12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)
P1: IKH
10.1146/annurev.polisci.6.121901.085655

Annu. Rev. Polit. Sci. 2003. 6:23356

doi: 10.1146/annurev.polisci.6.121901.085655
c 2003 by Annual Reviews. All rights reserved
Copyright
First published online as a Review in Advance on Feb. 6, 2003

ELECTORAL FRAUD: Causes, Types, and

Consequences

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

Fabrice Lehoucq
Division of Political Studies, Centro de Investigaciones y Docencia Economica (CIDE),
Carret. Mexico-Toluca 3655, Lomas de Santa Fe, Mexico City, DF, Mexico 01210;
email: [email protected]

Key Words electoral corruption, ballot rigging, vote buying, electoral laws
Abstract This article reviews research on electoral fraudclandestine and illegal
efforts to shape election results. Only a handful of works classify reports on electoral
fraud to identify its nature, magnitude, and causes. This review therefore looks at the
larger number of historical works (as well as some ethnographies and surveys) that
discuss ballot rigging. Its conclusions are threefold. First, fraud takes on a panoply of
forms; it ranges from procedural violations of electoral law (that may or may not intend
to distort results) to the outright use of violence against voters. Second, even when ballot
rigging is an integral part of electoral competition, it is infrequently decisive. Fraud,
nevertheless, undermines political stability because, in close races, it can be crucial.
Third, political competition shapes the rhythm and nature of electoral fraud. Efforts to
steal elections increase with inequality, but competitivenesswhich institutions help
to shapedetermines the ballot-rigging strategies parties adopt.

INTRODUCTION
The illusion of transparency (Gueniffey 1993) is responsible for the scarcity
of scholarship on the operationand corruptionof balloting procedures. For
decades, researchers assumed that electoral laws straightforwardly converted preferences into outcomes, a notion that encouraged many social scientists and historians to use election results to identify the social bases of political parties. Yet, as
social choice research demonstrates (Riker 1982), the procedures for converting
votes into political power can distort the meaning of the general will and, worse
still, produce a collective choice that violates the preferences of many citizens. Indeed, the opportunities generated by voter registration, polling station operation,
and tallying procedures often lead parties to stuff the ballot box, an outcome that
can wreak havoc with more sociologically minded views of politics.
If, as a first approximation, we define electoral fraud as clandestine efforts to
shape election results, it becomes clearer why scholarship on electoral fraud is so
uncommon. No one who stuffs the ballot box wants to leave a trail of incriminating evidence. Fraud is also difficult for social scientists to study because the most
1094-2939/03/0615-0233$14.00

233

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

234

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

abundant sources for studying fraudfrom first-hand accounts to partisan denunciations of electoral shenanigansare not objective. These obstacles, however,
are not insurmountable. Indeed, once we recognize that important areas of human
activity are typically beyond the gaze of the social scientist, yet nevertheless are
studied (e.g., crime and human sexuality), there is no reason to suppose that fraud
is empirically intractable.
In this review, I critically assess the handful of studies that discuss the types,
magnitude, and determinants of electoral fraud. Like entomologists, historians
and some social scientists concentrate on cataloguing and describing fraudulent
activities. Historians, in fact, have been among the few analysts to move beyond
the anecdotal or prurient interests of so many chronicles of the electorally illicit
accounts that titillate and amuse us. Several historians and social scientists have
also explored what I call the ethnography of electoral fraud. The ethnographers
of fraud help us understand what separated right from wrong to the people who
participated in fraud-tainted elections. Understanding justifications for activities
that the contemporary observer condemns (perhaps too easily), the anthropologists
of fraud try to make sense of the dilemmas that fraud posed to its perpetrators, its
audience, and its observers.
I also survey efforts to assess the consequences or the magnitude and causes
of electoral fraud. Efforts to assess the role of electoral fraud in politics range
from claims that fraud is an isolated, random event to claims that ballot rigging
was rampant in, for example, the United States until the 1960s. Understanding the
extent of fraud is indispensable for deciding whether it merits systematic attention.
I then examine the handful of accounts that conjecture why and how it happens.
Finally, I classify hypotheses about the causes of fraud.
Selecting work for review was difficult for two reasons. In the first place, social
scientists and historians have not written much about electoral fraud. I found fewer
than two dozen articles with titles that contained such terms as electoral fraud,
electoral corruption, vote buying, or ballot box stuffing. I found a slightly larger
number of legal texts on fraud and vote buying, which are typically normative
discussions of what should be considered illegal electoral behavior. In the second
place, a great number of articles and books refer to electoral fraud in very general
terms. Indeed, it is hard to study any democratic or quasidemocratic system of the
past 200 years without acknowledging that governments or their rivals resorted to
less than transparent methods to alter the results of the ballot box.
I therefore base my review on articles and books that analyze ballot rigging
from an empirical perspective. By empirical, I mean any effort to document
and theorize about the types, magnitude, and spatial distribution of ballot rigging.
Though electoral fraud plagues political systems around the world, I only found
research about the countries of the Americas, Asia, and Europe and available in
English, French, and Spanish. I also discuss books and articles that have important
things to say about electoral fraud, even if their subject matter revolves around
clientelism, political parties, elections, or the operation of the formal political
arena.

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

ELECTORAL FRAUD

P1: IKH

235

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

WHAT IS ELECTORAL FRAUD? HOW DO WE STUDY IT?

Common sense initially suggests that an activity is fraudulent if its perpetrator
wants it hidden from the public gaze. Manifestly fraudulent behaviorssuch as
coercing voters at the polling station to cast ballots for party X or filling the ballot
box with votes for party Xare things that only its victims want publicized. Even
procedural violations, such as a polling station opening late and closing early or
failing to advertise its location before election day, sound like fraud because the
accused would prefer that no one learn of these facts. Both blatantly coercive acts
and voting irregularities are also fraudulent because they can throw election results.
There is a fine line between fraud and political pressure. To take a particularly
glaring example, landlords are not guilty of electoral fraud when all their retainers
vote for their party. Regardless of whether peasants have been duped into voting in
favor of the landlords party or wish to avoid the reprisals associated with voting
against their employer, these activities are not really fraudulent, even if they are
morally reprehensible. Even if we can demonstrate that retainers voted against
their own interests, we cannot call this fraudulent unless a law has been broken.
So, in addition to being concealed and potentially affecting election results, an
act is fraudulent if it breaks the law. Indeed, parties go out of their way to do things
in the dark precisely because they are doing something wrong before or on election
day. They are taking advantage of the legal machinery of the electoral process to
steal an election they believe they cannot win fairly. A key advantage of a legalistic conception of fraud is that it permits assessing the location of the boundaries
between acceptable and unacceptable political activity. Indeed, ethnographers of
electoral fraud need to juxtapose what citizens and parties believe is morally reprehensible with what is legally possible to understand why, at particular places and
times, many activities are no longer deemed appropriate to win an election.
There are six sources for the study of electoral fraud. The first two are
objectivethat is, not partisan in inspiration, even if they are, like any source,
imperfect reflections of reality. One largely unexploited source of information is
scientific surveys. Cornelius (2002) identifies the social and partisan correlates of
vote buying from surveys of the 2000 Mexican election, the one in which the Institutional Revolutionary Party (PRI) was defeated after 70 years in power. To my
knowledge, only a research team at the Latin American Social Sciences Faculty
(Aparicio 2002) in Mexico City has also used surveys to make sense of electoral
fraud. A second and similarly underused method is ethnographic. Like pollsters,
ethnographers use interviews to make sense of electoral law violations. Schaffer
(1998, 2001) is one of the few political scientists to sit with villagers in developing
countries to plumb their views of democratic practices; he has done this in Senegal
and the Philippines. Schedler (2002b) also uses in-depth interviews to gauge the
attitudes of Mexicans toward what is known as electoral alchemy.
The next three information sources are partisan. One consists of the memoirs left by the participants or the victims of electoral chicanery. Like ethnographic accounts, memoirs often sacrifice statistical representativeness for detailed

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

236

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

coverage of fraud of particular places and times. Former U.S. President Jimmy
Carters (1992) tale of how he personally witnessed his rivals in the Georgia
Democratic Party stuff the ballot box against him is a marvelous example of this
genre. A second partisan source consists of the denunciations of fraud that parties
file with political authorities. Historically, parties sought legal redress of violations
of electoral law with legislatures. According to eighteenth- and nineteenth-century
constitutional doctrine (Lehoucq 2002a), executives ran elections and legislatures
certified their resultsan institutional framework that often placed losers in the
paradoxical position of seeking justice from their aggressors. Bensel,1 Anderson
(2000), Charnay (1964), and Lehoucq & Molina (2002) study complaints filed with
the U.S. Congress, Imperial German Reichstag, French National Assembly, and
Costa Rican Congress, respectively. A third partisan source consists of newspaper
accounts, a body of information that Cox & Kousser (1981) use to understand
electoral corruption in late-nineteenth-century rural New York.
A final source of information consists of the reports of organized civil society.
As international electoral observation took off in the last decades of the twentieth century (Middlebrook 1998, Pastor 1999), nongovernmental organizations
(NGOs), with domestic and international funding, have often displaced parties as
poll watchers. Although NGOs may be more impartial than parties (then again, no
one really knows) in monitoring activities on election day, their increasing visibility deprives parties of a principaland unrecognizedfunction, as I explain later
in this review. The Mexican Civic Alliance organized more than 18,000 observers
to monitor more than 10,000 polling stations in 1994, an effort that has generated
a voluminous record (Calderon Altzati & Cazes 1996, Cazes 1996).

