15 - DEVALUING THE VOTE IN LATIN AMERICA (Nam)

DEVALUING THE VOTE IN

LATIN AMERICA

Richard Snyder and David Samuels

Richard Snyder is associate professor of political science at Brown University. He is the author of Politics after Neoliberalism (2001) and coauthor with Gerardo L. Munck of Passion, Craft and Method in Com-parative Politics (2006). David Samuels, Benjamin E. Lippincott Associate Professor of political science at the University of Minnesota, specializes in Latin American politics and the comparative study of political institutions, with particular emphasis on Brazilian politics, electoral systems, political parties, legislatures, and federalism. This essay originally appeared in the January 2001 issue of the Journal of Democracy.

Although the exact definition of democracy is vigorously disputed, wide agreement exists that free and fair elections are the cornerstone of any democratic system of government. An indispensable characteristic of free and fair elections is that each citizen’s vote counts equally. This notion of fairness is embodied in the well-known principle of “one person, one vote” that theorists such as Robert Dahl consider to be an essential ingredient of democracy.1

Democratic rule now prevails across Latin America, and most Latin American countries embrace the principle of “one person, one vote” through constitutional provisions that explicitly guarantee the equal-ity of each citizen’s vote. Nevertheless, in practice, these same countries often fall far short of achieving such equality. In the lower chambers of their national legislatures, many Latin American countries have ex-traordinarily high levels of malapportionment—a wide discrepancy between the share of legislative seats and the share of population held by electoral districts. A malapportioned chamber means that the votes of some citizens weigh far more heavily than those of others. Although a longstanding federalist tradition acknowledges the value of having a bicameral legislature with an upper chamber that represents territorial units equally, a situation that usually requires a significant degree of

Richard Snyder and David Samuels

169

malapportionment, there is a broad consensus that at least one chamber should weigh the votes of citizens equally. Thus there is no normative justification for malapportionment in the lower chamber.

Apart from these important normative issues, lower-chamber malap-portionment has troubling practical consequences for democracy in Latin America. High levels of lower-chamber malapportionment have con-tributed to a syndrome characterized by four interrelated elements: 1) a distinct rural and conservative bias in legislatures; 2) a tendency to-ward estrangement between the executive and legislative branches; 3) a growing capacity for subnational political actors to hold national gov-ernments hostage on important policy issues; and 4) the proliferation of subnational authoritarian enclaves. Moreover, in countries with high levels of lower-chamber malapportionment, a malapportioned upper chamber can greatly exacerbate these same problems. Malapportion-ment thus poses a serious, yet hitherto neglected, challenge to the qual-ity and fairness of democracy in many Latin American countries.

Free and Unfair Elections

The Table on the following page shows the degree of malapportion-ment in 19 Latin American countries in 1999.2 The entries indicate the percentage of seats in the legislature that are not apportioned according to the principle of “one person, one vote.” Thus a score of 0.00 would indicate a perfectly apportioned chamber in which no citizen’s vote weighs more than another’s. By contrast, a score of 1.00 would indicate a fully malapportioned chamber in which all of the legislative seats are allocated to a single electoral district with just one voter. And a score of 0.50 would mean that 50 percent of the seats are allocated to districts that would not receive those seats if there were no malapportionment.

The Table shows that the overall level of malapportionment for lower and upper chambers in Latin America is significantly higher than in the rest of the world. Indeed, Latin America has some of the world’s most malapportioned legislative chambers. For example, Argentina’s Senate has a score of 0.49, making it the world’s most malapportioned legisla-tive chamber. Ecuador’s lower chamber has a score of 0.20, making it the third most malapportioned lower chamber in the world. Although the Table shows that malapportionment is not a problem everywhere in Latin America, a number of prominent countries—Argentina, Bolivia, Brazil, Chile, Colombia, and Ecuador—do have extremely high levels of lower-chamber malapportionment. These countries together account for almost two-thirds of the population of Latin America, underlining how important a focus on malapportionment is for understanding the problems of democracy in the region.

