a whole. In fact, a Constitutional Amendment that Janeiro, Raquel Branquinho Nascimento, who had participated in
is part of a larger judicial reform bill is currently being examined the investigation of possible irregularities committed by the Cen
in the National Congress and contains a seri.es of proposals that tral Bank in the selling of dollars below market rate to the Marka
seek to diminish the power of the Public Prosecution, particularly and FonteCindam banks, stated in an interview, 'criticism eonsti
with regard to its capacity to question, accuse, or conduct the pros tutes a right of defense, but they come from a privileged segment
ecution of authorities, and therefore to affect public policy. of society-politically, economically; and socially-which is not used
Despite such criticisms, it is clear that as a result of the Public to receiving the same treatment as others. Our fight is for all to be
Prosecution's actions, Brazilian society is gaining a sense that treated in an equal manner within constitutional principles'
crimes committed by the 'powerful' will not be ignored. This is· <:Folha de S. Paulo, 26 April 1999).
clearly reflected in this editorial;·appearingin the newspaper Folha The institutional performance of the Public Prosecution of indi
de S. Paulo: vidual states varies greatly from state to state. As Ada Grinover
In the past, the flagrant disobedience of the law only occasionally brought observes, 'the Public Prosecution's model employed in the state of
consequences. Corruption and crime won't and will never end because of Sao Paulo or Parana, for example, is not a reality throughout the
action taken by prosecutors. Nevertheless, a certain sense of impunity: entire country.... Not all Public Prosecutions are as independent or
has been broken. The imprisonment of a banker and the stripping of a: autonomous as they could be' (Sadek 1997). Such variation across
rich and powerful senator's legislative mandate has acted as an Olympic: states is well illustrated by contrasting the cases of Paraiba and
benchmark. They serve to establish a new parameter of morality in which Bahia. In Parafba, for example, the Public Prosecution focuses on
society, little by little, begins to think it possible to attain. (Folha de S:: administrative improprieties, and has created a 'Permanent Com
Paulo, 9 July 2000) · mission in the Combat Against Administrative Impropriety' whose
Despite recent successes, members of the Federal Public Prosl. main function is to receive and evaluate accusations of administra
ecution face a number of obstacles. One problem is insufficient tive irregularities practiced by mayors or ex-mayors, brought forth
staff-as mentioned earlier, there are only 338 federal prosecutors; by any citizen, the Tribunal de Contas, 10 or by the County Legisla
and since 1996 the institution has been awaiting approval of legi~~ ture. In contrast, the Public Prosecution in the state of Bahia has
lation that would create 304 new positions. The Public Prosecution, demonstrated a clear preference to shy away from activities which
also suffers from a lack of technical support. Prosecutors must; have the potential to run against local political interests. In Bahia
depend on police investigations to obtain evidence and, accordmgi the Public Prosecution has focused its efforts on monitoring the
to 93 per cent of federal prosecutors, one of the primary obstacl~~:: police departments and responding -to grievances pertaining to
to attaining justice is a lack of police resources. According to 8,!f family disputes like alimony payments and paternity tests. While
per cent of federal prosecutors, police investigations yield insuffll the oversight of the police department can be considered 'political',
cient results (Castillio and Sadek 1998).9 In addition, the federal such an effort does not come close to the political repercussions
associated with judicial action which can eventually lead to the
9 These figures were obtained from a 1997 survey sent to all members of. $.~
. . ~ . . ' .
Federal Public Proseeution. More than half of the federal public prosecutors-.re1i · w The Tribun!ll._dd-9P.n,tok ,is, rospouaible for auditing public budgets and it is an
sponded (244 of 516). institution subordinatad to. UtQJ ...ogia(ul:~lrc. ·
66 Charles D. Kenney
relative immunity from executive control, the ombudsman's will Horizontal Accountability: Conceptsand Conflicts 67.
ingness and ability to exercise a degree of horizontal accountability reasons, we are left with a conceptual gap. 7 In these cases, ~ents of
was due both to the quality of its leadership and to the support the_state are empowered to p9.li_tj_(?al!J: s<:!nction other agents of the
both in terms of material resources and public backing-given by a state for actions or omissions that are not ggalified as unlawful..
· number of foreign governments, including the United States. The =---:---:-:-------.. ------·--~-··-- - --
Clearly, these are accountability relati?ns_hips and_ not 12ower- .
ombudsman's office became a vigorous and widely respected de ~haring checlcs and balances. Clearly too, tliey are hor,=--- . _ ~' •
fender of citizens, rights and a thorn in the side of the Fujimori tionships within the state, not vertical relationships between soci
administration. At the same time, Peru's ombudsman's office con ety and state. In other words, they meet all the requirements for
stituted but an island in a truncated network of horizontal account O'Donnell's definition of horizontal accountability, save one--they
ability, and-without underestimating the political significance of need not be responses to allegedly unlawful acts and omissions.
its actions-the accountability it exercised was limited. The om We are faced, then, with a choice: either we broaden .O'Donnell's
budsman could ·investigate and publicize alleged violations of citi definition of horizontal accountability to include some relation
zen's rights, but it was the attorney general's office and the ships in which one state agent holds another state agent account-
judiciary that were in the end responsible for further investigating abl~!£r acts th~t~~. ~~-~~~!!~ge<f1>_J?!_,~nla~ur,or-welioidonio
and acting on the allegations presented by the ombudsman. In OTionnell's unlawfulness requirement and are left with a set of
other words, the accountability exercised by the ombudsman's accountability relationships that fall outside both the categories
office was a diminished, limited form of horizontal accountability, of vertical and of horizontal accountability. Given the importance of
because the other powers on which accountability depended often lawfulness in O'DonneU's theoretical work on democracy and the
failed to act.