TYPES OF BALLOT RIGGING

It is hard to find a book on unreformed political systems that does not refer to electoral fraud. Pre-reform systems are polities that do not meet minimal democratic
standards, that is, they restrict franchise rights to something less than all adults of
age 18 or older and civil liberties are unenforced. More important, elections are
imperfectly administered in unreformed systemswhich leads to outcomes whose
credibility ranges from nonexistent to partial. Notable accounts of unreformed systems that, to some extent, analyze electoral fraud include books by Botana (1979)
and Sabato (2001) on Argentina, Anderson on Germany (2000), Basadre on Peru
(1980), Graham on Brazil (1990), Hanham (1959) and Seymour (1915) on England,
Hoppen (1984) on Ireland, Kousser (1974) and Perman (2001) on the U.S. South,
and Varela-Ortega on Spain (2001a,b). Much of this work suggests that fraud was
commonplace, even if relatively few authors try to classify and count acts of ballot
rigging. I begin by discussing the works that refer to electoral fraud in the most
general terms before moving on to more systematic efforts.
1

Bensel RL. 2002. The American Ballot Box: Law, Identity, and Voting, 18501900. Unpublished manuscript.

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD

P1: IKH

237

Both Kousser (1974) and Perman (2001) rely on legislative records, newspapers,
and census records to relate how the Democratic Party in the 11 states of the former
Confederacy deprived African Americans and poor whites of their suffrage rights
through the end of the nineteenth century. Even though the Republican Party
added the thirteenth (1865), fourteenth (1868), and fifteenth (1870) amendments
to the Constitution to ensure the suffrage rights of recently emancipated slaves,
the planter-dominated Democratic Party gained control of southern governorships
and legislatures to enact poll taxes and other restrictive procedures to eradicate the
electoral base of their Populist and Republican opponents. Kousser (1974) tells
this story particularly well, with abundant examples of the Democratic Partys use
of fraud and violence to regain control of southern politics.
Students of U.S. politics, nevertheless, have not written systematic accounts
of ballot rigging, even though irregularities continue to plague U.S. elections
(National Commission on Federal Election Reform 2002). Bensels (2002) study
is an exception. Based on the appeals losers filed with the U.S. House of Representatives during the heyday of the public ballot (18501900), Bensel shows
that the most common allegations of fraud involved either intimidating or using
violence against undecided or opposition voters and/or questioning whether voters
met suffrage requirements. In urban areas, citizens cast party-supplied ballots in
precincts that their party and/or ethnic group controlled. Population densities necessitated multiple polling stations in the nineteenth-century urban United States,
an administrative fact that allowed rival parties and factions to capture precincts
that parties made safe for their followers and dangerous for their adversaries. In
a rural community, peer pressure could replace blunter forms of intimidation to
ensure results at its sole polling station.
A rare and pioneering study of rural New York examines why parties went from
paying citizens to vote to paying them to stay at home on election day (Cox &
Kousser 1981) after this state adopted the Australian Ballotthe use of a single,
government-supplied ballotin 1890. Another paper documents how a Court of
Common Appeals judge disenfranchised some 1700 voters or 26% of the electorate
in Adams County, Ohio in 1910 for having sold their votes for prices ranging from
a drink of whiskey to $25 (Gist 1961, p. 77), some 20 years after Ohio residents
got Australian Ballots. The overall absence of scholarship nevertheless is notable
because Congressional and Judicial archives contain a wealth of evidence about
these activities (Argersinger 1992). Unlike in many other democracies, legislative
bodies still certify election results in the United States, a process that has produced
a record of contested House elections and recounts (Garber & Frank 1990).
Historians have produced studies that shed light on pre-reform political systems
in several Latin American countries (Annino 1995, Malamud 2000, Posada-Carbo
1996, Sabato 1999). In his review of electoral corruption in Latin America between 1830 and 1930, Posada-Carbo (2000) lists the research that examines how
governments and parties stuffed the ballot box. The author draws attention to how
commonplace fraud was and how little we understand it. In an all-too-general
analysis of Guatemalan elections, Sloan (1970, p. 78; see also McDonald 1967)

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

238

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

also points out that electoral fraud has never been studied carefully, even though
the case for choosing the revolutionary path of development is partly dependent
upon proof that the electoral system is corrupt.
In his own work on Colombia, Posada-Carbo (1994, 1997, 2000) uses newspapers, party pamphlets, and period publications to show how Colombian conservatives and liberals were organized into political machines and that fabricating
votes was an integral part of electoral strategy. Graham (1990) shows that the
constitutional monarchy of nineteenth-century Brazil had a political life remarkably similar to that of nineteenth-century England. An electorate consisting of
50.6 percent of all free males, 20 years or older, regardless of race or literacy
voted for parliaments that governed alongside Brazilian emperors (Graham 1990,
p. 108). Graham shows that cabinets took elections very seriously because they
needed to maintain the parliamentary majorities necessary to enact the federations
laws. As a result, elections were contests among regional power brokers, where
fraud and violence determined who occupied legislative and other elected posts.
In a study of the 1872 election in Peru, Mucke (2001) shows that parties struggled to gain control of town squares, where winners held elections. Losers, in
turn, would organize their own polling stations in squares of lesser importance.
Rival polling stations often squared off against each other, and did little to collect
the votes from all voters. Indeed, the idea of an impartial polling station receiving the votes of all citizens seemed to be the farthest thing from parties minds.
Both then forwarded their results to parallel provincial boards, which would each
then send them to Congress, in which partisan majorities certified the results they
found most acceptable. Basadre (1980), the great Peruvian historian, describes
how electoral fraud prompted legislators to reform laws to make it harder to steal
elections.
Ballot rigging is an integral part of Argentine political history. Argentine historians typically argue that electoral fraud has been pervasive (Ferrero 1983). Tjarks
(1963) relates how, in the 1876 elections in the Province of Salta, government
forces and their opponents surrounded polling stations with militias to prevent
their rivals from voting, purchased votes and polling station officials, and purged
the electoral rolls of rival voters. Using legislative records, newspapers, and private
correspondence, Sambruccetti (1980) mentions that parties packed the electoral
rolls, set up parallel polling stations, bought votes, and destroyed opposition ballots
in the 1886 elections. Both also discuss government parties use of their legislative majorities to dismiss allegations of fraud to certify the vote. The accounts
of Tjarks (1963) and Sambruccetti (1980) also suggest that, in Argentina, packing the electoral registry with the names of ineligible males was a key source of
fraud. In a society receiving massive numbers of immigrants from the Old World,
parties faced powerful incentives to register any male aged 18 years or older (all
of whom possessed suffrage rights), regardless of whether he had become a naturalized Argentinian. According to Botanas (1979) classic study of the Argentine
conservative order between 1886 and 1916, fraud permitted provincial governors
to deliver the requisite votes so that their party could maintain control of the