Most legislatures have some malapportionment, especially in the upper chamber (where one exists).3 Moreover, countries typically expe-

170	Devaluing the Vote in Latin America
TABLE—MALAPPORTIONMENT IN LATIN AMERICA, 1999
COUNTRY	LOWER CHAMBER	UPPER CHAMBER
Argentina*	0.14	0.49
Belize	0.08	–
Bolivia	0.17	0.38
Brazil*	0.09	0.40
Chile	0.15	0.31
Colombia	0.13	0.00
Costa Rica	0.02	–
Dominican Republic	0.08	0.38
Ecuador	0.20	–
El Salvador	0.07	–
Guatemala	0.06	–
Honduras	0.04	–
Mexico*	0.06	0.23
Nicaragua	0.06	–
Panama	0.06	–
Paraguay	0.04	0.00
Peru	0.00	–
Uruguay	0.03	0.00
Venezuela*	0.07	0.33
Latin America Average	0.08	0.25
Average for Advanced Industrial Democracies	0.04	0.18
World Average Without Latin America	0.06	0.18
United States*	0.01	0.36

* Countries with federal systems

Note: A score of 0.00 would indicate perfect apportionment; a score of 1.00 would indicate perfect malapportionment.

Source: David Samuels and Richard Snyder, “The Value of a Vote: Malapportionment in Comparative Perspective,” British Journal of Political Science 31 (October 2001): 627–49.

rience rising levels of “natural” malapportionment over time because of demographic changes across electoral districts. Nevertheless, many coun-tries—especially the advanced industrial democracies of Western Eu-rope and North America—have achieved low levels of malapportionment in one or both legislative chambers. Low levels of malapportionment in these cases reflect the implementation of procedures for periodic reap-portionment. By contrast, the far higher levels of malapportionment characterizing both upper and lower chambers in many Latin American countries indicate that they have not implemented such corrective pro-cedures as part of their transitions to democracy.

A Formal Flaw

Given the return to formal democracy in Latin America over the last two decades, much of the debate about the shortcomings of democracy

Richard Snyder and David Samuels

171

in the region has focused on informal rules and practices. For example, Guillermo O’Donnell calls attention to “brown areas,” where the rule of law is extremely attenuated because of a loose fit between formal rules and actual practices. Similarly, Jonathan Fox argues that “semi-clientelist” politics (especially at the local level) are an important obstacle to full-fledged democracy in Latin America.4

These informal defects are clearly important. Yet the problems of contemporary Latin American democracies are by no means limited to the informal realm. Formal institutions can also diminish the quality of democracy. A focus on malapportionment highlights how formalized, detailed, and explicit electoral rules in many countries have resulted in large inequalities in the weighting of citizens’ votes. Consequently, the limitations of many contemporary Latin American democracies go well beyond the problem that “the games played ‘inside’ the democratic institutions are different from the ones dictated by their formal rules.”5 Where formal electoral rules lead to unfair electoral results, even games that do abide by the rules may be undemocratic.

Malapportionment should be regarded as a formal institutional flaw that significantly weakens the quality of democracy in many countries in the region,6 yet it has attracted surprisingly little attention among scholars, politicians, or engaged citizens. In Chile, for example, the current method for electing members to the Chamber of Deputies has been the focus of much analysis and debate because it was designed by General Augusto Pinochet to favor the second-place finisher (typically, the conservative parties’ list), but the extreme malapportionment in the system, which is biased against the country’s metropolitan region, has attracted little notice. And in Brazil, where the issue of malapportion-ment has drawn some scholarly attention, a major recent report about proposed political reforms prepared with encouragement from the Cardoso administration contained “not one word about decreasing the malapportionment of the Senate.”7

This lack of attention highlights an important characteristic of mal-apportionment: It is often hidden from public view. In contrast to traditional “low-tech” methods for rigging elections, such as vote-buy-ing, ballot-stuffing, and techniques for tampering with the actual counting of votes, which are easier for election monitors and opposi-tion parties to detect, malapportionment has not been perceived as a proximate cause of unfair elections. Because the mechanisms for appor-tioning legislative seats are often arcane mathematical functions, they do not provide the kind of vivid material that easily lends itself to front-page news or to use as a lightning rod for mobilizing mass indignation. By contrast, lurid charges of ballot-stuffing and vote-buying are far more attractive issues for opposition parties, election losers, and inter-national election monitors. The virtual absence of political debate about malapportionment even in countries where it is most egregious, such as

172 Devaluing the Vote in Latin America

Argentina, Chile, Ecuador, Bolivia, and Brazil, underscores the hidden nature of this problem.