rule ofla,v, the stipulation that horizontal accountability be always
a response to improper and unlawful behavior cannot be set aside
Horizontal Accountability and Unlawfulness lightly. In parliamentary regimes, however, it makes sense to relax
this requirement and include parliaments holding ministers polit
A third disputed aspect of O'Donnell's definition is the scope of ically accountable within the concept of horizontal accountability:
horizontal accountability. ~ argues-th-af""liorizontal ac This move allows us to restate the unlawfulness requirement in
countability should be construed so as to 'hold rulers accountable terms of lawfulness: horizontal accountability involves state actors
_ furtfiepolit1cal··~~~?tlustle~_so~_q~e~s~f their behavior or agencies willing and able to sanction" other state actors and
in office'_ (1999: 60). Under what circumstances might state agen a enc1es-for t1ieir actsa.ilif"oiiuss1oiis'"In-accordance with the la.;
cies exercise horizontal accountability for actions that are not g - ----~------ ...... ---.----·----··--·-·-··----·
aiicttlie- constitution. For presidential
..-1,--regimes, this definition de-
-~
u_aj.!_~).,_~!,lt_9.p.J_y_~li~ Certainl~~gal _ anctions would..-.....
inal?.E!.P..Priate i~such cases. Although it is-
be_,
n difficult to disen
notes the same set of cases as O'Donnell's definition, since allega
tions of unlawfulness are usually required in such regimes for
tangle personal and political motives from arguments over the the application of sanctions. For parliamentary regimes, however,
lawfulness of acts and omissions, ]!!?,e im,Eosition of_legal sanctions. this definition would also denote cases in which state agents
\.vithou£ e~ence ofp.legality would its~e illegal. are held accountable for actions that are politically objectionable,
The question of politicaT sanctions is less clear. O'Donnell's but not illegal, as long as the authority to apply sanctions in
unlawfulness requirement works well within the framework of pre such
tution.cases is recognized as conlorming to the law and the consti
sidential regimes, in which the survival of office-holders in each
branch of government is politically independent of the other
branches, and impeachment requires allegations ofunlawful behav 1
ior. When we consider parliamentary regimes, however, or regimes This problem also arises where legislatures can remove presidents from office
such as Peru's in which ministers can be constitutionally censured for mentaJ, moral, or physical incapacity, without alleging unlawful actions. As
noted above, the Ecuadorian Legislature used such a measure to remove President
and removed from office by the legislature for purely political Bucaram from office i.n 1997, and similar provisions exist in a number of Latin
American constitutions.
,:
214 Maria Tereza Sadek and Roetingela Batista Caoalcanti The New Brazilian Public Prosecution 215
mayors have testified to date and 108 of those testifying have . • In the state of Para, there were four convictions, with the
been convicted (Veja, No. 27, 5 July 2000). ': Public Prosecution bringing charges against 125 mayors and
• In the state of Parafba, there were 18 convictions, and the ex-mayors, the majority for misuse of public funds (data from
Public Prosecution was able to induce mayors and ex-mayors the Public Prosecutor and Justice Tribunals published in Veja,
to return one million reais of illegally appropriated funds to No. 27, 5 July 2000).
state coffers rather than face charges (Veja, No. 27, 5 July • In the state of Parana, located in the southern region of the
2000). country and containing 399 counties, 141 mayors have had
• In Sao Paulo, the Public · Prosecution has levied 436 accus- criminal and civil charges levied against them due to irregu
ations against mayors since January 1997, the majority of which lar practices in their public administrations. In the criminal
were based on Law 201, which refers to crimes of adminis sphere, there are 250 suits against 101 mayors currently
trative responsibility, and can be enacted in cases of corrup in office. In the civil sphere, there are 110 suits against
tion, misuse of public funds, irregular investment of federal 40 mayors (Ministerio Publico do Parana, as of 30 June
and state funds, irregular use of municipal property, and so 2000). ·
on. The number of mayors who have been stripped of office in The Law of Administrative Impropriety (Law 8.429), passed in
the state has increased from seven in the previous adminis 1992, proved critical to the Public Prosecution. In the civil sphere,
tration (1993-96) to at least 41 in this last administra- ·. this law draws the greatest fear from mayors, because it enables
tion (1996-2000). Such a rate over the last three and a half public prosecutors to levy charges of illicit enrichment, financial
... years .is e.qajyalent to one mayor per month losing his or her · losses to the public sphere, and offences against principles of public
political office . due .to.. the Public Prosecution's work. Some of - . adroinist1;atiori. Iri the· criminal sphere, the most· frequent charges
these dispossessed mayors, however, were able to re-assume consist of appropriating public funds, u fraud, and hiring county
office through countervailing legal action. The primary employees without a public entrance exam.
charges brought against mayors were the misuse of public The state of Rio Grande do Sul deserves special attention. In
funds, irregularities in public employee hiring, inflating the .'. this state, the Public Prosecution has received the least amount of
costs of public works projects (which means the money could resistance from the Judiciary, a traditionally conservative institu
go to undisclosed third parties), and not transferring due rev tion that shies away from political conflicts. In 1994, the Justice
enue to the County Chamber. With the exception of mayors • Tribunal in Rio Grande do Sul established a Criminal Court for
from the cities of Bauni, with more than 300,000 inhabitants, the sole purpose of judging mayors.13 This Court has already con
and Guarulhos, with more than one million inhabitantsr victed 100 mayors, with punishments ranging from required com
available data indicates the majority of mayors who wer~i munity service to prison sentences for crimes like the misuse of
stripped of their mandate governed counties with less tha#,,~ public funds or irregular contracting. As of February 2000 there
20,000 inhabitants. Finally, public prosecutors have demons. were two mayors in prison in Rio Grande do Sul.
