Corruption and Democratic Governability in Latin America:

Issues of Types, Arenas, Perceptions, and Linkages1

John Bailey

Prepared for delivery at the 2006 Meeting of the Latin American Studies Association, San
Juan, Puerto Rico, March 15-18, 2006.

Abstract: In this paper I explore the effects of corruption on democratic governability in

Latin America, emphasizing different types of corrupt exchanges that affect various
arenas of politics and policy-making. My hypothesis is that some types of corruption are
more important than others for democratic governability. Linkages between democracy
and corruption need to be considered in both analytical and public opinion perspectives.
At the level of public opinion we find that the terms are multi-dimensional and conflated
in ordinary usage. Rather than specific findings, my goal is to sketch a map of the
corruption-democracy landscape that marks the sites that merit priority exploration.
Among these priority sites I include political parties, law enforcement and judiciary, and
grand corruption.

Introduction

Corruption regularly appears among the top two or three problems cited in opinion polls

in most Latin American countries. Along with violence, crime, inequality and

institutional weaknesses, it figures prominently in scholarly work on quality of

democracy as well. Researchers have made considerable progress in measuring aspects of

corruption and in linking these to reduced support for incumbents and policies. But to this

point, much of the work applies rather blunt categories and focuses on specific pieces of

the broader problem complex, usually with little attention to how the pieces might fit

together. Is there a useful way to reframe the corruption-democracy problem by

differentiating among types of corruption and the various ways in which these affect

democracy considered as regime as well as state institutions and processes?

1
 In this paper I explore the effects of corruption on democratic governability in

Latin America, emphasizing the types of corrupt exchanges that affect different arenas of

politics and policy-making. Causes of corruption, effects on economic performance, and

possible anti-corruption strategies and remedies lie beyond the present scope. As I use the

term, democratic governability encompasses not only how power is achieved and the

rules of the game (democracy as regime) but also the exercise of power by state agencies

acting within a legal framework to address priority problems in a society (governability).2

To anticipate theoretical-empirical challenges, we need to assess which kinds of

vulnerabilities are more important to the functioning of regime and state. Also, we shall

find that mass publics attach diverse meanings to corruption and also tend to conflate the

procedures and performance dimensions in a broad notion of democracy, complicating

our efforts to make democracy-corruption linkages.

I begin with a discussion of definitional and analytical issues and proceed to

examine the direct and perceived effects of corruption in relation to various arenas. My

hypothesis is that some types of corruption are more important than others for democratic

governability. Rather than specific findings, my goal is to sketch a map of the corruption-

democracy landscape that marks the sites that merit priority exploration.

Corruption: Meanings, Types and Arenas

Corruption is a highly contested term, loaded with normative baggage. The concept

doesn’t travel well across different cultures, through time, or across different individuals

and social categories. Following Michael Johnston (2002), I find it useful to combine

public-office and public-opinion approaches to defining corruption.3 Joseph Nye’s oft-

cited public-office centered definition is a useful beginning point:

2
 Corruption is behavior which deviates from the formal duties of a public role

because of private-regarding (personal, close family, private clique) pecuniary or

status gains; or violates rules against the exercise of certain types of private-

regarding influence. This includes such behavior as bribery (use of a reward to

pervert the judgment of a person in a position of trust); nepotism (bestowal of

patronage by reason of ascriptive relationship rather than merit); and

misappropriation (illegal appropriation of public resources for private-regarding

uses) (Nye 2002, 284).

Nye’s approach emphasizes the public-private distinction that characterizes the

modern, Western nation-state, which is the point of reference of most Latin American

experience. The main benefit of this approach is its relatively narrow scope: we focus on

government, public policy-making, and public-private exchanges. A limitation is that

formal-legal duties of a public office vary considerably across political units and over

time. In the neo-liberal era of state restructuring we find varieties of public functions

(welfare, education, law enforcement come to mind) delegated to private firms. Also, we

will probably find gaps between formal-legal rules about corruption, the informal rules

that hold among political activists and elites, and the meanings that mass publics attach to

corruption. To address these gaps, Michael Johnston suggests that we supplement the

analytical definition by including opinion:

A better approach is to use a formal analytical definition [in this case Nye’s] to

identify corruption, to posit social conceptions of corruption as an important

factor affecting political response (or non-response) to corruption, and then to

examine the divergence between the two outlooks. This strategy, rather than

3
 attempting somehow to merge two fundamentally contrasting conceptions of

corruption preserves the differences between them, and focuses our attention upon

those contrasts as interesting political question in their own right (Johnston 2002,

189-190).

Johnston’s idea is helpful because we can return to Nye’s definition and interpret

his notion of “rules” broadly to include both the formal-legal and informal customs. That

is, public actors may engage in behavior that is technically corrupt, but which they may

view as ethically acceptable. The contrasts between social conceptions and laws might go

the other way: the public may see as corrupt the behavior of public officials that is

technically lawful. Despite the utility of narrow definitions, if public opinion views some

types of misbehavior in the private sector as “corrupt” this becomes a relevant datum.

Further, perceptions of what constitutes trivial versus serious corruption vary over time

and across groups and strata defined in different ways.4 These perceptions, in turn,

influence attitudes related to regime legitimacy and individual behavior (e.g., voting,

obeying the law, and the like). In sum, it’s the interplay between the analytical and

opinion perspectives that opens up useful lines of inquiry.

Thus, as a research strategy we travel two paths. First, in an analytical sense we

explore the ways in which corruption negatively affects regime and state through a

variety of linkages. Second, with a public opinion approach, we investigate the ways in

which perceptions of corrupt exchanges can undermine regime and state.