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD

P1: IKH

239

presidency, the Chamber of Deputies, and the provincial legislatures (which appointed the members of the federal Senate).
Vote buying was the key source of fraud in nineteenth-century England (Hanham
1959, OGorman 1989, OLeary 1962, Seymour 1915) and Ireland (Hoppen 1984).
Even after the 1832 Reform Bill, which outlawed many of the districts with less
than a handful of registered voters, English grandees used their parliamentary
influence to maintain what were known as the rotten boroughs. The continuation
of public voting allowed landlords to control the behavior of their retainers in
the counties, the largely rural constituencies, without overt fraud. In the urban
boroughs, the enfranchisement of an increasing number of voters fueled a market
of votes, an outcome that did not begin to decline until after the establishment of
the secret franchise in 1872 in England and Ireland. Along with small district size
and the casual enforcement of anticorruption laws (until 1868, the English House
of Commons reviewed petitions to nullify electoral results, few of which it ever
endorsed), rampant electoral corruption began to dissipate after the dismantling of
the public franchise and the enactment of anticorruption Acts.
Vote buying is a major issue in several Asian political systems as well, even
if specialists complain about scarcity of evidence for the market of votes. Rigger
(2002) cites a survey that found more than two thirds of voters and politicians in
Taiwan believed candidates offered favors and/or cash to vote on election day. The
consensus among specialists on the island, Rigger notes, is that vote buying declined in the 1990s, which she attributes to growing electoral competition and the
urbanization of Taiwanese society. Callahan (2000, Callahan & McCargo 1996)
and Hicken (2002) discuss the centrality of vote buying in Thailand. Before the
1997 constitutional reform, Thai voters cast as many ballots as there were seats in
their plurality districts. Hicken argues that a surplus of votes encouraged citizens
to sell votes to the highest bidder. Both Callahan and Hicken contend that vote buying has declined since 1997, when Thailand adopted a mixed-member electoral
system (400 single-member, plurality seats plus 100 seats selected in a nationwide district through proportional representation) and an independent electoral
commission that aggressively prosecuted vote buyers and vote sellers.
Several historians use complaints filed with the Reichstag to make sense of
nineteenth-century German elections (Anderson 2000; Fairbairn 1990, 1997). In
her study on Imperial Germany, Anderson (2000) uses the election scrutinies or
audits (Wahlprufungen) that parties filed with the Reichstag between 1867, when
the German Empire established suffrage rights for men 25 years or older, and 1912,
the last elections held before the Kaisers abdication. She seeks to understand
how the Germans experienced their new franchise (p. 18). This is an important
point: Andersons book does not classify acts of fraud to understand how and why
votes were fabricated. Her key finding is that Germans did not complain about
bribery, physical force, or the falsification of results on election day. Without providing any hard numbers, she claims that electoral complaints condemned the
Catholic hierarchy, landlords, and factory owners for using influence on voters. And influence ranged from the pressure of the congregation on each of its

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

240

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

members to the threat of eviction that landlords used against retainers and the
possible dismissal that capitalists used against their employees. Much of the controversy surrounding electoral misconduct, Andersons book suggests, revolved
around ways of using procedural violations of electoral law to overturn election
results, especially since the Reichstag considered employer pressure to be a private
matter until 1912 (Anderson 2000, p. 233)11 years after the Reichstag replaced
party-supplied ballots with the Australian Ballot to safeguard the privacy rights
of voters. Fairbairns work is consistent with Andersons portrait; his research
on the 1898 and 1903 Imperial elections indicates that many election protests
involved purely technical errors by officials untrained for their tasks, quirks of
human behaviour, and opportunistic allegations by candidates who had narrowly
lost elections (Fairbairn 1990, p. 817).
Spanish historians have produced the largest national historiography that discusses ballot rigging as part of the post-Francoist effort to come to terms with
the failure of earlier republican experiments in Spanish history (for reviews, see
Darde 1996, Tusell 1991, Varela Ortega 2001a,b). Varela Ortegas (2001a,b) study,
in particular, of restoration Spain between 1875 and 1900 describes how the central
government and locally based strongmen (caciques) manipulated voting returns.
Varela Ortegas book is particularly noteworthy because it combines a narrative
of the election cycle with a listing of violations of electoral laws. He shows that
the government and the caciques, which often opposed each other, tried to gain
the upper hand before election day by purging the electoral rolls of the names
of opposition citizens, a process that the establishment of universal suffrage for
men 25 years or older in 1890 did not thwart. According to Varela Ortega, there
was little fraud in most (largely rural) districts, because collusion between the
government and caciques cooked results in favor of mutually agreed-upon candidates. In more competitive districts, fraud included stuffing ballot boxes with
false votes; ballot substitution; preventing opposition voters from casting ballots; locating polling stations in disagreeable places, such as a hospitals infectious wards; having citizens vote repeatedly or on behalf of deceased, nonexistent,
or opposition voters (this practice when done collectively became known as
escuadrillas volantes or flying squads); and, during the tally of the vote, falsely
claiming that rival candidates had withdrawn from races (Varela Ortega 2001a,
pp. 47879).
Civic Alliance of Mexico uses reports of more than 1870 observer teams of a
stratified sample of rural polling stations to produce a comprehensive portrait of
election-day irregularities and legal violations in 1994, the first time in Mexico
that an independent court and administrative system ran an election (Calderon
Alzati & Cazes 1996, pp. 16989). These observers note that more than two thirds
of the polling stations witnessed the arrival of citizens whose names were not on
the electoral rolls (and who possessed electoral identification). They also find that
more than a third of the voting precincts could not maintain the secrecy of the
ballot and that parties or electoral authorities pressured voters in a quarter of the
polling stations. The observers report numerous procedural violations, including

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD

P1: IKH

241

that more than a third of the polling stations opened after 9:00 a.m. and that results
were not posted outside the polling station after officials and party observers tallied
the vote in 6% of these stations. Civic Alliance also finds that 7.5% of the voting precincts had voters whose thumbs were not dipped in indelible ink; staining
thumbs is a common way of distinguishing voters from nonvoters and thus preventing voters from casting more than one ballot per election. What distinguishes
Civic Alliances work from previous efforts is that it presents a classification of
acts of fraud and quantitatively measures them, objectives necessary to a more
systematic understanding of electoral fraud.
Lehoucq & Molina (2002) use the petitions to nullify electoral results parties
presented to Costa Ricas unicameral Congress between 1901 and 1948 to provide
a comprehensive listing of electoral fraud. Table 1 catalogs the charges parties
made in more than 123 petitions, from which we extract 1131 complaints (and to
whose total the authors add 235 accusations made during the tally of the vote).
Accusations range from the procedural (e.g., opening or closing polling stations
out of schedule) to the hardcore (e.g., stealing ballot boxes) to expelling voters
or poll watchers from the voting precinct. Table 2 lists the charges parties lodged
against polling stations, which rise from 30% of all complaints filed between
1901 and 1913 to 68% of all accusations during the 1940s. I include these tables
to summarize the types of fraud that incumbents, parties, and machines seem to
perpetrate everywhere, not just in prereform Costa Rica.
These tables reveal that the petitions denounce increasingly blatant acts of
fraud through time. Table 2 indicates that the absence of formal requisites and
questions about whether voters met suffrage requirements fell from 74% to roughly
36% of all accusations lodged against polling stations between the 1910s and the
1940s. Table 1 reveals that acts of coercion against voters and polling stations
went from 10% to 15% between these periods. These trends suggest that parties
increasingly began to condemn more blatant types of electoral fraud, a process
fueled by electoral reform and, in particular, the shift from a public voting system
where parties supplied voters with ballots to one where voters cast Australian
ballots in secret.
Under the heading of the nature of fraud, I also want to mention several examples of ethnographic research, efforts that depict the nature and meaning of
ballot rigging. To date, one of the best works in this genre is Aziz-Nassifs
(1987) fine study of the 1986 Governors race in the northern Mexican state of
Chihuahua. Trained as an anthropologist, Aziz-Nassif describes a highly controversial election, one that the opposition claimed the PRI stole from them.
Aziz-Nassif pays careful attention to pro- and antigovernment understandings of
their behavior, apparently interviewing many of the participants before and after election day. His account reflects the moral outrage that both left and right
felt about the PRIs control over yet another election. Aziz-Nassif describes how
a broad antireform coalition formed to contest the PRIs hegemony. He relates
that PRI manipulation of the electoral arena included a biased voter registry, the
pre-election reform of state electoral law to deprive opposition parties of poll

12 Apr 2003

18:35

242

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

TABLE 1 Accusations of electoral fraud by type and time period in Costa Rica, 19011946
(in percentages and numbers)a
Time periods

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

Accusation

19011912

19131923

19251938

19401946

Total

2nd-stage elector
excluded

2.8%
(4)

0.3%
(4)

2nd-stage elector not

qualified to vote

7.8%
(11)

11.9%
(11)

Formal requirements for

2nd-stage election
were violated

5.6%
(8)

0.6%
(8)

2nd-stage elections held

outside of official
time period

0.7%
(1)

0.1%
(1)

Authorities did not take

an electoral census

0.7%
(1)

Official coercion against

2nd-stage electors

0.7%
(1)

Official coercion
against voters

0.7%
(1)

10.9%
(48)

5.5%
(20)

11.3%
(33)

8.2%
(102)

Official coercion against

polling station

3.5%
(5)

0.9%
(4)

0.8%
(3)

4.8%
(14)

2.1%
(26)

Nonofficial coercion
against voters

0.2%
(1)

0.8%
(3)

Official intimidation
against voters

0.3%
(3)

1.4%
(5)

2.8%
(8)

1.3%
(16)

0.3%
(1)

3.4%
(10)

0.9%
(11)

3.1%
(9)

8.7%
(108)

0.2%
(1)

0.2%
(2)
0.1%
(1)

Nonofficial intimidation
against voters

0.3%
(4)

Officials show favoritism

toward a party

45.1%
(64)

3.9%
(17)

5%
(18)

Elected official not

qualified for post

2.1%
(3)

1.4%
(6)

1.6%
(6)

1.2%
(15)

2.3%
(10)
1.1%
(5)

13.7%
(50)
2.5%
(9)

2.4%
(7)

4.8%
(60)
1.7%
(21)

73.5%
(325)

65.1%
(237)

68.7%
(200)

65%
(805)

Purchase of votes
Liquor distributed
on election day
Accusations against
polling stations

30.3%
(43)

Unknown

2.9%
(13)

Other
Total
a

Source: Molina & Lehoucq (1999).