strated an exemplary performance in the city of Sao Paulor' Due to the Public Prosecution's efforts, the majority of mayors
the economic centre of the country. A group of public prosec;'!f and council members in the country face an unparalleled level of
tors created the GAECCO (Special Group to Combat Organ{ oversight. Whereas in the past mayors were accust.omed to govern
ized Crime), which has been responsible for the investigati,01;( ing without limits and to using their mandates either for personal
of a political scandal with national repercussions. The scandaf enrichment or as a springboard for higher offices at the state or
came to be known as mafia dos fiecais" of the city, and led t~
stripping the mandates of two county legislators and oµ,~
i:t Known as peculate in Brazilian legal terminology.
state deputy and to the temporary, but profoundly emblems ,a The fourth Camara (a court within the Justice Tribunal) was created by the
atic, removal of the mayor of Brazil's largest city. ·' " Judge Luiz Melibio Machado 'in 1994·, and became known as the 'mayors' court'. It
must be noted that· potent:inl punishments to infractions committed by mayors
wore delimited well bdf<lro ·t.h.o new Constitution through Law 20 of 1967. This
11 Or, 'Mafia of the Auditors'. law, however, had no real a(foctive impact.
)
..
216 Maria Tereza Sadek and Rostingela Batista Cavalcanti The New Brazilian Public Prosecution 217
federal level, today they find themselves in an uncomfortable wagered by the opposition, have all created substantial opportun
position at best." County chief executives unanimously recognize ities for the Public Prosecution. In the majority of cases the Public
that administering cities has become increasingly difficult each Prosecution has known how to take advantage of such opportun
passing yeaf/t In this statement, one of them encapsulated the gen ities, and they have done so through the aid of mass media and the
eral mayoral sentiment, while attempting to minimize the real opposition. Such positioning has affected not only the process by
motives behind the charges being levied against him, 'If the cur which
. the institution creates its own identity, but also its public
rent economic difficulties weren't enough, we still have to put up image,
with envy .... Since now one can't hunt birds and animals, they
have declared open hunting season on mayors.'16
It should also be noted that the Public Prosecution appears to Who Guards the Guardians?
have a greater ability to oversee the Executive than the Legisla
ture does. According to existing law, the National Congress is em While the Constitution undoubtedly delegates to the Public Pros
powered to create a Parliamentary Investigative Committee (CPI) ecution responsibility to oversee and control political and juridical
only after obtaining the approval of 171 of 513 federal deputies, institutions (the Executive, Legislature, and Judiciary), it does not
and 27 of 81 senators. Such a requirement effectively means that define who should oversee the Public Prosecution. Such ambiguity
in times of high government support in the Congress it becomes begs the traditional question: 'Who shall guard the guardians?' As
nearly impossible to create a CPL The opposition would have to a result, there is growing concern over possible 'excesses' in the
find support amongst legislators from parties who support the actions of members of the institution. Some political analysts, for
president. Recent experience indicates that not even broad-based example, have criticized the Public Prosecution for paralyzing public
popular support is sufficient to bring about a CPI in a Congress decision-making, and more seriously, they have questioned the
that supports the government's interest. According to Datafolha," very nature of such oversight capabilities (Kerche 1999). Drawing
in 1997, 91 per cent of the population was in favour of creating a upon certain ominous parallels from the not-so-distant past, ana
GPI to investigate charges that the government had 'bought' votes lysts have raised concerns about the ability of non-elected officials
in the Congress to obtain approval of a Constitutional Amendment to intervene in acts legitimated by the popular vote. Nevertheless,
to allow the re-election of all executive posts. Despite such popular the new structure of the Brazilian Public Prosecution-despite
support, the CPI was never created. In a similar vein, despite 86 representing a novelty from an institutional point of view-is con
per cent popular support to investigate a current political scandal · sistent with a larger process of creating a new type of democracy.
surrounding the construction of the building that would house SaQ ·. This 'new' democracy is one where law increasingly inserts itself
Paulo's Labor Tribunal (Forum Trabalhista de Sao Paulo, or T~T);.: into politics, and a growing number of oversight agencies exert
a CPI most likely will not be created. Very similar difficulties are · control over all branches of government. Political analysts are still
also found at the municipal and state levels. . groping with how to analyse this new dynamic and the internal
It can be argued that the confluence of allegations, popular dis- and/or external logic of these new institutions of control, which
satisfaction with politicians, difficulties encountered by the Legis-. do not seem to fit easily within the classic separation of powers
lature to effectively monitor the Executive, and severe criticisms· model.
These issues require discussion. First, the institution's particu
.1 arities within the justice system should be observed; second, the
14 The possibility tor oversight always existed, whether it be by the Legislature Public Prosecution's limitations within the justice system should
over the Executive or vice-versa, or through the Tribunal de Contes, an organ be examined; and finally, the existence of internal and external
created with the sole function of overseeing public spending. The Tribunal de controls over the Public Prosecution should be determined.