In terms of types of corruption, we might think of abuse of public office in two

basic ways: bribery in various forms and at various levels (money, property, or some

“good” is given or extorted in exchange for a publicly controlled good), and political

4
transactions (duties are violated to exchange a public good for a resource that benefits the

public official’s power or influence).5 The two ends may be served in the same

transaction, or there might be separate acts and different motives.

We need to make two more sets of distinctions: where does corruption take place

(in which “arena” of politics)? And, is the transaction voluntary or coerced? Much of the

work on corruption, especially that of international financial institutions and donor

organizations, focuses on the “output” side of politics, the administrative-regulatory-

judicial arena.6 Characteristic of this arena, the public duties of officials are set out in

laws, decrees and codes with one or another degree of clarity and precision. The

discretion granted to the public official ranges from virtually none (the toll for a two-axel

vehicle is 50 centavos, exactly) to extensive (building permits should take into account

the structure’s aesthetic impact on the neighborhood). In all, the output side of politics is

dense, complicated, and important; but it’s only one of several arenas.

On the “input” side and in the “conversion” process (the “black box”) we need to

consider two key processes that are typically subject to corrupt exchanges. First, bribery

and political corruption may influence the core processes of democracy taken in the

procedural sense. That is, corruption may influence access to the ballot, party

competition, election processes, mass participation and the menu of legal rights that these

require.7 Second, with respect to conversion, both bribery and political corruption may

influence rule-making aspects of public policy-making. Policy-makers may decide an

issue in light of its impact on their personal finances or on their power resources (or

both), rather than in light of legitimate constituency interests or the “public good” as

decided through a deliberative process. Issues in this arena may be less relevant to grass-

5
roots level corruption and more relevant to particular interests, especially wealthier

individuals and groups (including organized crime).8 Perceptions of corruption in this

arena, however, may have important effects on attitudes at the grassroots level.

Finally, we need to examine the corrupt exchange itself.9 Two parties engage in a

transaction in which something of value is offered or requested for something under

public control. The various possibilities are summarized in Figure 1, which can be

illustrated by the routine traffic stop (although the basic logic holds across arenas and

levels of office). The best of worlds is the “N,N” pair: nothing offered and nothing

requested. Here the officer simply writes up the speeder. The worst of worlds is the

“Y,Y” combination: the client offers and the agent requests. Officer and speeder proceed

directly to negotiate a bribe. The “N,Y” pair suggests coercion; the agent requests the

bribe. The officer sets the terms and the speeder must pay, or else. The “Y,N” pair

suggests a client corrupting the public agent. The speeder cajoles the reluctant officer into

a negotiation. Traffic enforcement is a simple (and non-trivial) example, but the basic

logic holds with public works contracts, campaign contributions, or judicial decisions.

Figure 1 about here

The analysis can be carried into a second round of decision-making in the mixed

cases. That is, the speeder accepts/refuses to pay; or the officer refuses/accepts the offer.

The differences are important in relation to both research methods and policy remedies.

We cannot simply assume that either government or civil society is victim or predator;

we need to observe. As to policy, anti-corruption measures can be fashioned to educate

an uncivil society or to temper a predatory government. For present purposes, the basic

6
point is that the corrupt exchange is a complex transaction in which government and civil

society can be both victims and corruptors.

With the groundwork laid, we can proceed to address the relationships between

corruption and (1) democracy as regime, (2) policy-making processes, and (3) policy-

implementation by state agencies. The orienting question is: What are the direct effects of

corruption and of the perceptions of corruption on democratic governability? In mapping

this, my approach is to provide examples of corruption in the various arenas and

speculate about their direct effects on regime and state. With the arenas sketched, we can

consider the effects of corruption as filtered through perceptions. An assumption is that

the majority of the Latin American cases meet minimum standards of functioning

polyarchy, operating in generally weak but coherent states.10 As to limits of the analysis,

the issues of the pervasiveness of corrupt practices in a polity (contained versus systemic)

and the long-term effects of inertia deserve extensive treatment but can only be noted in

passing.

Corruption and Regime: Democratic Procedures

The mainstream, procedural approach views democracy in terms of: (1) competing

parties, (2) periodic elections, and (3) extensive participation. To these are added a series

of requisites. From an analytical perspective, it is evident that the forms of corruption that

distort these fundamental processes are the most serious.11 “Corruption . . . breaks the

link between collective decision making and people’s powers to influence collective

decisions through speaking and voting, the very link that defines democracy” (Warren

2004, p. 328).

7
 Beyond anecdotes, we don’t know much about the prevalence of corrupt practices

in gaining access to the ballot, influencing votes, or distorting information as provided by

mass media.12 Presumably, closed list proportional representation systems, with

nominations controlled by party leaders, would tend to centralize corruption. Primaries,

especially open primaries, would act to decentralize corruption.13 Open list PR general

elections would seem to decentralize corruption as several candidates compete to attract

support from party followers. Given widespread poverty, income inequality, and

increased funding flowing into more competitive elections, be these primary or general

elections, one would expect that forms of bribery are increasing as well. Scattered

examples of vote-buying in the region suggest a mixed picture:

In Brazil’s municipal elections in March 2001, for example, 7 percent of voters

were offered money for their votes. Different surveys in Mexico place the

frequency of vote buying at between 5 percent and 26 percent, while a 1999

Gallup survey in Argentina found that 24 percent of interviewees knew someone

who sold his or her vote (Pfeiffer 2004, 76-77).