100%
(142)

1.1%
(13)

2%
(9)

3.3%
(12)

3.5%
(10)

2.5%
(31)

100%
(442)

100%
(364)

100%
(291)

100%
(1239)

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

ELECTORAL FRAUD

243

TABLE 2 Accusations of fraud against polling stations by type and time period, 19011946 (in
percentages and numbers)a
Time periods

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

Accusation

19011912

19131923

19251938

19401946

Total

Inappropriate
exclusion of voters

11.6%
(5)

3.7%
(12)

Party observer
expelled or threatened

4.7%
(2)

8.9%
(29)

10.6%
(25)

12.5%
(25)

10.1%
(81)

Absence of formal
requisites

48.8%
(21)

56.9%
(185)

35%
(83)

10.5%
(21)

38.5%
(310)

Voter cast more

than one ballot

2.3%
(1)

3.1%
(10)

5.9%
(14)

3%
(6)

3.9%
(31)

Voting booth in an
inappropriate place

2.3%
(1)

3.1%
(10)

3.4%
(8)

1
(2)

2.6%
(21)

Elections held outside

of official time period

4.7%
(2)

2.8%
(9)

2.1%
(5)

9.5%
(19)

4.3%
(35)

Voters did not meet

requirements
(too young,
not citizens, etc.)

26%
(11)

12.3%
(40)

0.8%
(2)

2.1%
(17)

Voters prevented from

casting ballots

1.2%
(4)

Number of votes
inflated

2.5%
(8)

Number of votes
exceeds number
of voters

0.6%
(2)

Elections were
not held

0.9%
(3)

0.8%
(2)

Votes not
received

0.6%
(2)

0.4%
(1)

Substitution of votes
was permitted

3.1%
(10)

9.3%
(22)

Location of
polling stations
changed on
election day

0.3%
(1)

6.6%
(53)

0.5%
(4)
7.2%
(17)

10%
(20)

5.6%
(45)
0.2%
(2)

1%
(2)

0.9%
(7)
0.4%
(3)

18.5%
(37)

8.6%
(69)
0.1%
(1)

Voters were
intimidated

0.4%
(1)

0.1%
(1)

Ballots were altered

3.4%
(8)

1%
(8)
(Continued )

12 Apr 2003

18:35

244

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

TABLE 2 (Continued)
Time periods

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

Accusation

19011912 19131923 19251938 19401946 Total

Votes were
annulled

1.7%
(4)

4.5%
(9)

1.6%
(13)

Number of ballots
exceeds number of voters

3.8%
(9)

7%
(14)

2.9%
(23)

Electoral identification
was rejected

3.4%
(8)

1%
(8)

Electoral identification not

demanded of a citizen

0.4%
(1)

0.1%
(1)

Number of ballots does not

equal number of
identification cards

0.8%
(2)

0.2%
(2)

Ballots substituted

0.4%
(1)

2%
(4)

0.6%
(5)

Electoral identification removed

2.1%
(5)

3.5%
(7)

1.5%
(12)

Ballots removed

0.4%
(1)

1%
(2)

0.4%
(3)

Ballot box altered

1.3%
(3)

5%
(10)

1.6%
(13)

Voting was public

6.4%
(15)

6%
(12)

3.4%
(27)

2%
(4)

0.5%
(4)

Electoral documentation not

surrendered for legislative
elections

1.5%
(3)

0.4%
(3)

Electoral documentation
opened before election day

0.5%
(1)

0.1%
(1)

Electoral documentation collected

by unauthorized individuals

0.5%
(1)

0.1%
(1)

Wrongfully counted absentee

ballots from other provinces

0.5%
(1)

0.1%
(1)

100%
(200)

100%
(805)

Vote tally conducted by

unauthorized individuals

Total
a

Source: Molina & Lehoucq (1999).

100%
(43)

100%
(325)

100%
(237)

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD

P1: IKH

245

watchers at all voting tables, and the certification of the results by the state legislature (where only the PRI had Representatives) instead of by an impartial electoral
tribunal.
In his exploration of 81 open-ended interviews with poor Mexicans before the
2000 elections, Schedler (2002b) shows that respondents had much more ambiguous views regarding vote trading than images of pliant clients portray. Many
of them expressed misgivings regarding public authorities and parties that exchange individual benefits for political support. Although they find it acceptable
that parties promise to improve social conditionsuniversalistic goods, according
to Schedlermany disapproved of candidates who trade votes for money or benefits. In Senagal, although community leaders and government officials spoke of
the traditional African virtues of public voting, Schaffer (1998) finds that voters
preferred the secret ballot because it allowed them to vote their conscience and
to remain on good terms with members of their family, fellow villagers, and, of
course, local notables and public officials. Most Senagalese, therefore, welcomed
the 1988 reform that reintroduced the mandatory use of (secret) ballot boxes.
The colonial French had established this practice in 1914 in urban areas, and it
cost them the election of the colonys representative in the mainlands National
Assembly.
Through in-depth interviews with 139 registered voters randomly selected
from four areas of Barangay Commonwealth, Quezon City, Philippines, Schaffer
(2001) also shows that many respondents also disapproved of vote tradesexplicit
promises to vote for X in exchange for Yor undeniable acts of corruption
(Rose-Ackerman 1999). They, however, expressed empathy for individuals who
made such deals. They also find nothing controversial about politicians promising to make peoples lot better as long as there is no explicit quid pro quo for
their support. Distinguishing between what Rose-Ackerman calls bribes (payments to voters with a quid pro quo) and tips (payments with no formal payback) is, in fact, the objective of a handful of lawyers (Karlan 1994, Hasen
2000) and two social scientists (Kochin & Kochin 1998), who concede that, if
politics is about the exchange of support for goods, then it is not easy to argue that vote trades are morally wrong. According to Schaffer, the failure of
middle- and upper-class reformers to understand these distinctions dooms their
efforts to create a political consensus to uproot corruption from Filipino political
life.
This overview suggests that electoral fraud is commonplace in many political
systems. Furthermore, there are many clandestine ways to increase vote totals,
ranging from procedural violations to outright coercion on election day. The study
of Costa Rica indicates that parties used some combination of at least 47 different
types of fraud (Lehoucq & Molina 2002). Though there may not be an infinite
number of ways to rig electoral results, the techniques for manipulating the vote
are varied and artful. This review also points out that blatant types of fraud
including the use of intimidation or the stealing of ballot boxesdo not comprise
a majority of fraudulent acts. Most efforts to fabricate votes are mundane; they

12 Apr 2003

18:35

246

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

involve surreptitious efforts to increase vote totals. Finally, this survey suggests
that electoral fraud incenses parties and citizens. There is a normative dimension
to electoral fraud that makes it more than simple violations of electoral law; in
many places, the accusation of fraud has become part of the discourse of democratic
movements, an accusation that refers to the desecration of civil liberties and liberal
democratic principles.

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

THE MAGNITUDE AND CONSEQUENCES

OF ELECTORAL FRAUD
How common is electoral fraud? How important is it? Although much of the
research on unreformed electoral systems paints a rough portrait of ballot rigging,
only a few of these studies try to assess its weight in electoral competition. This is
an important issue because certifiers of electoral returns often dismiss allegations
of fraud by claiming that legal violations do not affect results.
Both Schedler (2002b) and Schaffer (2001) cite surveys that ask citizens whether
they were offered benefits for their votes. Schedler draws on work by Cornelius
(2002), whose post-election survey of a panel of Mexicans finds that 26.1%
of respondents received gifts from parties or candidates in the 2000 elections.
Cornelius uses several statistical models to show that PRI operatives were much
less successful than either of the main opposition parties in nailing down vote
trades, perhaps because electoral safeguards allowed voters to take advantage of
the PRIs largesse and vote their consciences. According to an unpublished report
(cited in Schaffer 2001), 10% of Filipinos say a party offered them a gift before
the May 2001 election day. As in contemporary Mexico, vote buying showed that
it was not often effective: Only 38% of the Filipino poor mentioned voting for the
candidate that offered them money.
As all these authors recognize, however, these surveys do not permit inferring whether vote buying occurs. Vote buying involves an explicit quid pro quo
for votes, an exchange that only seems to proliferate when parties can ensure
that voters or their agents cast ballots in prearranged ways (Lehoucq 2002b).
Depending on how the question is phrased and how voters are approached (via
telephones, face-to-face on the street, in their homes, etc.), surveys can generate
contradictory findings. Three surveys of Mexican voters in 2000 found that some
respondentsestimates ranged from 4.7% to 26.1%felt pressure or got something of value for their votes (cited in Cornelius 2002, p. 17). Fine-tuning questions and improving sampling techniques will go a long way toward making surveys a highly useful means of understanding the nature and dynamics of electoral
fraud.
Domnguez and McCanns study of Mexican public opinion is perhaps the most
ambitious attempt to use surveys to determine the decisiveness of electoral fraud
(Domnguez & McCann 1996, McCann & Domnguez 1998). They demonstrate
that, contrary to conventional wisdom, the opposition probably did not win the