Contos is an auxiliary agency of the Legislature. Recruitment and promotion patterns within the Public Prosecu
15 The statement was made by a mayor from the interior region of the state or,:
Sao Paulo and quoted in O Estadode Siio Paulo, 4 January 2000. ·
tion are key elements with regard to the institution's particular
rs Datafolha is a public opinion polling department within the news organiza1:: ities within the justice system. Much like the Judiciary, entry into
tion Folha de Silo Paulo. the Public Prosecution is not conducted by election, but through
218 Maria TerezaSadek and Roeiingela Batista Caualcanti The New Brazilian Public Prosecution 219
civil service examinations. Potential candidates require only a tenure rewards career experience or the length of time one holds
law degree. The Constitution (Art. 129, Par. 3) states: 'Admission office, and is an objective criterion that allows ascension to individ
into the career shall take place by means of a civil service entrance uals who do not move within the organization's powerful or influ
examination of tests and presentation of academic and profes ential circles. In contrast to tenure, promotion based upon merit
sional credentials, ensuring participation by the Brazilian Bar As can lead to a mechanism of internal control. Members who occupy
sociation in such examination, and observing, for appointment, the the highest positions within the institution-the colegio dos
order of classification.' Tenure and merit are the two criteria used procuradores (the prosecutor's college)--can exert internal control
for promotion. over their fellow members because they are responsible for
Entry and promotion in the Public Prosecution are obviously not designing merit-based promotion evaluations and criteria.
controlled through a system consistent with a classic conception of Internal mechanisms of control are exerted by two agents: the
vertical accountability. A lack of internal vertical accountability, Corregedoria and the Attorney General, the institution's highest
however, is not necessarily positive or negative. There are strong office. The Corregedoria-Gerai of the Public Prosecution, a high
arguments in favour of entry through a civil service examination. ranking department within the institution, is broadly responsible
These include accessibility, and also a set of incentives which for personnel orientation and oversight. Its primary formal respon
discourage (but obviously do not eliminate) political influence and/ sibilities include conducting inspections of prosecutorial depart
or nepotism, and promotion of competence. This manner of selec ments, keeping track of progress toward goals established by public
tion has shown itself to be 'democratic' through being 'open'. Most -: prosecutors within their respective programmes, and implementa
recent data show there is a high and growing number of select can- tion of disciplinary action against members of the institution,
didates "who come"from 'adiverse pool-of public and-private univer- . .. .through use .of.the appropriate administrative S_a.il<!Ji<>P.~. (I,c!-W Ne>.
sities. Studies conducted by ID ESP (Sadek 1997; Castillio and 8.625, Art. 17, 12 February 1993). The highest office within the
Sadek 1998) indicate that recruitment has increasingly been exter- . Public Prosecution of the Union is held by the Attorney General of
nal or, in other words, there has been an increase in the proportion the Republic (Procurador-Geral da Reptiblica), and at the state
of entrants whose parents do not have an affiliation with the· level it is held by the Attorney General of Justice (Procurador
Public Prosecution (62 per cent of entrants to state Public Prosecu Geral de Justica). According to Article 128 of the 1988 ·constitu
tions and 78 per cent of entrants to the federal Public Prosecution, tion, the Attorney General of the Republic is 'appointed by the
do not have family ties to the justice system.) Furthermore, a. President of the Republic from among career members over 35
significant segment of those who are appointed to the Public Pros years of age, after his name has been approved by the absolute
ecution have parents with a medium or low level of education,. majority of the members of the Federal Senate, for a term of office
thus indicating that entry is not limited to the social, political, oi; of two years, reappointment being allowed. The removal of the
economic elite." · .·1i Attorney General of the Republic, on the initiative of the President
Promotion patterns, as mentioned earlier, are determined through of the Republic, shall be subject to prior authorization by the abso
tenure and merit. These criteria also reflect positively on th~: lute majority of the Federal Senate' (Federal Constitution of
Public Prosecution because they reduce political and party inter.\· Brazil, Art. 128, Par. 1, 2).18 The Attorney General of Justice is in
ference over the institution's leadership. Promotion based on' turn nominated by the state governor from a list of three candi
dates, whose names are determined by election from among all
members of the respective state Public Prosecutions. 19
17 A study conducted by IDESP in 1997 reveals that a majority of prosecutor~·::
from the Public Prosecutor of the Union come from middle class and lower middl~t
. "'
class families. Of all interviewees, only 21 per cent had a parent in a white 1,:01\ru,i.\ rn According tq a study conducted by IDESP, only 5 per cent of Public Prosecu
profession, while 57 per cent had parents in blue collar professions (Castilbo ;mdJ tor members agree with the selection rule for the Attorney General of the Republic.
Sadek 1998). With regard to members of the Public Prosecutor at the state levef,t: 'I'he overw},i.elming 111;ajority of members agree the selection should be conducted
data reveal a very similar picture: 67 per cent come from middle class and low~f? without political irite('.fiaroncc (Custilho and Sadek 1998).
middle class family backgrounds, and only 15 per cent of all prosecutors did not'. . t!t Differences in a~!.Q~tioi:i rules for the federal and state Attorney General pos
work prior to entering the institution. · · ition can act as inci.i.q·qf~-~.o.fJl~oio.~titution's level of real autonomy.
l
220 Maria Tereza Sadek and Rostingeia Batista Caoalcanti The New Brazilian Public Prosecution 221
As was mentioned earlier, the first paragraph of Article 127 of ceded to requests made by prosecutors. In Italy, however, prosecu
the 1988 Constitution stipulates the institutional principles of tors can issue temporary arrest of suspects, they have access to
the Public Prosecution to be unity, indivisibility, and functional classified banking and tax records, and they have the right to
independence. Unity refers to the fact that each prosecutor can install telephone wiretaps and confiscate documents associated with
have only one institutional affiliation, indivisibility means that their investigation. In Germany, prosecutors can confiscate goods
members are interchangeable, and functional autonomy means in and documents, intercept correspondence and telephone calls, in
dependence in the exercise of its functions (Mazzilli 1993). stall undercover agents in organized crime organizations, and issue
The large degree of freedom and autonomy given to members temporary prison sentences. Finally, in the United States, police
of the Public Prosecution, who are not obligated to submit to the officers work under the command of prosecutors. While American
guidelines established by the Attorney General, constitutes the prosecutors also need a judicial warrant in order to have access to
most important criticism of the institution. According to legal de classified information, their requests for warrants are granted in
terminations, each prosecutor has the right to select his or her 90 per cent of all petitions.