The overall costs of campaigning are generally rising throughout the region. Our

knowledge about campaign finance is improving rapidly, for example with respect to

methods of calculating abuses of campaign finance rules. But the sources of the abuses,

that is, who are breaking the rules and what is received in exchange, remain hazy (Speck

2004). Also, we know a bit about incumbents’ abuse of office to reward or threaten

voters. With respect to distorting information, we know less about favorable deals that

media outlets offer parties in exchange for special concessions. In sum, from an

8
analytical perspective, corrupt practices that infect basic procedures of democracy as

regime merit priority attention.

Corruption and Policy-making

In a short paper I can only illustrate the vast arena of rule making with two typical

examples. First, in divided government where the president’s party lacks a majority in the

assembly, the president goes beyond traditional pork barrel or favors and resorts to direct

cash payments to pass legislation. Second, legislators simply abuse their duties to pocket

money through multiple schemes.

Majority-building in divided government:

Mario Pontaquarto, a former senate clerk, reportedly admitted to having been

dispatched in April 2000 by the government of Fernando de la Rua (December

1999-December 2001) to bribe a number of senators into voting for a law that

effectively reduced workers’ employment rights and which passed in May 2000.

The total amount of the bribe was apparently 5 million pesos (the equivalent of

US $5 million . . . ), which was allegedly siphoned from the state intelligence

service’s secret funds (TI 2005, pp. 97-98).

Legislative theft:

In 1993 in a bombshell interview with Veja magazine, a congressional staffer

alleged that congressmen were systematically defrauding the Treasury by writing

budget amendments that would benefit specific senators, congressmen, and

businessmen. The budget scheme was allegedly run by a number of congressmen

whose short stature gave rise to the “dwarves” nickname used in the press, who

9
 allegedly defrauded the Treasury in two ways. First, congressmen wrote budget

amendments that benefited specific construction companies, in exchange for

bribes and illicit campaign funds. Second, the congressmen wrote budget

amendments that transferred funds to “phantom” social organizations created by

friends and family members, and appropriated the transfers (Taylor and Buranelli,

forthcoming).

With respect to divided government with weak party discipline one could make a

functionalist argument that vote-buying may be necessary to rule making. Even more, the

direct purchase of legislators’ votes may be less expensive overall than the quantities of

pork cum kickbacks needed to achieve legislative majorities. Elements of a

counterargument, however, are more persuasive. First, vote buying obviates the need to

employ transparent arguments and engage in negotiations to design good quality

legislation. Second, as the Argentine example shows, the use of money from the

intelligence service’s budget inevitably brings the state security apparatus (with its own

methods and agendas) into the legislative arena.

Corruption and Policy Implementation

I sketch two overlapping arenas here. The first is a broad category of administration that

runs the gamut from ordinary service delivery through the top levels of grand corruption.

The second involves both rule implementation and rule adjudication: police and

administration of justice.

Administrative corruption

This involves abuse of office in the quotidian operations of government in

extracting resources, distributing goods and services and regulating behavior. It can be

10
quite significant, depending on the incidence and locus of the abuse. Most services the

state provides have some rationale, but some might be considered more important than

others. In terms of basic state functions and societal survival, one would assign priority to

taxation and budget administration, national defense and public security, and to the

regulatory agencies that oversee public health and safety.

In terms of incidence, we might differentiate between overall levels of corruption.

Corruption might be “contained” in specific offices or agencies, or it might be systemic,

present throughout the administration from the ministers at the top to bottom-level

support staff. There seem to be dynamics of vicious cycles and tipping points that can

trap a system in high levels of governmental and societal corruption. Manzetti

differentiates between systems of low corruption, with functioning checks and balances,

internal constraints and an exigent civil society, versus those with high corruption, which

is the case of most Latin American countries:

High corruption takes place when: (1) many checks and balances among the three

branches of government and the institutional mechanisms to combat corruption

are weak or not used; (2) there are not self-restraints in profiting from corruption

as commissions reach extremely high levels; and (3) corruption is so widespread

at any societal level as to be accepted and tolerated. . . . . (M)any LDCs are likely

to fall into the second model (Manzetti 2000, p. 139).

One might add that there seem to be varieties of mechanisms at work that appear to

reproduce high corruption, trapping governments and societies in a kind of low-level

equilibrium of poor quality administration. Two examples serve to illustrate ordinary

administration:

11
Municipal government in La Paz, Bolivia:

The city government was in effect a huge “construction company” that wasn’t

constructing much. The city owned tractors, trucks, and all kinds of construction

machinery. There were 4,000 city laborers, who were paid meager, fixed salaries

and were only coming to work an average of five hours a day. Machinery was

also used for a similar amount of time, rendering it extremely inefficient given its

high capital cost. But I found that the use of gasoline, oil and spare parts was

abnormally high. Surely they were being sold in the black market, I thought, and

soon this suspicion was sadly verified. New tires and expensive machinery parts

such a fuel injectors, pumps, and Caterpillar parts were available for sale; and in

exchange broken and used parts were “replaced” on the city’s machinery (Mayor

MacLean-Albora quoted in Klitgaard 2000, p. 7).

Milk distribution in Peru:

[Leakages] in Peru are significant and far more pervasive and extensive at the

bottom of the [distribution] chain than at the top. From the entire amount of

public funds intended for the Vaso de Leche program, barely 29 percent get to

their intended beneficiaries. This does not mean that 71 cents from each dollar are

fully lost in corruption costs. The rest of the resources get rather leaked away

through a combination of administrative costs, ineligible products and

beneficiaries, and other modalities. Results also challenge the predominant view

of the last decade that organizations that are closer to the people they serve are

inherently better in service delivery (Seligson et al. 2006, p. 27).