Although malapportionment has attracted little public attention, it can give some groups important strategic advantages in political com-petition. Politicians across the world have long recognized this, and they have honed well-known techniques for manipulating how elec-toral systems are apportioned, including gerrymandering and the pres-ervation of “rotten boroughs.” As is the case elsewhere, the use of such blatant methods of manipulating apportionment seems increasingly unlikely in contemporary Latin America. Yet this does not mean that apportionment decisions have been depoliticized, nor does it mean that apportionment procedures are fully democratic.

In contemporary Latin America, unlike in the advanced industrial democracies, high levels of malapportionment readily coexist with the core package of democratic rules and institutions—freedom of associa-tion, assembly, and speech; universal suffrage; absence of massive fraud; and contested elections.8 In a malapportioned system, all citizens can enjoy a free and equal opportunity to formulate and signify their prefer-ences, but they are denied the opportunity to have their preferences weighed equally.9 Consequently, even genuinely competitive elections held in systems with universal suffrage and where basic civil and politi-cal freedoms are effectively guaranteed may nevertheless be unfair. This apparent compatibility between malapportionment and the core ele-ments of democratic politics can help sustain an illusion of robust democracy in contexts where the votes of some citizens actually weigh far more than the votes of others.

The Malapportionment Syndrome

The high levels of lower-chamber malapportionment in many Latin American countries have a negative effect on the quality of democracy. Let us now consider in greater detail the four practical consequences of malapportionment noted at the outset of this essay:

1) Rural-conservative bias. Malapportionment has resulted in an overrepresentation of rural interests in both lower and upper chambers of national legislatures in a number of Latin American countries. This rural bias is partially an institutional legacy of efforts by conservative landed elites to undercut the electoral strength of emerging urban classes in the early twentieth century. The military regimes of the 1960s and 1970s, especially in Argentina, Brazil, and Chile, significantly rein-forced this rural bias, thereby guaranteeing that it would persist into the contemporary democratic period.10

Examples from several countries illustrate how malapportionment fosters a rural bias in legislatures. In Chile, the Santiago and Valparaíso metropolitan areas together encompass 50 percent of the voting popu-

Richard Snyder and David Samuels

173

lation, but they elect only 37 percent of the national Chamber of Depu-ties. Moreover, the population of Chile’s most populous electoral district (Maipu) is five times greater than that of its least populous district (Coihaique), yet both districts get two seats, making one vote in Coihaique worth the same as five votes in Maipu. In Brazil, the six densely populated southern states hold 55 percent of the population but elect only 48 percent of the deputies, and one vote in the country’s least populous state (Roraima) has 16 times the weight in electing a deputy of a vote cast in the most populous state (S~ao Paulo). This in-equality reaches even greater heights in Ecuador, where the two largest districts (Guayas and Pichincha) hold 45 percent of the population yet command only 23 percent of the seats. One vote in Ecuador’s least populous district (Galápagos) has 50 times the weight of a vote cast in the most populous district (Guayas).

This rural bias associated with malapportionment leads to distor-tions in the ideological balance of national legislatures. In Brazil in 1998, for example, leftist parties captured 26 percent of all votes for federal deputies, but because a large share of those votes was won in underrepresented states, these parties earned only 22 percent of the seats. And in Argentina, overrepresentation of rural provinces has long served to protect the Peronist party and its conservative allies in the lower chamber at the expense of urban interests, including the Peronist party’s own urban supporters.11

2) Estrangement between the legislative and executive branches.

Some scholars praise presidential systems for their ability to combine distinct kinds of representation—the legislature can be “representative of the diversity of the society and polity,” while the executive repre-sents the nation as a whole.12 In the context of highly malapportioned legislatures, however, such combinations can actually lead to paralyz-ing impasses between the two branches. Because the president’s electoral district is essentially the entire nation, presidential candidates have strong incentives to build their electoral coalitions and focus their cam-paigns in regions with the largest absolute number of voters. Consequently, presidents today seek and gain most of their support from urban constituencies, whereas legislators in malapportioned legis-latures are more likely to represent rural interests. This difference in their bases of support can contribute to an estrangement between the executive and legislative branches.

In Brazil, for example, no presidential candidate can hope to win without doing very well in the heavily urban states of S~ao Paulo, Rio de Janeiro, and Minas Gerais, which together have 43 percent of the elec-torate. Yet because of malapportionment, those three states control only 33 percent of the seats in the Chamber of Deputies. Thus a successful presidential candidate who concentrates his campaign in those states must inevitably cultivate new allies from less populous states in order

174 Devaluing the Vote in Latin America

to construct a viable governing coalition. This is precisely what current president Fernando Henrique Cardoso, who makes his home in S~ao Paulo, has done.