own cases according to his or her own criteria. Furthermore, very One can further argue, at least in the Brazilian case, that
few, if any, mechanisms of popular control over the Public Prosecu dangers posed to democracy from possible excesses committed by
tion exist. members of the Public Prosecution are fewer and of lesser degree
The Constitution granted a number of guarantees to members of than the dangers from a lack of transparency in public adminis
the Public Prosecution (such as life tenure and functional auton tration. In other words, the ransacking of public patrimony, im
omy) to assure the institution's independence-principallyagainst punity, and corruption of all sorts poses a much larger threat to
Executive intervention. Those same guarantees, however, can con democratic values. 21 As a public prosecutor from the state of
tribute to 'excessive' behavior in which the institution begins to Parana has stated, 'Corruption has always existed, but it hasn't
operate beyond the limits of its prescribed mission. But, much like increased. What changed is that now the Public Prosecution and
the Judiciary, without these constitutional guarantees the institu the population possess the means to control such irregularities'
tion would be open to pressure from the private and political arena; (Folha de S. Paulo, 9 July 2000).
making it very difficult to carry out its responsibilities. The media has contributed toward the dissemination of two con
Concern over potentially excessive behaviour on behalf of the tradictory images of the Public Prosecution. On the one hand, the
Public Prosecution, however, should be attenuated by the fact that public prosecutor is portrayed as the 'people's true defender', while
the institution cannot sentence verdicts. Members of the institu on the other hand the same public prosecutor is portrayed as an
tion have the power to initiate legal action, conduct investiga 'irresponsible exhibitionist'. The image of the public prosecutor as
tions, and open public inquiries, but ultimately the Judiciary, the 'people's true defender' is already commonplace in the main
decides whether the accused party is guilty or innocent. In this news media channels. He or she is a prosecutor who fights the abuse
manner, the Judiciary, while not being able to prevent eventual of economic power and corruption, defends minority rights, is not
abuses conducted by the Public Prosecution, certainly has the . intimidated by other authorities, and 'hunts' mayors, council
ability to correct eventual abuses. Furthermore, prosecutors are members, deputies, senators, and ministers. In contrast to this
not above the law and are bound by ordinary legal constraints. positive depiction, the media also portray a Public Prosecution
In comparison to other countries where the Public Prosecution is·
active, Brazilian prosecutors have relatively limited resources at
their disposal. As stated earlier, in order to conduct an investiga 21
According to Werneck Vianna (1999), 'there doesn't exist any record of an
tion Brazilian prosecutors not only need support from the police, equivalent to the ujudicialization" of politics [or, stated in other words, of the
but also authorization from a judge to obtain access to classified· affirmasio« of Lai», its procedures and institutions] in a nondemocratic context,
despite the somber predictions of some who interpret it as a threat to majority
information. According to CONAMP (National Confederation of rule and democratic sovereignty.' He further stipulates that judicial institutions
: ••• '1'. •'. , . • . • ' .
the Public Prosecution)," in 60 per cent of cases the Judiciary has . 'don't substitute. p,~1liti.cs [or its in.....tii.u,tionsl, but fill a void that, in mass societies
with intense socl~l.in'obHfu1tiou (Iike Brazil), could come to substitute democratic
• ' • 1 ··: ••••• .•: • :
20 ConfederactioNacional do Ministerio Publico is an association of public pros- essentials wit,h Ii Vt~,;Jt\nt (If popular sovereignty which escapes the institution of
""'... 4-"-=-- f'..r-~ 4>11 of.otoc "'tth.1> rn11nt.rv su Ifraae.' ·· ·
The. New Brazilian Public Prosecution 223
222 Maria Tereza Sadek and Rosiingela Batista Cooalcanti
that does not respect due process of law, makes bold statements Conclusion
based on weak· evidence, treats suspects as already guilty, and
does not hesitate to give interviews. The institutions encompassing the system of justice have under
One must, however, attenuate concerns that the Public Prosecu gone a process of profound transformation. A quick analysis of
tion is excessivelyinterfering, or at least has the potential to inter Brazilian politics in the past few years demonstrates the extent
fere, in public policy making. According to CONAMP, in a majority to which the Judiciary and the Public Prosecution have stepped
of cases the Judiciary has ruled in favour of public administrators. out of the shadows to become relevant political actors. Both insti
In most cases judges have ruled that lawsuits against public tutions have been present in practically all significant reforms en
adJirinistrators constitute a violation of the public mandate, and compassing Brazil's social, economic, and political spheres.