12
Woven into these examples are points that merit mention. Corruption is embedded in

administrative systems marked by low pay and poor performance; also, decentralization

doesn’t offer an automatic cure for corrupt practices.

Political corruption in program administration

Apart from issues of theft and bribery, governments and political parties

frequently use public programs as resources to generate support. Widespread poverty and

weak institutions in the region create special problems of clientelism in the delivery of

public goods and services that are supposed to be universally available and distributed

according to neutral criteria. Clientelism is a form of asymmetric relationship in which a

patron provides goods or benefits in exchange for the client’s support.14 The patron can

take diverse forms: an individual, group, corporation, or—rather commonly—a political

party. Argentina’s Partido Justicialista (PJ) provides an interesting example of the

reinforcement of clientelism in a context of neoliberal adjustment.

A 1997 survey of 112 PJ base units in the capital and Greater Buenos Aires found

that 96 percent engaged in some form of social assistance, including food

distribution, medical and legal services, child care, and programs for the elderly.

Base units also implemented government social policies. In Buenos Aires

province, for example, PJ activists participated actively in the Life Plan, which

distributed a daily ration of eggs, milk, and other basic goods to nearly 400,000

people through a network of 10,000 block workers, or manzaneras (Levitsky

2003, p.28).

Clientelism typically hinders efforts to professionalize the public bureaucracy. It also

undermines notions of citizenship, a status with legally established rights and duties. The

13
hypothesis is that citizens are likely to be more active, exigent participants in public life,

while clients will tend to remain more passive.

Grand corruption

In general usage this refers to high-level officials pocketing money from

kickbacks, theft, embezzlement, insider deals, and the like. Privatizations of public

enterprises provide some of the most spectacular cases of grand corruption in the region.

They often tend to be massive and convoluted, highly visible and yet opaque.15 The more

ordinary, recurring types of grand corruption provide simpler examples.

Variations on grand theft:

One of the biggest cases was the alleged embezzlement of 4.5 million quetzals

(US $600,000) of public funds by [Guatemalan] President Portillo, Vice-President

Juan Francisco Reyes and his private secretary, Julio Girón. They were accused of

setting up 13 bank accounts and four ghost companies in Panama to launder the

money (TI 2004, p. 197).

Bolivian Defense Minister in 2000, Fernando Kieffer, allegedly “diverted

international donations for the people of Aiguila and Mizqu who suffered an

earthquake in 1998; that some of the money was used to buy an overpriced

executive jet; and that he was behind the irregular purchase of Galil weapons for

the army” (TI 2005, p. 108).

[Nicaraguan President] Alemán’s alleged crimes were not unique to the region,

either in terms of the amounts involved or the methods used. We are talking of

allegations of some US $100 million in public funds obtained through fraud,

14
 embezzlement and misappropriation, though congressman Leonel Teller estimates

that the figure is closer to US $250 million (TI 2004, 222).

As the Bolivian example illustrates, grand corruption often involves transnational

actors, usually multi-national corporations. This form of corruption could, at some point,

undermine an entire government’s finances. Grand corruption frequently triggers

scandals, and the more typical effect is the devastating example it sets for all government

employees and the citizenry at large. If top officials are shown to be corrupt (and, even

worse, to suffer no consequences) the lesson to all others is a negative one.16 The inertia

of the vicious cycle of corruption is thus reinforced.

Corruption in law enforcement and justice administration

Law enforcement straddles rule implementation and rule adjudication, while

administration of justice falls into the latter category. As armed agents authorized to use

lethal force and operating generally out of public view, police officers are the law in a

real sense. Police engage in many different types of activities, and I would return to the

discussion of types of corruption to reiterate a basic point. Transit police may prey upon

the innocent, but many motorists break the law. The exchange is usually one of bribe-

seeking and bribe-offering. Other types of police, whether preventive (uniformed) or

investigative (plain clothes), are notorious throughout the region for engaging in violent,

abusive activities with virtually complete impunity, and for constituting key elements of

organized crime. That is, active or retired (or purged) police officers are routinely found

to be involved in a wide variety of violent racketeering, including extortion, kidnapping,

drug-trafficking, and “social cleansing.”17 In some activities, the police are simply

15
uniformed criminals. In other activities, the police extort payments from victims. The

exchange here is bribe-extorting and intimidated bribe-offering.

The judicial system plays fundamental roles in democratic governability. It can

provide for “horizontal accountability,” i.e., guaranteeing the checks and balances needed

to contain executive and legislative power. It constitutes the institutional framework and

functioning agencies within which citizens can translate their abstract legal rights into

practical action.18

In sum, police and courts play central roles in creating basic order and justice or

the oft-cited “rule of law,” which is the bedrock of any political system. With recent

upsurge of criminal violence throughout much of Latin America, these roles are even

more important.

These few examples are offered to illustrate the various sites on a map of policy

making and implementation. Here we are interested in the direct effects of corruption on

democratic governability. We turn now to look at the effects as mediated by perceptions.