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD

P1: IKH

247

1988 general election, one that resulted in Carlos Salinass controversial election
to the presidency. By taking into account the preferences of nonvoterswho disproportionately favored opposition candidatesthey suggest that Salinas would
have won anyway, but perhaps by less than an absolute majority of the popular vote
(Domnguez & McCann 1996, pp. 16264). The PRIs nervousness on election
night, when the central computer system mysteriously shut downits concern
that it would simply outpoll its rivals and perhaps lose the Federal District (where
1 out of 5 voters resides)led to shenanigans that robbed the election result of
legitimacy and ushered in a decade of electoral reforms. In a review of public
opinion polls, Schedler (1999) shows that the share of respondents believing that
fraud would mar elections went from over 50% in the 1988 elections to less than
30% in the 1997 midterm elections.
Another way to gauge the magnitude of fraud is to correct for the impact of ballot rigging on vote totals. The most creative and sensitive effort to assess the impact
of fraud is Powells (1989) statistical analysis of the 1868 Mississippi ratification
election. In June 1868, 53.4% of the voters rejected the Republican-inspired constitution, a charter that threatened to bar former white (and Democratic) officials
from ever holding office again. Powell uses regression analysis of the 61 counties
to identify outliersplaces with large shares of black registered voters who rejected the proposed constitution. By correcting results in these counties based on
vote totals and approval rates in normal counties, Powell plausibly argues that the
vote should have been 52.5% in favor and 47.5% against the constitution. Because
the outlier counties tend to be the places with allegations of fraud, Powell (1989,
p. 651) suggests that the swing vote was the stolen vote. King (2001) also uses
ecological analysis of 32 counties to argue that South Carolinian Democrats stole
the gubernatorial election from the Republican candidate, Daniel Chamberlain, in
the 1876 elections.
Other statistical analyses of voting returns include Baums (1991) analysis of
the 1861 Texas secession referendum and Oberst & Weilages (1990) study of the
1982 Sri Lankan referendum. Baum suggests that apathy, intimidation, or possible
counting out may have caused underrepresentation of voters wishing to remain
part of the United States in three counties of Texas. His paper implies, however,
that fraud was not responsible for the 3 to 1 vote in favor of joining the U.S.
Confederacy. Oberst & Weilage use simple tests of significance to demonstrate
that the difference between the yes and no votes was far greater in the 16 (of 133)
districts where there were allegations of fraud than in the districts with no such
allegations. Although they do not claim that the government stole the referendum to
ensure that parliamentarians could remain in office until 1989 (and thus inaugurate
a semipresidential system), they show how t-tests can identify areas ripe for more
intensive study.
These academic studies have their political counterparts. Hotly contested elections in Africa, the Americas, Asia, and Eastern Europe often involve NGOs
that organize a parallel count of the vote to generate an independent basis for
evaluating the performance of authoritarian systems forced to hold elections. In

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

248

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

1994, Civic Alliance of Mexico fielded more than 10,000 observers, covering
1 out of 10 polling stations with their sample. Though their parallel vote count
was within 1% of official returns, Civic Alliance documented numerous infractions of electoral law. According to their observers (Calderon-Alzati & Cazes
1996, p. 179), for example, 39% of Mexican polling stations violated the secrecy
of the ballot, whether because some voters showed their ballots (18.58 of these
polling stations) or someone saw for whom they voted (16.53 of the total). What
is unclear from these reports is whether these and other irregularities constitute
grounds to invalidate the results of the 1994 Mexican presidential or legislative
elections.
Using the legal challenges parties file regarding election outcomes is another
way to assess the impact of fraud on election outcomes. Gomez-Tagle (1988,
pp. 2124) reveals that the share of impugned votes in Mexicos single-member
legislative districts went from 46% in the 1979 elections to 62% in the 1985
elections. Renzato (1991, p. 126) points out that parties, on average, complained
about results in five times more districts in Spain than in Italy between 1904 and
1914. On average, parties complained about results in more than a third of Spains
mostly single-member districts; in 1914, they filed complaints in nearly half of
them. Similarly, Tusell (1970, p. 618) notes that this ratio remained steady until
1923 in Spain, even with the establishment of an Electoral Tribunal in 1907 to
judge complaints about election outcomes (even as the Cortes remained the body
that ultimately certified election results). Lehoucq & Molina (2002, p. 242) show
that parties presented an average of nearly 6 petitions to nullify electoral results in
each of the 20 elections held in Costa Rica between 1901 and 1948. In a different
vein, Eisenstadt (2002) shows that the rule-of-law ratiothe ratio of street protests,
or what he calls postelectoral conflicts, to electoral court cases filed by opposition
partiesbegan to approach zero by the second half of the 1990s among Mexican
states, suggesting that opposition parties began to trust autonomous electoral courts
to resolve their disputes about the results from the ballot box.
Identifying the number of impugned races must be complemented with an
analysis of the charges raised in the petitions to gauge the effects of this electoral
litigation on the magnitude of fraud. On the basis of such an analysis, Lehoucq
& Molina (2002, p. 22) conclude that three Costa Rican presidentsout of the
12 elected at four-year intervals between 1901 and 1948owed their election to
fraudulent practices. Even with very permissive assumptions about the extent of
fraud, the proportion of impugned votes does not typically exceed the percentage
of votes separating winners from losers. Yet, 20% of 12 presidential elections is
not an insignificant share over a 50-year period. Though the constant accusations
of fraud that accompany elections in unreformed systems exaggerate the impact of
fraud, the fact that ballot rigging could be decisive suggests that its denunciation
is justified.
Knowledge about the magnitude of fraud suggests that it is not always efficacious. The handful of surveys suggests that parties offer an array of promises, gifts,
and even cash for votes, but that does not mean that vote buying always works.

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD

P1: IKH

249

Analysis of election results raises the possibility that outcomes can be fixedthat,
given past performance or the behavior of voters, recorded outcomes do not always express the unfettered will of the people. Legal petitions indicate that parties
will denounce an array of practices in the hopes of overturning unfair (or even
fair) election returns. All three types of evidence suggest that complaints are not
infrequently exaggerated; the 1988 Mexican survey and the legal data on Costa
Rica suggest that fraud is usually not the cause of an electoral defeat. Nevertheless,
ballot rigging can reduce turnout, contribute to cynicism, and therefore fabricate
enough votes to tip the scales in one partys favor. Because electoral fraud can
be decisive in close races, its ultimate cost may be that it undermines democratic
stability. Regardless of whether fraud is decisive, it encourages incumbents and
opponents to discredit elections and their outcomes. Precisely because it is so hard
to determine the efficacy of ballot rigging, electoral fraud and its denunciation
corrode the democratic body politic.

THE CAUSES OF BALLOT RIGGING

If stuffing the ballot box can be decisive and acts of electoral fraud can be classified and counted, how can fraud be explained? Why do parties and machines rig
ballot results? Though few studies explicitly analyze motives, the underlying argument seems to be that incumbents, parties, and machines will try to get away with
anything to retain or obtain control of the state. Less consensus exists about the interests these agents protect. Some analysts underscore the importance of economic
interestslandlords and industrialists come to mindwhile others emphasize the
role of partisan interests. In this section, I review the evidence about why parties fabricate fraud, which is even scarcer than the research on its magnitude and
consequences.
Of the historical studies about fraud, Andersons (2000) is the best exponent
of the argument that economic interests perpetrate electoral fraud. She uses the
German Imperial electoral audits to show how eastern Junkers, coal mine owners,
and factory owners policed their peasants and employees to ensure that conservative parties won elections. Graham (1990), Hoppen (1984), Kousser (1974), and
OGorman (1989) imply that landlords use fraud to ensure that the representatives
elected from their districts maintain their class power.
Darde (1996), Botana (1979), Tusell (1991), and Varela Ortega (2001a) suggest
that partisanship explains the incidence of fraud. Botana argues that incumbent manipulation of the electoral system in Argentina preserved the conservative orders
control of a political system until 1916, when the opposition Radical Civic Union
won the general elections. The 1916 defeat of the conservative candidate was the
result of the 1912 Saenz Pena electoral law, which, among other reforms, had
cleaned up the voter registry. Darde surveys much of the Spanish historiography
to identify economic and partisan motives of the cacique. Tusell (1991) and Varela
Ortega (2001a) adopt a political angle, that of the local machine operator; the