that only the executive branch has the right to determine adminis Reforms were implemented or attempted in varied arenas: eco
trative priorities. Statements made by members of the Public nomic reforms, consisting of privatizations, the breaking of state
prosee,"Ution reveal how the Judiciary has effectively acted to monopolies, and economic opening; social. reforms, consisting of
curb potential excesses by the institution. From the point of changes in the system of social security and wage policy; and
vjew of members of the Public Prosecution and their respective political reforms, consisting of permitting the re-election of execu
the Judiciary is obviously seen as an institution that tive posts. Magistrates and members of the Public Prosecution
5yrnpathizers,
has hindered the Public Prosecution's oversight responsibilities. came to occupy a position in the public spotlight that was previ
'fhrough the use of jurisprudence, dozens of class action suits against ously reserved only for politicians, whether from the government
e,
· ·· · ·· · · ·p1iblic·authorities have-been ..annulled These af~__ tjyil_ ~~~o.n~ _~1:11-ch at or the opposition, the Executive or the Legislature. Such a situ-
···-ifi a..··· ., dicialactivisn , a . ······-·-····· a· ·t d
IS c ass e as JU 1c1c:U ac vism an unprece en e
a ti on, whi.·ch···-·-···1···
t,etnpted to obligate the State to implement public policy that effectively_
delivered on rights already granted to citizens, like protection of the en according to traditional standards, challenges conventional
yironment, children and adolescents, and public assets. (Public Prosecu notions of majoritarian democracy, divisions of power, and the
tor for the State of Sao Paulo, XI National Association of the Public very identity of judicial institutions. Today, one would be hard
rrosecution, interview by R. Cavalcanti, 23-26 September 1996). pressed to speak of the Judiciary as a silent power, along the
Public authorities can be condemned to carry out their responsibilities, lines of Montesquieu, or of the Public Prosecution as an institu
whether it be through reparation of effective or potential wrongdoing. tion primarily responsible for penal law, as was foreseen in
guch a ruling can further be made without violatingdue administrative Roman law. The Judiciary has been forced to render its opinion
djscretion or budgetary predictability. (Public Prosecutor for the State of
over public policy and governmental decisions. The Public Pros
Rio Grande do Sul, interview, March 1999).
ecution, in turn, has activated the Judiciary on questions that
All of these issues clearly impact upon the classic division of, reach beyond the limits of criminal law to focus its efforts on the
powers model. As O'Donnell points out, a new type of democracy is' defense of diffuse and collective rights. Under this new scenario,
etnerging, which implements additional mechanisms of control it comes as no surprise that the institutional design of judicial
and oversight of public authorities: 'These mechanisms include the. institutions is often identified as an additional consociational
classic institutions of the Executive, Legislature. and Judiciary; element in Brazilian politics (see, for instance, Lamounier 1992).
but in modern polyarchies they also include various other agencies The Judiciary and the Public Prosecution are seen as providing
which act as "ombudsman" and which are responsible for oversee» additional evidence for claims that Brazil's political system has
jng and auditing public agencies' (O'Donnell 1998b: 42). He warns/ adopted an institutional design characterized by actors and/or
however, that only under rare exceptions can these agencies func institutions with veto power.
tion properly when acting in isolation. They may be able to mobile, The Public Prosecution may be the Brazilian institution that has
ize public opinion with their proceedings, but ultimately thei# changed most from a constitutional perspective, and as a result it
effectiveness rests upon decisions made in court. 'Horizontal a~,~ is still attempting to define itself internally and externally. As the
countl3.bility isn't a product of isolated agencies, but of a networ W, current Attorney General from the state of Sao Paulo bas stated,
of agencies which has at its head a justice system devoted to th.iii its ri.ew · atttib«i]~S,' '\bting to. the Public Prosecution a field not yet
,; ~: ,.; '
. . . . : ... .~ :.. ... ~.. ·
accountability' (O'Donnell 1998b: 43). .: . ,: ,•,,· · · .· '·
.,
224 Maria Tereza Sadek and Rosiingeio. Batista Caoalcanti The New Brazilian Public Prosecution 225
delimited, a field of work which we are still exploring; a historic An independent Public Prosecution has been inconvenient to certain
ally recent field' (Sadek 1997). political sectors, both within and outside government. The fearless and
In the light of these recent developments, analysts should in consistent work carried out by prosecutors, who have brought charges,
creasingly turn their attention toward the difficult task of how to lawsuits and judicial indictments, is without a doubt responsible for a new
reconcile institutional independence with both internal and exter sense of hope that we can end public impunity. (0 Estado de S. Paulo,
21 February 2000)
nal accountability (Cappelletti 1995). As we have stated, the Public
Prosecution already exhibits the necessary conditions to independ In addition to the Muzzle Law, there are three other amend
ently exert its functions. Such -independence, however, should not ments being discussed in the Congress that seek to restrict the
be exerted in an irresponsible fashion, without mechanisms that Public Prosecution's autonomy. All have generated intense debate
allow the institution to be held accountable. Independence with an within and outside the Congress. The first amendment stipulates
effective delivery ofjustice cannot exist without responsibility vis-a that candidates to the Attorney General of the Republic 'can in
vis the law and society at large. On the other hand, a call for effect clude any person who is not employed by the Public Prosecution of
ive action in addition to responsibility on behalf of its members the Union, but who is at least 35 years of age, has considerable
depends on the existence of clear and predictable laws. Brazil is juridical knowledge, and an unblemished reputation.' The second
therefore a considerable distance away from attaining the right bal amendment gives the president of the Republic direct authority to
ance between independence, efficiency, efficacy, and responsibility. determine prosecutor promotions. The third proposal seeks to limit
The current judicial reform proposal in the Brazilian Congress is the power of prosecutors who preside at the lower echelons of the
interpreted by some analysts as a step toward attaining this bal Public Prosecution's organizational hierarchy through implement
ance, while others see it as an obstacle to reconciling judicial effi ing a 'special forum'" for mayors. According to this proposal, only
ciency with accountability. An important item in the reform the head of the institution, the Attorney General, would have the
proposal 'prohibits members of the Public Prosecution, the Judi ability to initiate legal action against mayors. 24
ciary, and the police to unduly disseminate to the media, or other Leaving the merits of these reform proposals aside, these amend
third parties, information gathered from carrying out their re ments provide an obvious indicator that legal reforms implemen
spective responsibilities which violates confidentiality laws to the ted in the 1988 Constitution have changed Brazil's justice system
detriment of personal privacy and the public image of the individ performance. Simultaneously, however, there has been a (re)dis
ual in question.' covery of the rule of law by society (Telles 1997). A significant and
According to the reform proposal that intends to impose a growing segment of the population is discovering the justice
gag rule, the National Council of the Public Prosecution is em system to be an effective means to oblige government to carry out
powered to fire members of the institution who violate the above its responsibilities. Furthermore, recourse to the justice system
clause. Known as the 'Muzzle Law', this reform proposal is · has a multiplier effect, which extends benefits granted to the
supported by the federal Executive and practically all mayors, underprivileged population.