Effects of Corruption as Mediated by Public Opinion

The effects of corruption on democratic governability will depend in good part on how

the populace evaluates what they perceive to be corrupt exchanges, assuming that

evaluations influence opinions and behavior. What do we think we know about the

corruption-democracy linkage in a general sense? Findings to date generally point in the

same direction: perceptions of or (self-reported) experience with corruption are generally

correlated with reduced support for (or less confidence in) democracy (taken as regime),

incumbents, policies, or specific agencies. Canache and Allison, for example, use data

from the 1995-1997 World Values Survey and the 1997 Transparency International

16
survey. They find that citizens perceive corruption negatively and they connect those

perceptions to their evaluations of political leaders and institutions. They find no

evidence, however, that perceptions of corruption have soured mass opinion on

democracy as a form of government. (Canache and Allison 2003, pp. 18-19). Seligson

(2002) finds that self-reported experience with corruption (taken as forms of bribery) is

correlated with reduced support for key institutions of government in four Latin

American countries. Other researchers report that perceptions of corruption are associated

with withdrawal and voter abstention. “In brief, perceptions of corruption can reinforce

the existing distribution of political power by effectively removing from the political

arena citizens who may have reason to challenge the system” (Davis et al. 2004, p. 701).

A deepening of democracy means the opposite, the opening of channels through which

discontented citizens can find effective voice through democratic participation.

These studies provide a useful point of departure. How might we reformulate the

democracy-corruption relationship to examine the perception linkages in a more

disaggregated way? We begin with the basic question: What do individuals perceive as

“democracy” and “corruption”? What do the terms mean in everyday life, in different

countries, and at different times?

The little bit of survey material at hand suggests that mass publics conflate several

attributes in their perceptions of democracy. With respect to Costa Rica, for example,

Jorge Vargas Cullel (2005, p. 115) reports that about 30 percent of respondents in a 2001

national survey defined democracy in basically procedural terms, about 21 percent chose

an “expanded definition” that included better representation and improved citizen control

over the government, about eight percent responded in behavioral terms (better

17
leadership), another 10 percent emphasized economic progress and equality, and the

remaining 31 percent mixed in elements from all the preceding definitions.

Camp (2001, p. 9) reports that surveys on Costa Rica, Mexico, and Chile suggest

that most Latin Americans don’t conceptualize democracy or have the same expectations

from democracy as do North American theorists or citizens. “(W)hat most distinguishes

the Latin American version of democracy from that of the United States is its emphasis

on social and economic equality and progress.” The value of Camp’s comparative

approach is to remind us about the important national differences in conceptions of

democracy. His findings suggest that of the three countries Costa Rica most nearly

resembles the U.S. emphasis on democracy in political terms, while the other two

emphasize government performance (Camp 2001, p. 15).

What meanings do mass publics assign to corruption? Based on data from

Transparencia Mexicana surveys done in 2001 and 2003, Bailey and Paras (forthcoming)

found that Mexican respondents identified a wide range of different phenomena as

corrupt. “Mordidas” (bribes) make up the most important category, and this meaning

coincides with most scholars’ definitions. But it is cited by only a fifth to a quarter of the

respondents. Dishonesty, by either government or citizens, makes up the next largest

category with 15 percent. Then comes a laundry list of items with no obvious logical

connections, ending in a large “don’t know” response. In sum, a respondent who cites

“corruption” may have in mind bribery, dishonesty, failure to implement a program, poor

quality administration, or something else. And the mix probably varies from one country

to the next.

18
 Where is “corruption” seen to exist? A 2005 survey carried out by Transparency

International in 15 Latin American countries asked about the degree to which corruption

affects different spheres of life in the country.19 The term “corruption” is undefined but

was scored on a 5-point scale from least to most present. Two sets of findings are

pertinent. First, respondents make no sharp public/private distinctions. Corruption is seen

in both society and government. Both the private sector (3.5) and communications media

(3.3) fare rather poorly, with religious bodies seen as least corrupt (2.8). Second, the most

corrupt public sectors are: customs (4), legal system/judiciary (4.3), police (4.3), the

legislature (4.4), and political parties (4.5) (TILAC 2005).

Noting the presence of corruption in different sites need not mean that those are

considered the most serious. Transparency International’s Global Corruption Barometer

surveyed some 40,838 people in 47 countries in July 2002. The barometer tapped into the

impacts of corruption on different spheres of life as well as perceptions of change over

time. “The most striking finding came when respondents were asked from which

institution they would choose to eliminate corruption first if they had a magic wand. The

overwhelming first choice was political parties, followed by the courts and the police”

(TI 2004, p. 288). We’ll reconsider below this apparently strange assortment of actors.

In a survey done in Nicaragua that presented various situations of corruption,

Seligson (1997, pp. 9-12) found that respondents were most negative in their judgment of

legislators who take bribes from private companies. Respondents were less critical about

teachers accepting gifts, transit police soliciting bribes, or public employees accepting

payments. They were least critical of a mayor who loans out the town’s tractor to

construct a baseball field. With respect to judgments about corrupt practices, respondents

19
were most negative about businesses paying bribes and about students giving gifts, but

were less so about using connections to get a job, paying bribes to transit police, or

paying a bribe to get a birth certificate.20

Limited as they are, these bits of data suggest interesting possibilities for

reformulating the democracy-corruption relationship in terms of perceptions. With

respect to democracy as regime, respondents appear to be particularly negative about

political parties; with respect to state institutions they focus on police and the courts.

Political parties play the central role of linking mass preferences to elite

competition for control of government. Public estimation of parties is generally low

across most countries. It isn’t obvious what precisely is perceived to be corrupt about the

parties themselves. Probably more at issue is the perception of some mix of campaign

finance abuses, forms of electoral fraud, and rhetoric that is considered to be deceptive.