Yet presidents with urban bases of support who face a malappor-tioned legislature favoring rural regions are not always as successful as Cardoso has been in forging a governing coalition. The distinct geo-graphic support bases for the Brazilian legislative and executive branches led to conflicts that contributed to the collapse of democracy in 1964.13 A similar process led to the overthrow of President Velasco Ibarra in Ecuador in 1961.14 These examples suggest that a failure to manage the coalition-building challenges posed by malapportioned legislatures can undermine the stability of democratic regimes.

3. Holding the center hostage. When rural areas are overrepresented in the legislature, a president who wins election with an urban base of support may not be able to govern unless he can build alliances with rural groups. Thus he may face compelling pressures to channel patron-age payoffs to rural areas in order to purchase legislative support for his policy proposals.15 As a result, the president is potentially vulnerable to being held hostage by rural interests.

Moreover, in highly malapportioned systems, an increase in the share of the total population residing in urban areas can ironically strengthen the hand of overrepresented rural interests. This is because the amount of pork-barrel funds required to secure electoral support in rural areas should decrease as they acquire relatively smaller populations, thereby making them an attractive source of “cheap” support for coalition-builders at the national level. The pro-rural effects of existing malapportionment, com-bined with increasing migration from rural to urban areas, have greatly strengthened the influence of rural interests in many Latin American leg-islatures, even as those societies have become increasingly urbanized.

Edward Gibson and his colleagues have found statistical evidence showing that lower-chamber malapportionment in federal systems tends to distort government spending significantly in ways that favor sparsely populated regions.16 In countries such as Argentina and Brazil, they argue, the central government strategically directs pork-barrel spend-ing toward overrepresented rural states in order to shore up its majority in the congress. Thus malapportionment has strengthened the capacity of rural interests in these countries to extract fiscal favors from the central government.

4. Proliferation of subnational authoritarian enclaves. Malappor-tionment also has an important impact on subnational politics in Latin America. In new democracies, overrepresentation of rural districts can contribute to the maintenance—and even the proliferation—of non-democratic enclaves at the subnational level. Malapportionment can compel democratic elites at the national level to tolerate subnational authoritarian enclaves because these elites rely on overrepresented re-

Richard Snyder and David Samuels

175

gions to secure the national legislative majorities they need to achieve their policy goals. Ironically, the ability of leaders at the national level to implement and consolidate democratic reforms in a highly malappor-tioned system may therefore depend on winning the overvalued support of subnational authoritarian elites. At the same time, overrepresentation of subnational authoritarian enclaves in national legislatures may strengthen the ability of these subnational elites to fend off efforts by external groups seeking to reform local politics. Such a dynamic has been especially apparent in the case of Brazil, where, as Alfred Stepan notes, “Many of the states that are overrepresented in the federal legis-lature are precisely those states with particularly unequal income distribution and strong traditions of local oligarchic control.”17

Strategies for Reducing Malapportionment

Is there a cure for the malapportionment syndrome? Can Latin Ameri-can countries solve the problem of high malapportionment? We address these questions by considering the strengths and limitations of two strategies for reducing malapportionment that have been employed suc-cessfully by many countries around the world: 1) judicial oversight of reapportionment and 2) electoral-law reform.

1) Judicial oversight of reapportionment. Courts play a central role in the reapportionment of electoral systems in several countries, most notably the United States. Until the 1960s, the U.S. House of Represen-tatives had significant levels of malapportionment within states, but since a series of Supreme Court decisions in the 1960s mandating that the votes of all citizens should count equally and that electoral districts should, as far as possible, have equal populations, the House has had remarkably low levels of malapportionment.

These Supreme Court rulings, which began in 1962 with Baker v. Carr, gave federal and state courts the authority to declare a state’s redistricting plan null and void. This judicial review of districting plans dramatically altered the strategic situation facing politicians (state leg-islators and governors) involved in the reapportionment process. When the courts rejected a state’s districting plan, they gave the legislature and the governor a deadline, and if the legislature and governor failed to reach a satisfactory agreement by the deadline, the court imposed its own plan. The threat of losing control of the reapportionment process to the courts gave politicians a powerful incentive to agree to a districting plan that conformed closely to the “one person, one vote” principle. In Germany, the High Court has played a similar oversight role since the early 1960s that has also helped keep lower-chamber malapportion-ment at low levels.