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

250

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

cacique served his own interests as well as those of the local grandees and of the
central government. Indeed, the cacique cut deals with landlords and government
officials to obtain particularistic benefits (pork), such as roads, bridges, and related construction projects, useful for holding his voters together and therefore
maintaining his power. Like other Spanish historians, they usefully distinguish between rural areas, where, because of some collusion between central government
officials, landlords, and caciques, there was no effective political competitionand
therefore no election-day fraudand competitive urban areas, where all manner
of electoral atrocities occurred (Varela Ortega 2001b).
Several studies of electoral fraud shed light on the relative importance of sociological and political factors. Using survey data, Domnguez & McCann (1996,
pp. 16470) argue that accusations were more common in competitive, typically
urban, districts. Exit polls recorded fewer denunciations of voting irregularities
as the share of the vote that incumbent PRI gubernatorial candidates obtained
in districts increased (Domnguez & McCann 1996). Eisenstadt (1998) demonstrates that infractions of electoral law picked up in the 1980s and 1990s, precisely
when the PRIs decades-long domination of politics began to crumble (also, see
Gomez-Tagle 1988, Molinar-Horcasitas 1991).
Three studies explicitly vary social structural and institutional factors to identify the determinants of electoral fraud. First, Eisenstadts (1999) analysis of more
than 3000 municipal elections between 1989 and 1998 in Mexico reveals that
post-election conflicts increased where political competition was more intense
and where citizens had a tradition of engaging in collective action. He measures
competitiveness as the ratio of the votes the two principal opposition parties receive
to those of the PRI. Eisenstadt claims this variable performs better than the conventional indicator for competitiveness, the difference in votes between the firstand second-place winners (Cox 1988). Eisenstadt also uses an invaluable database
from the federal Attorney General for Agrarian Issues. The relationship between
collective agrarian conflicts per capita and post-election conflicts is positive.
Second, Trejo & Aguilar-Rivera (2002) show that election conflict in 803 municipalities (or approximately a third of all municipalitiesthose whose population
is at least 30% indigenous) in Mexico during the 1997 federal elections is largely a
function of sociological factors. They use ordinal probit models to show that procedural violations of electoral law and reports of intimidation in the organization of
elections are negatively related to geographic dispersion. The models also demonstrate that voting irregularities are positively related to the monolingual share of
the population, the effective number of ethnic groups, gender discrimination, and
the effective number of religions. Their study, which also includes 10 focus groups,
finds that an institutional factor also leads to election conflict. The use of customary law in local elections, a common practice in the State of Oaxaca, augments
difficulties for voters using different procedures and rules in federal elections.
Third, in a study of 1300 accusations of ballot rigging between 1901 and 1946
in Costa Rica, Lehoucq & Molina (2002) discover that institutional as well as
sociological factors explain the rhythm and spatial distribution of accusations of

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD

P1: IKH

251

fraud. Slightly less than half of the accusations of fraud took place in the three
poorest and least populated provinces, which never contained more than 25% of
the electorate. Social differentiation was much more pronounced in the periphery, where laborers were often completely landless and worked on large cattle or
banana estates. In contrast, citizens of the four central provinceswhere most
voters residedtended to live in tightly knit, highly literate communities capable
of organizing themselves to protect their civil liberties. Nevertheless, Lehoucq &
Molina (2002) argue that the structure of electoral competition was equallyif
not moreresponsible for regional distributions of electoral fraud. Unlike most
races in the center, most elections in the periphery were single- or double-member
districts whose seats went to the parties that only needed to obtain one more vote
than their rivals. In the center, in contrast, proportional representation typically
allocated seats among parties because most races involved parties competing for
three or more legislative seats at a time. Unless the quotientthe number of seats
divided by the number of valid voteswas small and/or the number of impugned
votes was large, parties stood little chance of ever electing one of their members
to Congress. Parties, in other words, faced incentives to commit and, most importantly, to denounce acts of fraud precisely where voters were disproportionately
poorer, less literate, and more ethnically discriminated against.
The paucity of findings makes it harder to generalize about the causes of fraud
than about its nature or magnitude. Though the study of Spain and one study of
Mexico (i.e., Domnguez & McCann 1996) suggest that fraud is more common
in urban than in rural areas, this conclusion has to be qualified. Ballot rigging
may be less common in rural areas because political or economic domination
makes it pointless to complain about fixed results. The urban bias in reports, in
other words, seems to be a product of an incumbent monopoly. Civic Alliances
finding (Calderon Altzati & Cazes 1996, Cazes 1996) that violations of the secret
ballot are more common in rural polling stations suggests that local machines use
subtle techniques to control voters, tactics that electoral observation picks up in an
increasingly competitive electoral environment. Existing studies also indicate that
properties of electoral lawsdistrict size and electoral formulacan increase the
uncertainty of political competition, the factor through which social and political
variables shape the nature and frequency of ballot rigging.

CONCLUSION: DIRECTIONS FOR FUTURE RESEARCH

The research on electoral fraud tells a consistent story. First, fraud takes on a
panoply of forms. It ranges from procedural violations of electoral law (that may
or may not intend to distort results) to the outright use of violence to intimidate
voters and poll watchers. Moreover, only a minority of the accusations involves
blatant types of fraud. Second, ballot rigging does not appear to be decisive most
of the time. The colorful history of vote fabrication probably exaggerates its role in
determining election outcomes. Fraud, nevertheless, undermines political stability

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

252

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

because, in close races, it can be decisive. Even when elections are not competitive,
ballot rigging robs elections of credibility and therefore prevents the consolidation
of democratic institutions. Third, political competition shapes the rhythm and
nature of electoral fraud. Efforts to steal elections increase with social inequality,
but political competitivenesswhich institutions help to shapedetermines the
ballot rigging strategies parties adopt.
Studying electoral fraud is not easy because it is supposed to remain clandestine. Yet, analyzing ballot rigging is far from impossible. There is a wealth of
legal material about the nature and incidence of ballot rigging. In virtually all
places, parties filed complaints about election returns with legislatures until the
late nineteenth century. By the beginning of the twentieth century, electoral tribunals or commissions began to run elections and to judge the complaints parties
filed about them (Lehoucq 2002a). Whatever the system a country uses (or used),
candidates and parties therefore generate the detailed legal trail that researchers,
to date, rarely exploit. Indeed, most of the historical work on elections refers to
such materials but almost never tries to classify, count, and analyze them. True,
these petitions cut both ways: They exaggerate as well as denounce acts of fraud.
Carefully used, though, accusations of fraud provide a wealth of information about
ballot rigginghow it worked, how extensive it wasand help shed light on what
it means to its targets and its plaintiffs.
Surveys can also uncover the facts and the causes of electoral fraud. To date,
only a handful of surveys cover ballot rigging. Questions need to focus on activities
deemed fraudulent and to encourage respondents to report how they are cajoled,
pressured, or bought. Focus groups or even more ethnographic research techniques
can also reveal how citizens react to clandestine efforts to shape election outcomes
and even participate in them.
Future studies of electoral fraud need to be theoretically focused. Researchers
tend to select countries and time periods for idiographic reasons or on the dependent
variable. The study of country X for country Xs sake, however, can lead researchers
to neglect issues that are causally important. For example, Dardes (1996) effort to
come to grips with a history of controlled rural elections or fraudulent urban ones
in Spain pushes him to focus on cultural limitations. That the existing research
emphasizes the impact of sociological and institutional factorsand that, to date,
there is no systematic study of ballot rigging for Spainsuggests that students of
this case may not be choosing the best approach to make sense of electoral fraud in
Spain. Similarly, picking cases on the dependent variable can prejudice efforts to
distinguish actual from spurious causes. A study limited to places with charges of
fraud runs the risk of producing conclusions that do not hold true for districts with
no such reports. Comparison between countries or of districts within a country is a
research strategy useful to overcome the selection bias problem (King et al. 1994).
I end this review with several hypotheses worth investigating. The first is
whether majoritarian electoral laws encourage higher levels of ballot rigging and
election litigation than more proportional systems. If seats are allocated on a firstpast-the-post basis, parties may commitand denounceballot rigging because a