The negative evaluation of the police is a recurring finding throughout most of the

region. That the judiciary is also cited as a priority concern among government functions

is a bit of a puzzle. If citizens come into daily contact with the police, and thus have some

basis for judgment, many fewer would seem to interact with the courts considered

narrowly as judges and judicial proceedings. One possibility is that respondents are

including as “judicial power” (poder judicial) the offices of the ministerios públicos

(roughly prosecuting attorneys) and the policía judicial (investigative police). Citizens

would have more possibilities of contacts with these offices in the course of filing

complaints or of being party to a criminal investigation. In many countries in the region

these two offices have ample track records of corrupt behavior.21 Another possibility

involves the upsurge of concern over criminal violence since the mid 1980s and the

20
demonstrated inadequacy of the legal system to respond. “Corruption” here may be the

respondents’ way to condemn ineffectiveness, as well as to convey dissatisfaction judges

and court officials on the take.

Bo Rothstein (forthcoming, p. 149) offers a hint about why law enforcement and

judiciary are critical to democracy in terms of linkages of perception. He suggests strong

connections between law enforcement and social trust, a crucial element of democratic

governability:

The institutions of law and order have a particularly important task in a civilized

society compared with other institutions, i.e., to track down and punish people

who have committed acts that make them undeserving of social trust. If people

believe that these institutions perform that task effectively and fairly, they also

have reason to believe that people have little chance of getting away with criminal

behavior. Accordingly, it is reasonable to conclude that most people refrain from

such behavior, from which it follow that people believe that it is generally

possible to “trust most people.”

As a category separate from functions (parties, law enforcement) I would also

underline the importance of grand corruption. Recall Seligson’s findings about the

seriousness that respondents attach to corrupt behavior by legislators who accept bribes

and by the business firms that offer them. These exchanges approximate our notion of

grand corruption: misbehavior by high-level officials involving substantial sums.

Perceptions of grand corruption also give us clues about an interesting puzzle in the

democracy-corruption linkage. Self-reported experience with corruption (which runs

from the low teens to the mid-20s percentage range) is much less than perceptions of

21
corruption in government as a whole or in specific institutions (which run in the mid-60s

into the 70-percent range).22 Why do relatively few respondents say they experienced

corruption yet so many express negative evaluations of distant actors and agencies? A

three-part hypothesis can account for the disparity. First, respondents probably

underreport their experience with corruption. Second, respondents base their views about

distant institutions on media reports, which may or may not be accurate but which tend to

emphasize incidents of grand corruption, often in the context of scandal. Third,

respondents may be dissatisfied with government performance or about their particular

circumstances and choose “corruption” as an all-purpose label to express a negative

feeling.23

Conclusions: Recasting the Democracy-Corruption Linkages

Corruption is rightly seen as a prominent cause of low quality democracy throughout

much of Latin America. The scholarly and policy work in this field has progressed

considerably over the past twenty years. In this paper I have tried to make four basic

points. First, it is useful to differentiate between democracy as regime apart from state

institutions and the legal system in which the regime operates. Second, it is also useful

make some basic differentiations among arenas of policy-making and to consider the

nature of corruption as it operates in these sites. Third, we need to posit analytical

linkages between corruption and regime and state in ways that set orders of priority

among functions and processes. Fourth, apart from the direct linkages, the perceptions of

corrupt exchanges constitute an important set of mediated linkages. A challenge inherent

in examining perceptions is that the terms “democracy” and “corruption” in daily usage

are complex, vague, multi-dimensional, and confused.

22
 Space limitations excluded discussion of the mixes and interactions of different

types of corruption. Do scandal and grand corruption, for example, magnify negative

perceptions of ordinary corruption? Nor was the broader context considered. For

example, is corruption viewed less negatively in an economy experiencing solid growth

and low inflation? Is the problem less serious early in a fixed term than later?24

I have ventured some speculations about which types of corruption operating in

particular arenas might be somehow more important to the operations of democracy and

state institutions. These particular guesses might prove wrong as our understanding of the

origins and maintenance of democracy advances. Even so, I think the rationale behind the

speculations remains sound: We need to unpack and disaggregate the broad concepts that

we have employed up to this point in order to analyze more effectively their

interrelationships.

Figure

Figure 1. Dynamics of Corrupt Transaction

Agent
Not Request
Request Bribe
Bribe
Client

Offer
Y,Y Y,N
Bribe

Not Offer
N,Y N,N
Bribe
Source: Antolova (2003, p. 2)

23
Bibliography

Antolova, Lucia. 2003. “Typology of Corruption,” (unpublished paper, December 16).

Bailey, John and Lucía Dammert. 2006. Public Security and Police Reform in the
Americas (Pittsburgh: University of Pittsburgh Press).

----- and Pablo Paras. forthcoming. “Perceptions and Attitudes about Corruption and
Democracy in Mexico,” Journal of Mexican Studies/Estudios Mexicanos.

Bayart, Jean-Francois, Stephen Ellis and Beatrice Hibou. 1999. “From Kleptocracy to the
Felonious State,” in Bayart, Ellis and Hibou, eds., The Criminalization of the State in
Africa (Bloomington: Indiana University Press), pp. 1-31.

Bhargava, Vinay and Emil Bolongaita. 2004. Challenging Corruption in Asia: Case
Studies and a Framework for Action (Washington, D.C.: The World Bank).

Camp, Roderic Ai. 2001. “Democracy through Latin American Lenses: An Appraisal,” in
Roderic Ai Camp, ed. Citizen Views of Democracy in Latin America (Pittsburgh:
University of Pittsburgh Press), pp. 5-23.

Canache, Damarys and Michael Allison. 2003. “Corrupted Perceptions: The Effect of
Corruption on Political Support in Latin American Democracies,” Prepared for delivery
at the XXIV International Congress of the Latin American Studies Association, Dallas,
Texas, March 27-29.