What are the implications of the U.S. and German cases for Latin American countries? First, it bears emphasizing that the strategy of re-

176 Devaluing the Vote in Latin America

ducing malapportionment via judicial oversight presupposes that the courts (whether local, state, or national) are insulated from partisan in-fluence. Otherwise, the prospect of judicial review would not necessarily induce politicians to act to avoid the legally defined default outcome (a court-imposed districting plan). If the courts are exposed to political pressures and operate as reliable agents of partisan interests, then judi-cial oversight of the reapportionment process is unlikely to serve as an effective mechanism for reducing malapportionment. In such a scenario, court-imposed plans, like politically engineered plans, are likely to manipulate apportionment for partisan purposes. Moreover, the effec-tiveness of the judicial oversight strategy depends on the courts’ capacity to make a credible threat either to redraw district boundaries themselves or to invalidate an election.

In Latin America, neither judicial autonomy nor capacity can be taken for granted. In many countries in the region, courts are notorious both for their politicization and for their weakness.18 Consequently, most Latin American judiciaries probably lack the ability to define a credible legal default outcome that could induce politicians to imple-ment districting plans conforming to the “one person, one vote” principle.

Since the judicial oversight strategy for reducing malapportionment does not seem especially feasible in contemporary Latin America, a more promising alternative would be to create a neutral, nonpartisan electoral commission that is legally obligated to reduce lower-chamber malapportionment. This strategy for reducing malapportionment has been employed effectively in Mexico, where, since 1996, the Federal Electoral Institute (IFE) has exercised full authority over redistricting decisions and has helped compensate for the weakness of the Mexican judiciary. In July 1996, IFE’s top governing body, the General Council, unanimously approved a lower-chamber redistricting plan—designed by a committee of nonpartisan technical experts with impressive aca-demic and professional credentials—that significantly reduced malap-portionment in anticipation of the 1997 elections. Mexico’s political parties also played a key role in designing the plan: Their representa-tives were invited to comment on drafts of the plan, and some of their suggestions were incorporated into the final version, contributing to the plan’s undisputed acceptance.19

This approach to reducing malapportionment could work in other Latin American countries where the judiciary lacks the capacity to en-force the “one person, one vote” principle. In assessing the prospects for replicating the “IFE model” in other countries, it is important to highlight that IFE had its roots in the allegedly stolen presidential vic-tory of Institutional Revolutionary Party candidate Carlos Salinas in 1988. In exchange for recognizing Salinas’s legitimacy, the opposition National Action Party successfully demanded a package of electoral

Richard Snyder and David Samuels

177

reforms, including the formation of an independent electoral commis-sion.20 This suggests that effective independent, nonpartisan electoral commissions emerge out of partisan competition and may require an interparty pact.

2) Electoral-law reform. Reforming the electoral laws is probably a more viable strategy than judicial oversight for reducing malapportion-ment in Latin America. We consider here three distinct types of electoral law reform: 1) reapportionment; 2) a single-district chamber; and 3) a mixed electoral system with tiers.

One method for reducing malapportionment involves reallocating seats across districts so that underrepresented districts receive a propor-tionally larger share of the total number of seats and overrepresented districts receive a proportionally smaller share. In the absence of non-partisan judicial oversight, however, such reapportionment may be extremely difficult to achieve. Efforts to reallocate seats typically re-quire complex negotiations in order to avoid a “winner-take-all” situation where losers would face permanent disenfranchisement and would therefore have compelling incentives to undermine any proposed agreement. Consequently, attempts to implement reapportionment are unlikely to succeed unless they are incorporated into a broader package of electoral and institutional reforms. For example, Britain’s landmark reapportionment in 1885 was but one piece of the Third Reform Act, which included a major expansion of suffrage and resulted from a com-plex interparty pact. To our knowledge, reapportionment initiatives have not been linked to broader proposals for electoral and institu-tional reform in Latin America, but this could be a viable strategy for reducing malapportionment. Such reform proposals might even include provisions for creating the kind of nonpartisan electoral commission discussed above to manage the apportionment process in a neutral fash-ion. Because the problem of malapportionment in countries like Brazil seems so closely connected with other institutional flaws, such as a fragmented party system and overly decentralized federalism, correct-ing malapportionment could become part of a wider package of reforms that would significantly enhance the quality of democracy, economic efficiency, and socioeconomic justice.