while it is opposed by members of the Public Prosecution and Work conducted by the Public Prosecution, specifically its over
the media." Those in favour of the reform justify it as a necessary sight and control responsibilities, has generated a significant re
means to protect individual rights. Opponents claim that if sponse from those groups that feel their interests have been
approved, the reform will effectively give impunity to public ad~ · endangered. The proposed 'Muzzle Law' is not only a first and
ministrators involved in charges of corruption, public contracting tentative reaction by opponents to the Public Prosecution, but an
fraud, misuse of public funds, and illicit personal enrichment. Ar.-· indicator that the 'formal-legal country' has had an impact on the
guments against such reform are well illustrated by this -editorial 'real country'. Clearly, we are referring only to the latent potential
in one of Brazil's most respected newspapers: 23
Known as the foro priuiligiado,
24
The Attorney General is also the only position within the Public Prosecution
22 Such a proposal is at the least unsettling because the SaIIl~ limitations are
over which the Executive has any influence in its nomination, AB stated earlier,
not extended to legislators who, through CPis, have frequently ignored the the Attorney General is nominated by the Executive from a list of three candi
principle of presumed innocence. · dates drawn up b.Y tho Public Prosecution.
74 Charles D. Kenney Horizontal Accountability: Concepts and Conflicts 75
accountability to win the battle for democratic legitimacy, it is and because those who raise the republican banner of horizontal
first necessary that they recognize the sources of their illegitimacy accountability have themselves often lost even the republican
in the eyes of many citizens, so that they might act to gain the bases for legitimacy,
legitimacy necessary to sustain themselves in a conflict with the
executive. One critical aspect (that calls us back to the conceptual
discussion earlier in this chapter) is the distinction between polit References
ically objectionable actions and those that are unlawful. To the
extent that agents of horizontal accountability attempt to check Dahl, R. A. 1989. Democracy and its Critics. New· Haven, CT: Yale Uni
lawful actions by presenting them as illicit or unconstitutional, versity Press.
they end up undermining the legitimacy that flows from the proper Elster, J. 1999. 'Accountability in Athenian Politics'. In A. Przeworski,
exercise of their duties. In presidential regimes, at least, the more S. Stokes and B. Manin, eds., Democracy, Accountability, and Represen
the exercise of horizontal accountability appears to be nothing but tation, pp. 253-78. Cambridge: Cambridge University Press.
The Federalist. 1990. The Federalist: A Collection of Essays Written in
a reflection of personal or partisan interests, the less legitimate it
Favor of the New Constitution As Agreed Upon By the Federal Conven
will appear. tion, September 17, 1787, by Alexander Hamilton, John Jay, James
Finally, democratic legitimacy accrues to caudillos like Fujimori Madison. George W. Carey and James McClellan, eds. Dubuque, IA:
and Chavez to the extent that they are able to avoid the appear Kendall/Hunt Publishing Company.
ance of being unambiguously repressive of liberal rights to prop Kenney, C. D. 1998. 'Institutionalized Instability? Questions for Democ
erty, free expression, and assembly. Never paragons of liberal rights, racy from Peru, 1980-1992'. Ph.D. dissertation, University of Notre
they nonetheless must avoid the wholesale repression common to Daine.
most dictatorships if they are not to forfeit an important piece of Lakoff, S. 1996. Democracy: History, Theory, Practice. Boulder, CO: West
this valuable democratic legitimacy. If these freedoms do become view Press.
significantly restricted and if this becomes widely known, this O'Donnell, G. 1999a. 'Delegative Democracy'. In Counterpoints: Selected
Essays on Authoritarianism and Democratization, pp. 159-74. Notre
source of democratic legitimacy will be eroded and eventually
Dame: University of Notre Dame Press.
lost. --. 1999b. 'Horizontal Accountability in New Democracies'. In A. Sche
dler, L. Diamond and M. F. Plattner, eds., The Self-Restraining State:
Conclusions Power and Accountability in New Democracies, pp. 29-51. Boulder, CO:
I have made two points in these reflections on horizontal account Lynne Rienner Publishers.
--. 1999c. 'A Response to my Commentators'. In A Schedler, L. Dia
ability in Latin America. I :first argued for the appropriateness of a
mond and M. F. Plattner, eds., The Self-Restraining State: Power and
definition of horizontal accounta61lity.In whiclitneagents of ac Accountability in New Democracies, pp. 68-71. Boulder, CO: Lynne
c~untability-are-fi:mited-to those within. tlie "state -(and not aclQri_ Rienner Publishers.