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD

P1: IKH

253

relatively small number of votes may determine an election outcome. To the extent
that majoritarian systems encourage personal as opposed to party-centered campaigns, first-past-the-post systems may generate additional incentives for candidates, for example, to buy votes and for voters to sell them (Hicken 2002). The cases
for which we know the mostEngland, Germany, Ireland, Spain, and the United
Statesuphold this claim; all of these cases used plurality to allocate seats among
candidates. The study of Costa Rica adds further support to this hypothesis: Plurality districts generate more accusations of fraud than proportional-representation
ones. Majoritarian formulas such as the Thai multiple-seat, multiple-vote plurality
system or the Taiwanese single-nontransferable vote (Grofman et al. 1999) also
appear to be linked with high levels of vote buying in both countries.
Electoral systems that dampen turnout may simultaneously encourage fraud.
Many electoral authoritarian regimes (Schedler 2002a)systems where incumbents hold elections but manipulate laws for partisan advantagehinder voter
registration or prevent opposition parties from observing the electoral process or
the tally of the vote in order to reduce turnout. If rates of voter participation fall,
then fabricating a handful of votes may be sufficient to retain power, a fact that
opposition or regional parties may exploit as elections become more competitive.
This claim, I should add, dovetails with the previous hypothesis. As Cox (1999)
points out, proportional systems increase turnout more than majoritarian laws because the expected utility of a vote cast for a third party is greater in the former
than in the latter systems. Thus, plurality electoral laws may decrease turnout so
that the impact of stealing votes becomes greater. Finally, both hypotheses are part
of the broader theory that political competition fuels ballot rigging. Indeed, only
the existence of anticorruption legislation and nonpartisan electoral commissions
can thwart competition from propagating all manner of strategies for winning
elections.
The Annual Review of Political Science is online at
http://polisci.annualreviews.org

LITERATURE CITED
Anderson ML. 2000. Practicing Democracy:
Elections and Political Culture in Imperial
Germany. Princeton, NJ: Princeton Univ.
Press. 483 pp.
Annino A, ed. 1995. Historia de las elecciones
en Iberoamerica, Siglo XIX. Buenos Aires:
FCE. 481pp.
Aparicio R. 2002. La magnitud de la manipulacion del voto en las elecciones federales
del ano 2000. Perfiles Latinoam. 20:79
100
Argersinger P. 1992. Structure, Process and

Party: Essays in American Political History.

Armonk, NY: Sharpe. 219 pp.
Aziz-Nassif A. 1987. Chihuahua y los lmites
de la democracia electoral. Revista Mexicana
de Sociologa 49:159226
Basadre J. 1980. Elecciones y centralismo en
el Peru: Apuntes para un esquema historico.
Lima: Univ. del Pacifico. 173 pp.
Baum D. 1991. Pinpointing apparent fraud in
the 1861 Texas secession referendum. J. Int.
Hist. 22:20121
Botana N. 1979. El orden conservador: la

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

254

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

poltica Argentina entre 18801916. Buenos

Aires: Ed. Sudamericana. 345 pp.
Calderon-Alzati E, Cazes D. 1996. Las elecciones presidenciales de 1994. Mexico, DF:
La Jornada/CIICH, UNAM. 257 pp.
Callahan W. 2000. Pollwatching, Elections and
Civil Society in Southeast Asia. Burlington,
VT: Ashgate. 218 pp.
Callahan W, McCargo D. 1996. Vote-buying in
Thailands northeast. Asian Surv. 26:37692
Carter J. 1992. Turning Point: a Candidate, a
State, and a Nation Come of Age. New York:
Times Books. 273 pp.
Cazes D, and 843 coauthors. 1996. Memorial
de las elecciones de 1994: testimonios de observadores. Mexico, DF: La Jornada/CIICH,
UNAM. 422 pp.
Charnay JP. 1964. Les scrutins politiques en
France de 1815 a` 1962: contestations et invalidations. Paris: Cah. De la Fond. Nat. Des
Sci. Polit. 281 pp.
Cornelius W. 2002. La eficacia de la compra
y coaccion del voto en las elecciones mexicanas de 2000. Perfiles Latinoam. 20:11
32
Cox GW. 1988. Closeness and turnout: a
methodological note. J. Polit. 50:76875
Cox GW. 1999. Electoral rules and the calculus
of mobilization. Leg. Stud. Q. 24:387419
Cox GW, Kousser JM. 1981. Turnout and rural
corruption: New York as a test case. Am. J.
Polit. Sci. 25:64663
Darde C. 1996. Fraud and passivity of the electorate in Spain, 18751923. In Elections Before Democracy, ed. E Posada-Carbo, pp.
20123. New York: St. Martins
Domnguez JI, McCann JA. 1996. Democratizing Mexico: Public Opinion and Electoral Choices. Baltimore, MD: Johns Hopkins Univ. Press. 269 pp.
Eisenstadt TA. 1998. Courting democracy in
Mexico: party strategies, electoral institution-building, and political opening. PhD
thesis. Univ. Calif., San Diego. 379 pp.
Eisenstadt TA. 1999. Weak electoral institutions or legacies of social conflict? Modeling causes of Mexicos local post-electoral
mobilizations, 19891998. Presented at

Annu. Meet. Am. Polit. Sci. Assoc., 95th,

Atlanta, GA
Eisenstadt TA. 2002. Measuring electoral court
failure in Mexico. Int. Polit. Sci. Rev. 23:47
68
Fairbairn B. 1990. Authority vs. democracy:
Prussian officials in the German elections of
1898 and 1903. Hist. J. 33:81138
Fairbairn B. 1997. Democracy in the Undemocratic State: the German Reichstag Elections
of 1898 and 1903. Toronto: Univ. Toronto
Press. 328 pp.
Ferrero RA. 1983. Los fraudes electorales. Todo
es Hist. 197:4864
Garber M, Frank A. 1990. Contested Elections
and Recounts. 2 vols. Washington, DC: Gov.
Print. Off.
Gist G. 1961. Progressive reform in a rural community: the Adams County vote-fraud case.
Miss. Valley Hist. Rev. 48:6078
Gomez-Tagle S. 1988. Conflictos y contradicciones en el sistema electoral mexicano. Estud. Sociol. 6:338
Graham R. 1990. Politics and Patronage in
Nineteenth Century Brazil. Stanford, CA:
Stanford Univ. Press 382 pp.
Grofman B, Lee SC, Winckler EA, Woodall B,
eds. 1999. Elections in Japan, Korea, and
Taiwan under the Single Non-Transferable
Vote: the Comparative Study of an Embedded
Institution. Ann Arbor: Univ. Mich. Press.
390 pp.
Gueniffey P. 1993. Le nombre et la raison: la
Revolution francaise et les e lections. Paris:

Ecole
des Hautes Etudes
en Ciencias Sociales. 559 pp.
Hanham HJ. 1959. Elections and Party Management: Politics in the Time of Disraeli and
Gladstone. London: Longmanns. 468 pp.
Hasen RL. 2000. Vote buying. Calif. Law Rev.
88:132371
Hicken AD. 2002. Parties, pork and policy: policymaking in developing democracies. PhD
thesis. Univ. Calif., San Diego
Hoppen KT. 1984. Elections, Politics and Society in Ireland, 18321885. Oxford, UK:
Clarendon. 569 pp.
Karlan P. 1994. Not by money but by virtue

12 Apr 2003

18:35

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ELECTORAL FRAUD
won? Vote trafficking and the voting rights
system. Va. Law Rev. 80:145575
King RF. 2001. Counting the votes: South Carolinas stolen election of 1876. J. Interdiscip.
Hist. 32:16991
King G, Keohane RO, Verba S. 1994. Designing
Social Inquiry: Scientific Inference in Qualitative Research. Princeton, NJ: Princeton
Univ. Press
Kochin MS, Kochin LA. 1998. When is buying
votes wrong? Public Choice 97:64562
Kousser JM. 1974. The Shaping of Southern
Politics: Suffrage Restriction and the Establishment of the One-Party South, 1880
1910. New Haven, CT: Yale Univ. Press.
319 pp.
Lehoucq FE. 2002a. Can parties police themselves? Electoral governance and democratization. Int. Polit. Sci. Rev. 23:2946
Lehoucq FE. 2002b. When do parties buy
votes? Theoretical and empirical perspectives on electoral corruption. Presented at
Annu. Meet. Am. Polit. Sci. Assoc., 98th,
Boston, MA, Aug. 28Sep. 1
Lehoucq FE, Molina I. 2002. Stuffing the Ballot
Box: Fraud, Electoral Reform, and Democratization in Costa Rica. New York: Cambridge Univ. Press. 277 pp.
Malamud C, ed. 2000. Legitimidad, representacion y alternancia en Espana y America
Latina: las reformas electorales (1880
1930). Mexico, DF: FCE. 318 pp.
McCann JA, Domnguez JI. 1998. Mexicans react to electoral fraud and political corruption:
an assessment of public opinion voting behavior. Elec. Stud. 17:483503
McDonald RH. 1967. Electoral fraud and
regime controls in Latin America. West. Polit.
Q. 20:694708
Middlebrook KJ, ed. 1998. Electoral Observation and Democratic Transitions in Latin
America. La Jolla, CA: Cent. US-Mexican
Stud., Univ. Calif., San Diego
Molina I, Lehoucq FE. 1999. Political competition and electoral fraud: a Latin American case study. J. Interdiscip. Hist. 30:199
234
Molinar-Horcasitas J. 1991. El tiempo de