A second method for reducing malapportionment in Latin America would be to abandon an electoral system with multiple districts by adopting a single nationwide district, as Israel and the Netherlands have done. In Latin America, only Peru has implemented a single-district format in its lower chamber. In 1993, President Alberto Fujimori abol-ished Peru’s Senate and created a unicameral legislature with a single nationwide district, thereby eliminating malapportionment altogether. Three Latin American countries (Colombia, Paraguay, and Uruguay) have recently eliminated malapportionment in their upper houses by adopting a single-district format.

178 Devaluing the Vote in Latin America

Although a single-district chamber eliminates malapportionment, this strategy has important disadvantages that outweigh the goal of perfect apportionment. Single-district chambers can weaken citizen control over their representatives by making it harder for voters to hold legislators responsible for their actions. The weakening of citizen control raises troubling issues about the accountability of elected officials in single-district chambers: Because citizens lose the sense that a particular legislator “belongs” to them, they may have weak incentives to monitor and punish the performance of individual legislators. Consequently, the ability of the “electoral connection” to serve as a mechanism of accountability may be severely attenuated in systems with single-dis-trict chambers. This would be especially unfortunate in countries where accountability is already a crucial problem.

Finally, several Latin American countries have recently adopted a “two-tiered” electoral system that combines elements of single-mem-ber-district (SMD) systems and proportional-representation (PR) systems. For example, Mexico now elects 300 deputies in a tier of SMDs and 200 additional deputies in a second tier of five 40-member PR districts. Venezuela and Bolivia have recently implemented similar systems, and a mixed system is often proposed as an alternative to Brazil’s open-list PR framework.

Countries do not typically adopt such systems with the goal of re-ducing malapportionment. Rather, the objective is to guarantee the rep-resentation of minority parties that would fail to win a plurality in any single district because their base of support is geographically dispersed. Nevertheless, a mixed system could serve to reduce malapportionment, depending on the number of seats added to the second tier and whether or not the second-tier seats are allocated to a nationwide district.

Mixed systems significantly reduce malapportionment only if the number of seats in the upper tier is substantial. For example, adding a tier with 20 seats elected in a nationwide district to a malapportioned legislature with 300 seats would have only a slight effect on the overall level of malapportionment. Moreover, a mixed system will reduce mal-apportionment only if the upper tier allocates seats exclusively to a nationwide district (as in El Salvador and Nicaragua). By contrast, tiers that distribute seats to subnational (provincial, state, or regional) dis-tricts may actually increase overall malapportionment (as in Bolivia and Venezuela).21

The strategy of allocating a large number of seats to a nationwide district has several advantages. First, it can be implemented in the con-text of a chamber with multiple, territorially defined electoral districts. Consequently, a far greater degree of citizen control over legislators is possible than in the case of a single-district chamber. Second, a tier of new seats can easily be added to the legislature without taking away any existing seats. This possibility reduces the likelihood of opposi-

Richard Snyder and David Samuels

179

tion from incumbent legislators concerned about protecting their dis-tricts. In short, because courts generally lack the capacity to supervise the reapportionment process in a nonpartisan fashion, the strategy of adding a tier may offer the best solution to the problem of malappor-tionment in Latin America.

Revaluing the Vote

Our critique of malapportionment in contemporary Latin America is anchored in the view that elections should, first and foremost, represent the will of individual citizens. Of course, alternative conceptions of representation exist—for example, that legislators should represent cor-porate or territorial units. Although few instances of corporate representation remain in Latin America,22 the constitutions of many Latin American countries do explicitly provide for representation of territorial units in the upper chamber. Such “territorial” chambers are intended to protect the interests of less-populated regions. A high de-gree of malapportionment may be necessary to achieve this objective, which is certainly a legitimate democratic goal.

Yet while malapportionment may be warranted in the upper cham-ber, there is no normative justification for unfairness in the lower chamber. The lower chamber should be based upon “one person, one vote,” with citizens represented as political equals. Because many Latin American countries have high levels of malapportionment in both their upper and lower chambers, they essentially have two territorial cham-bers, and none in which citizens’ votes count equally. As we have argued, this situation has had a negative effect on the quality of democracy in the region. Latin America’s democracies should revalue the vote by solving the problem of malapportionment in their lower chambers.