Pastor, R. A. 1999. 'The Third Dimension of Accountability: The Inter
ffi!s~de 1~~1i6~~~~~cTI~~?c!:d-~fjii~€~~-J6~~),!liJ1~!1~l::;.
__ pE_! .._ - - -·--·-----··-······----·---··········· - , .. national Community in National Elections'. In A. Schedler, L. Diamond
the
-·--·--li
scope ._of -- accountability
_ - is generally
" ,,. limited
,. ,, -. -.,to actions
- ._ . --·--· o~s-
or --·- - and M. F. Plattner, eds., The Self-Restraining State: Power and Ac
sioi:J.s _!_~-~-.13-:E.~ ~~-~~b._!?~! ~~~-~~-..~· ·~-~P.~ ~Y.~~~~---1:!!aY_!~.clude countability in New Democracies, pp. 123-42. Boulder, CO: Lynne
actions that are not unlawful. Rienner Publishers.
I then explored the ways in which an attention to the demo Plattner, M. 1999. 'Traditions of Accountability'. In A. Schedler, L. Dia
cratic, republican, and liberal dimensions of modern representa mond and M. F. Plattner, eds., The Self-Restraining State: Power and
tive democracy helps us understand the initial success of leaders Accountability in Neu: Democracies, pp. 63-7. Boulder, CO: Lynne
Rienner Publishers.
like Fujimori in Peru and Chavez in Venezuela. These contempor
Przeworski, A., S. Stokes, and B. Manin, eds. 1999. Democracy, Account
ary caudillos often win the battle for democratic legitimacy at crit ability, and Representation. Cambridge: Cambridge University Press.
ical moments because their evidently massive public support Sartori, G. 1987. The Theory of' Democracy Revisited. Chatham: Chatham
constitutes a democratic trump of their more republican opposition House Publishere.
Horizontal Accountability and the Rule of Law 229
.............. ............... systems throughout, much of Latin America' (Diamond 1999: 41).
Countries may also consolidate 'electoral democracy', but fail to
8 overcome 'reserved domains of power for the military or other
............................. actors not accountable to the electorate', or without establishing
the rule of law to assure individual freedom and civic pluralism
(Diamond 1999: 10). Such awareness explains what Thomas Car
:f]:orizontal Accountability and the Rule of others has called the 'rule of law revival' (1998: 101) and its enthu
Law in Central America siastic embrace by donor countries and international financial
institutions. It would also help account for why Central American
countries presently seem to be unusually dependent on external
scrutiny of their justice systems and relatively open to reform
Michael Dodson efforts, which often are led from abroad. As Carothers has argued
in his most recent book (1999: 164), 'the rule of law appeals as a
Donald W. Jackson remedy for every major political, economic, and social challenge
facingtransitional countries.'
Introduction Utilizing his influential distinction between 'vertical' and 'hori
zontal' accountability, Guillermo O'Donnell (1994) has pointed out
1fb.is chapter is concerned with the efficacy of legal reforms in that the 'electoral dimension of vertical accountability' is relatively
creating effective mechanisms of horizontal accountability and strong these days in many Latin American countries. However,
also with how such reforms are perceived by citizens in El Salva courts or human rights ombudsmen, which function to ensure that
dor and Guatemala, the two countries on which we focus. In as citizens' rights are not violated by persons holding public author
sessing the efficacy of reforms, it is important not only that they ity, and to enforce the legal and moral obligations of those who
are effective in practice; reforms also must be seen to be effective exercise power, generally are weaker. O'Donnell has suggested
50 that ordinary citizens will come to trust reformed institutions that in order to establish a constitutional state in which the rule of
as venues in which they can seek redress of their grievances. law prevails, the three distinct elements of 'polyarchy' must be
Io our research we ask this question: In Central America's cur combined (O'Donnell 1999a: 29-51). These elements include free
rent democratic transitions are reforms producing judiciaries and and fair elections, the protection of individual rights (liberalism),
ht1man rights ombudsmen that are strong enough and independ and insistence on the moral and legal obligations of those who govern
eot enough to sustain horizontal accountability of the executive (republicanism). Recently, O'Donnell has put the point this way. A
811d legislative branches? Are judiciaries and ombudsmen helping political democracy must include a legal system that 'enacts and
to eliminate or at least constrain the impunity previously enjoyed backs' the rights and freedoms that we ordinarily consider to be
bY those who misused public autb.ority? Our inquiry is part of the part of a democratic regime, and such a system also must prevent
even broader question being posed in the literature: Will these· anyone from being above the law (O'Donnell 2000a. 46). These last
transitions end in the consolidation of effective liberal democracies two elements are the primary focus of this chapter.
thl'oughout Latin America? Research on this question provoked a At the same time, O'Donnell recognizes that those who exercise
51;in1ulating
process of theoretical reflection, a surge in case stud-. power must have the capacity to govern effectively. The fact that
jeS, and a growing backlash of skepticism (Ames 1999; Schedler et James Madison was willing to tolerate a certain amount of ineffi
al. 1999; Becker 1999; Mendez et al. 1999; Hammargren 2000). ciency in government in order to protect individual rights does not
Scholars have recognized the many defects that plague third-wave provide much guidance today in finding the right balance, for
democracies and threaten to hamper the 'quality of democracy' in example, between the oft-competing values of 'crime control' and
[llBllY countries (Schedler et al. 1999: 2). 'due process'. As this chapter will show, striking the right balance
j\roongthe most prominent of these defects of third-wave democ in countries like: El Salvador and Guatemala can be agonizingly
racies has been the 'grossly dysfunctional character of judicial difficult. With-· the growth of the modern state, we have seen the