Dahl, Robert. 1971. Polyarchy: Participation and Opposition (New Haven: Yale
University Press).

Davis, Charles L., Roderic Ai Camp and Kenneth M. Coleman, 2004. “The Influence of
Party Systems on Citizens’ Perceptions of Corruption and Electoral Response in Latin
America,” Comparative Political Studies, 37:6 (August), 677-703.

Easton, David. 1965. A Systems Analysis of Political Life (New York: John Wiley &
Sons).

Heidenheimer, Arnold J. and Michael Johnston, eds. 2002. Political Corruption:

Concepts and Contexts (New Brunswick, N.J.: Transaction Publishers).

-----. 2002 [1970]. “Perspectives on the Perception of Corruption,” in Heidenheimer,

Arnold J. and Michael Johnston, eds. 2002. Political Corruption: Concepts and Contexts,
pp. 141-154.

Johnston, Michael. 2002. “Right and Wrong in American Politics: Popular Conceptions
of Corruption,” in Arnold J. Heidenheimer and Michael Johnston, eds. Political

24
Corruption: Concepts and Contexts (New Brunswick, N.J.: Transaction Publishers), pp.
173-194.

Karklins, Rasma. 2005. The System Made Me Do It: Corruption in Post-Communist

Societies (New York: M.E. Sharpe).

Levitsky, Steven. 2003. “The Labor Politics to Machine Politics: The Transformation of
Party-Union Linkages in Argentine Peronism, 1983-1999,” Latin American Research
Review, 38:3 (October), 3-36.

Manzetti, Luigi. 2000. “Market Reforms without Transparency,” in Joseph S. Tulchin

and Ralph H. Espach, eds., Combating Corruption in Latin America (WasHington, D.C.:
Woodrow Wilson International Center for Scholars), pp. 130-172.

Moreno, Alejandro. 2003. “Corruption and Democracy: A Cultural Assessment,” in

Ronald Inglehart, ed., Human Values and Social Change: Findings from the Values
Surveys (Boston: Brill), pp. 265-277.

Nye, Joseph S. 2002 [1967] . “Corruption and Political Development: A Cost-Benefit

Analysis,” in Arnold J. Heidenheimer and Michael Johnston, eds., Political Corruption:
Concepts and Contexts (New Brunswick, N.J.: Transaction Publishers), pp. 281-302.

O’Donnell, Guillermo. 2004. “Human Development, Human Rights, and Democracy,” in

Guillermo O’Donnell et al., eds., The Quality of Democracy: Theory and Applications
(University of Notre Dame), pp. 9-92.

Peters, John G. and Susan Welch. 2002. “Gradients of Corruption in Perceptions of

American Public Life,” in Arnold J. Heidenheimer and Michael Johnston, eds. Political
Corruption: Concepts and Contexts (New Brunswick, N.J.: Transaction Publishers), pp.
155-172.

Pfeiffer, Silke. 2004. “Vote Buying and Its Implications for Democracy: Evidence from
Latin America,” (TI 2004, pp. 76-83).

Rehren, Alfredo. 2006. “Corrupción y crisis del estado democrático latinoamericano,”

Paper prepared for delivery at the 2006 Meeting of the Latin American Studies
Association, San Juan, Puerto Rico, March 15-18, 2006.

Reno, William. 2000. “Clandestine Economies, Violence and States in Africa,” J. of

International Affairs, 53:2 (Spring), 433-459.

Speck, Bruno W. 2004. “Campaign Finance Reform: Is Latin America on the Road to
Transparency?” (TI 2004, pp. 32-35).

Rothstein, Bo. forthcoming. Social Traps and the Problem of Trust (Cambridge
University Press).

25
Seligson, Mitchell A. 1997. “Los Nicaraguenses hablan sobre la corrupción: Un estudio
de opinión pública,” (Casals & Associates: Latin American Public Opinion Project)
(March).

-----. 2001. “Transparency and Anti-Corruption Activities in Colombia: A Survey of

Citizen Experiences,” (Casals & Associates, October).

-----. 2002. “The Impact of Corruption on Regime Legitimacy: A Comparative Study of

Four Latin American Countries,” Journal of Politics, 64:2 (May), 408-433.

Seligson, Mitchell A., José R. López-Cálix and Lorena Alcázar. 2006. “Does Local
Accountability Work? Tracing ‘Leakages’ in the Peruvian ‘Vaso de leche’ Program,”
Paper prepared for delivery at the 2006 Meeting of the Latin American Studies
Association, San Juan, Puerto Rico, March 15-18, 2006.

Taylor, Matthew M. and Vinicius Buranelli. Forthcoming. “Ending Up in Pizza:

Accountability as a Problem of Institutional Arrangement in Brazil,” Latin American
Politics and Society.

Thompson, Dennis. 1993. “Mediated Corruption: The Case of the Keating Five,”
American Political Science Review, 369-381.

TI--Transparency International. 2004. Global Corruption Report (London: Pluto Press).

-----. 2005. Global Corruption Report (London: Pluto Press).

TILAC—Transparency International Latinoamérica y Caribe. 2005. Barómetro Global de

la Corrupción 2005; Latinoamérica: datos y conclusions (December 9).

Vargas Cullel, Jorge. 2004. “Democracy and the Quality of Democracy: Empirical
Findings and Methodological and Theoretical Issues Drawn from the Citizen Audit of the
Quality of Democracy in Costa Rica,” in Guillermo O’Donnell et al., eds., The Quality of
Democracy: Theory and Applications (University of Notre Dame), pp. 93-162.

Warren, Mark E. 2004. “What Does Corruption Mean in a Democracy?” American J. of

Political Science, 48:2 (April), 328-343.