P1: IKH

255

la legitimidad: elecciones, autoritarismo y

democracia en Mexico. Mexico, DF: Cal y
Arena. 265 pp.
Mucke U. 2001. Elections and political participation in nineteenth century Peru: the
1871/1872 presidential campaign. J. Latin
Am. Stud. 33:31146
National Commission on Federal Electoral Reform. 2002. To Assure Pride and Confidence
in the Electoral Process. Washington, DC:
Brookings Inst. 385 pp.
Oberst RC, Weilage A. 1990. Quantitative tests
of electoral fraud: the 1982 Sri Lankan referendum. Corrupt. Ref. 5:4962
OGorman F. 1989. Voters, Patrons and Parties: The Unreformed Electoral System of
Hanaverian England, 17341832. Oxford,
UK: Oxford Univ. Press. 445 pp.
OLeary C. 1962. The Elimination of Corrupt
Practices in British Elections, 18681911.
Oxford, UK: Clarendon. 253 pp.
Pastor R. 1999. The role of electoral administration in democratic transitions: implications for policy and research. Democratization 6:127
Perman M. 2001. Struggle for Mastery:
Disfranchisement of the South, 18881908.
Chapel Hill: Univ. North Carol. Press. 416 pp.
Posada-Carbo E. 1994. Elections and civil wars
in nineteenth-century Colombia: the 1875
presidential campaign. J. Latin Am. Stud.
26:62149
Posada-Carbo E, ed. 1996. Elections Before
Democracy. New York: St. Martins
Posada-Carbo E. 1997. Limits to power: elections under the conservative hegemony in
Colombia, 18861930. Hispan. Am. Hist.
Rev. 77:24579
Posada-Carbo E. 2000. Electoral juggling: a
comparative history of the corruption of the
suffrage in Latin America, 18301930. J.
Latin Am. Stud. 32:61144
Powell LN. 1989. Correcting for fraud: a quantitative reassessment of the Mississippi ratification election of 1868. J. South. Hist. 5:633
58
Renzato G. 1991. La forja de la soberana
nacional: las elecciones en los sistemas

12 Apr 2003

18:35

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

256

AR

AR186-PO06-12.tex

AR186-PO06-12.sgm

LaTeX2e(2002/01/18)

P1: IKH

LEHOUCQ

liberales italiano y espanol. Ayer 3:115

38
Rigger S. 2002. Weighing a shadow: toward a
technique for estimating the effects of votebuying in Taiwan. Presented at Trading Political Rights: The Comparative Politics of
Vote-Buying, Mass. Inst. Technol., Cambridge, MA
Riker WH. 1982. Liberalism Against Populism:
The Theory of Social Choice Confronts the
Theory of Populism. San Francisco: Freeman. 311 pp.
Rose-Ackerman S. 1999. Corruption and Government: Causes, Consequences, and Reform. New York: Cambridge Univ. Press. 266
pp.
Sabato H, ed. 1999. Ciudadana poltica y
formacion de las naciones: perspectivas
historicas de America Latina. Mexico, DF:
FCE. 449 pp.
Sabato H. 2001. The Many and the Few: Political Participation in Republican Buenos
Aires. Stanford, CA: Stanford Univ. Press
Sambruccetti SIR. 1980. El fraude electoral en
1886. Bol. Inst. Hist. Argent. y Am. Dr.
Emilio Ravgnon 16:41582
Schedler A. 1999. Percepciones publicas de
fraude electoral en Mexico. Perfiles Latinoam. 14:10327
Schedler A. 2002a. The menu of manipulation.
J. Democracy 13:3650
Schedler A. 2002b. Citizens resent clientelism:
on the moral economy of vote trading in
Mexico. Presented at Annu. Meet. Am. Polit.
Sci. Assoc., 98th, Boston, MA
Schaffer F. 1998. Democracy in Translation:
Understanding Politics in an Unfamiliar Culture. Ithaca, NY: Cornell Univ. Press. 168 pp.

Schaffer F. 2001. Clean elections and the great

unwashed: electoral reform and the class
divide in the Philippines. Presented at Annu.
Meet. Am. Pol. Sci. Assoc., 97th, San Francisco, CA
Seymour C. 1915. Electoral Reform in England
and Wales: The Development and Operation
of the Parliamentary Franchise, 18321885.
New Haven, CT: Yale Univ. Press. 564 pp.
Sloan JW. 1970. Electoral fraud and social
change: the Guatemalan example. Sci. Soc.
34:7891
Tjarks GOE. 1963. Las elecciones saltenas de
1876 (un estudio de fraude electoral). Anuario de Historia 1:41775
Trejo G, Aguilar-Rivera JA. 2002. Etnicidad y
consolidacion democratica: la organizacion
de las elecciones en las regiones indgenas
de Mexico. In Los Dilemas de la democracia en Mexico, ed. A Hemond, D Recondo,
pp. 195224. Mexico, DF: CEMCA and
IFE
Tusell J. 1970. Para la sociologa poltica de
la Espana contemporanea: el impacto de la
ley de 1907 en el comportamiento electoral.
Hispania 30:571631
Tusell J. 1991. El sufragio universal en Espana
(18911936): un balance historiografico.
Ayer 3:1362
Varela Ortega J. 2001a. Los amigos polticos:
partidos, elecciones y caciquismo en la
restauracion, 18751900. Madrid: Marcial
Pons/Junta de Castilla y Leon, Consejera de
Educacion y Cultura. 568 pp.
Varela-Ortega J, ed. 2001b. El poder de la influencia: geografa del caciquismo en Espana
(18751923). Madrid: Marcial Pons/CEPC.
776 pp.

P1: FDS

April 17, 2003

15:13

Annual Reviews

AR186-FM

Annual Review of Political Science

Volume 6, 2003

CONTENTS
RICHARD E. NEUSTADT: PUBLIC SERVANT AS SCHOLAR,
Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org
Access provided by Columbia University on 01/29/15. For personal use only.

Charles O. Jones

GOVERNMENT TERMINATION, Michael Laver

POLITICAL SCIENCE ON THE PERIPHERY: SWEDEN, Olof Ruin
FOLIE RE PUBLICAINE, Robert E. Goodin
THE COLD WAR AS HISTORY, Jonathan Haslam
ROBERT A. DAHLS PHILOSOPHY OF DEMOCRACY, EXHIBITED IN
HIS ESSAYS, Michael Bailey and David Braybrooke
BRINGING ROBERT A. DAHLS THEORY OF DEMOCRACY TO EUROPE,
Sergio Fabbrini

1
23
41
55
77
99
119

THE MEDIA AND DEMOCRACY: BEYOND MYTHS AND STEREOTYPES,

Doris Graber

THE SUPREME COURT IN AMERICAN POLITICS, Lawrence Baum

WHY THE UNITED STATES FOUGHT IN VIETNAM, Larry Berman and
Stephen R. Routh

What Is Third World Security? Raju G.C. Thomas

Electoral Fraud: Causes, Types, and Consequences, Fabrice Lehoucq
RAWLS AND HABERMAS ON PUBLIC REASON: HUMAN RIGHTS AND
GLOBAL JUSTICE, J. Donald Moon
DEMOCRATIC INDIVIDUALISM AND ITS CRITICS, George Kateb
DELIBERATIVE DEMOCRATIC THEORY, Simone Chambers
THE POLITICAL ECONOMY OF HEALTH IN THE UNITED STATES,
John D. Wilkerson

ANCIENT EMPIRES, MODERN STATES, AND THE STUDY OF

GOVERNMENT, George E. Von der Muhll
TRADE, FOREIGN INVESTMENT, AND SECURITY, Richard Rosecrance and
Peter Thompson

CURRENT CONTROVERSIES IN FEMINIST THEORY, Mary G. Dietz

HERBERT A. SIMON: POLITICAL SCIENTIST, Jonathan Bendor

139
161
181
205
233
257
275
307
327
345
377
399
433

vii

P1: FDS

April 17, 2003

viii

15:13

Annual Reviews

AR186-FM

CONTENTS

TERRORIST DECISION MAKING, Gordon H. McCormick

What Role Does Prejudice Play in Ethnic Conflict? Donald P. Green and
Rachel L. Seher

473
509

INDEXES
Subject Index
Cumulative Index of Contributing Authors, Volumes 26
Cumulative Index of Chapter Titles, Volumes 26

Annu. Rev. Polit. Sci. 2003.6:233-256. Downloaded from www.annualreviews.org

Access provided by Columbia University on 01/29/15. For personal use only.

ERRATA
An online log of corrections to Annual Review of Political Science
chapters (if any have yet been occasioned, 1998 to the present) may be
found at http://polisci.annualreviews.org/

533
553
555