Endnotes

1
Georgetown’s School of Foreign Service and its Center for Peace and Security Studies provided helpful
support for this project. Enrique Bravo contributed helpful research assistance.
2
O’Donnell (2004, pp. 15-20) emphasizes the need to differentiate between democratic regime and
democratic state.

26
3
Heidenheimer and Johnston (2002, pp. 3-14) discuss public office, market, public interest, and public
opinion as the main approaches to defining corruption. Warren (2004) is a good example of a broader
conception of corruption, one that involves abuse of a position of trust for private gain and to the detriment
of the shared interest. His approach encompasses corrupt exchanges in the private realm as well.
4
For example, Heidenheimer’s (2002 [1970]) notions of “black, gray and white” corruption.
5
World Bank terminology differentiates between administrative corruption: “ . . . private payments to
public officials to distort the prescribed implementation of official rules and policies,” and state capture: “ .
. . shaping the formation of the basic rules of the game (that is, laws, rules, decrees, and regulations)
through illicit and non-transparent private payments to public officials” (quoted in Bhargava and
Bolongaita 2004, p. 24). Thompson (1993) is a valuable discussion of what he calls “mediated corruption,”
or the types of corruption that actors engage in to gain influence in the political process through borderline
or unethical acts that may not benefit the actors’ personal finances but which strengthen their position.
6
The imagery is drawn from systems analysis, e.g., Easton (1965). The output side is arguably less
politically sensitive and more amenable to policy intervention.
7
Recall Robert Dahl’s influential list of requisites for polyarchy (Dahl 1971, Ch. 1).
8
My approach is vulnerable to the criticism of reviving a politics-administration dichotomy that has proven
to be misleading. Even so, one can make rough—but useful—distinctions between inputs, rule-making an
rule application.
9
The following discussion draws on Antalova (2003).
10
Conversations about corruption among our Africanist colleagues focus on failed states and near-anarchic
violence in which ruling elites deliberately undermine legal systems and public bureaucracies in order to
reinforce their own power and wealth. See for example Reno 2000 and Bayart et al., 1999.
11
Dahl (1971, Ch. 1) sets out the most frequently cited requisites of polyarchy. In an interesting exercise of
ranking the gravity of corrupt acts, Karklins (2005, p. 25) includes “undermining elections and political
competition” and “corruption in and of the media” among the most serious (Level III) forms of corruption.
12
By way of anecdotes, Da Silva (2000, pp. 188-189) describes the complex relationships between the
Brazilian media and government. The “price” of a journalist may be in terms of money, information or
praise, and a politician needs to be careful about what to offer. “Outside Sao Paulo and Rio, it is still the
rule rather than the exception that journalists work for both a newspaper and as public servants in
government, although as long as they write positive things, or at least not negative things, about their
government bosses, they need not show up at the office.” A Mexican friend once commented on journalists
who blackmail politicians and government officials to not publish information.
13
Gallup Argentina (1999, pp. 47-48), for example, finds vote-buying in primary elections of Justicialismo
and La Alianza.
14
See Alfredo Rehren’s useful discussion (Rehren 2006).
15
Manzetti 2000 is a useful discussion.
16
As one of my students in a seminar at the Universidad Católica in Cordoba, Argentina, put it in 2004:
“This analysis of principal-agent-client is interesting, but what do you do when the principal is the most
corrupt point in the triangle?”
17
This is quite a sweeping generalization. The Chilean Carabineros were something of an exception in the
six countries covered in Bailey and Dammert 2006.
18
The creation of special anti-corruption agencies in many Latin American countries is a reminder that the
judicial system is seen as inadequate to deal with corruption.
19
“?En qué medida cree usted que la corrupción afecta diferentes esferas de la vida en este país?”
20
Peters and Welch (2002) and Johnston (2002) provide useful discussions of gradients of “seriousness” of
corruption in the U.S. case. Their findings are roughly in line with Seligson’s.
21
Enrique Bravo made this useful observation.
22
Seligson (2001, p. 26) reports that “(O)ver 70% of our respondents believed that corruption was common
or very common. This finding contradicts the information provided by respondents regarding direct and
indirect experience with corruption, and suggests that Colombians as a general rule believe that the levels
of corruption that escape the public eye are very underreported.” Bailey and Paras (forthcoming) find a
similar disparity in the Mexican case.
23
I have no data from Latin America at hand to support the hypothesis. Eric M. Uslaner finds that high-
level corruption has greater negative effects on trust than “ordinary corruption.” He did a 2003 survey of

27
Romanians that distinguished between high- and low-level corruption. “Low-level corruption focuses on
the extra ‘gift’ payment people make to doctors, banks, the police and teachers. High-level corruption
involves bribery and corruption by politicians, business executives and the courts. We found that low-level
corruption did not lead to lower levels of trust. Most people do not see these payments as making others
‘rich,’ and often see such ‘gifts’ as making a bureaucratic system more efficient. In contrast, we found that
high-level corruption affected perceptions of growing inequality, the evaluations of the performance of the
government improving the quality of life, and trust both in other people and in government. Having to pay
off city officials and especially officers of the court, together with the beliefs that most politicians and
business people were corrupt, led people to believe that the system was stacked against them, that
government could not be trusted, and than even ordinary people were not trustworthy. There are different
types of corruption, and citizens in post-communist countries clearly distinguish among them” (TI 2005,
pp. 263-264).
24
Central America in the past decade would seem to be an interesting testing ground, especially since it
includes the recent experience of Costa Rica where several past presidents have been accused of grand
corruption.

28