NOTES

We appreciate helpful comments on earlier drafts from Alberto Diaz Cayeros, John Gerring, Edward Gibson, Juan Linz, Scott Mainwaring, N. Guillermo Molinelli, Gerardo Munck, Guillermo O’Donnell, Diego Reynoso, Andreas Schedler, and Steven Solnick.

1. Dahl writes that the “unimpaired opportunit[y]” of all full citizens to “have their preferences weighed equally” is a necessary condition for democracy. Robert Dahl, Polyarchy: Participation and Opposition (New Haven: Yale University Press, 1971), 2. Dahl also includes “voting equality” as one of his “five criteria for a democratic process.” Robert Dahl, Democracy and Its Critics (New Haven: Yale University Press, 1989), 109–11.

2. The table on p. 170 above is based on David Samuels and Richard Snyder, “The Value of a Vote: Malapportionment in Comparative Perspective,” British Journal of Political Science 31 (October 2001): 651–71. To calculate malap-portionment, we used a modified version of the Loosemore-Hanby index of elec-toral disproportionality, taking the absolute value of the difference between each

180 Devaluing the Vote in Latin America

district’s seat and population shares, adding them, and then dividing by two. Wherever available, we used population rather than registered voters per district, as most countries apportion seats on the basis of population.

3. In fact, the only way to avoid malapportionment completely is to elect representatives in a single nationwide district, as a few countries do (such as Israel, Namibia, the Netherlands, Peru, and Sierra Leone).

4. Guillermo O’Donnell, “On the State, Democratization, and Some Conceptual Problems: A Latin American View with Glances at Some Post-Communist Coun-tries,” World Development 21 (August 1993): 1355–70; and Jonathan Fox, “The Difficult Transition from Clientelism to Citizenship: Lessons from Mexico,” World Politics 46 (January 1994): 151–84.

5. Guillermo O’Donnell, “Illusions About Consolidation,” Journal of Democ-racy 7 (April 1996): 34–52, 41.

6. In terms of comparative research on democracy, recognizing this formal flaw offers important advantages. Precisely because it is a formal component of elec-toral systems, malapportionment is amenable to systematic measurement and statistical analysis. By contrast, informal flaws, like “brown areas” and the (un)rule of law, while critical, often frustrate scholars because of the difficulty of gathering systematic data that go beyond fertile anecdotal evidence.

7. Alfred Stepan, “Brazil’s Decentralized Federalism: Bringing Government Closer to the Citizens?” Daedalus 129 (Spring 2000): 145–69, 163. Part of the explanation for this omission surely involves the fact that solving the problem of malapportionment would require a constitutional amendment, and such an amend-ment would require the unlikely support of a supermajority of senators and deputies from the overrepresented states. For an insightful analysis of malapportionment in Brazil, see Jairo Nicolau, “As Distorç~oes na Representaç~ao dos Estados na Câmara dos Deputados Brasileira,” Dados 40 (1997): 441–64.

8. By contrast, informal practices such as clientelism and particularism may be less compatible with these rules and institutions.

9. Here, we follow Dahl’s threefold distinction between the formulating, signi-fying, and weighing of citizens’ preferences as necessary conditions for democracy. Robert Dahl, Polyarchy, 1–3.

10. See Richard Snyder and David Samuels, “Legislative Malapportionment in Latin America: Comparative and Historical Perspectives,” in Edward L. Gibson, ed., Federalism and Democracy in Latin America (Baltimore: Johns Hopkins Univer-sity Press, 2004).

11. See Edward L. Gibson, “The Populist Road to Market Reform: Policy and Electoral Coalitions in Mexico and Argentina,” World Politics 49 (April 1997): 339–70; and Edward L. Gibson and Ernesto Calvo, “Federalism and Low-Mainte-nance Constituencies: Territorial Dimensions of Economic Reform in Argentina,”

Studies in Comparative International Development 35 (September 2000): 32–55.

12. Matthew Shugart and John Carey, Presidents and Assemblies (Cambridge: Cambridge University Press, 1992), 286.

13. David V. Fleischer, “Manipulaç~oes Casuísticas do Sistema Eleitoral Durante o Período Militar, ou Como Usualmente o Fetiço Voltava contra o Feitiçeiro,”

Cadernos de Ci^encia Política 10 (1994).

14. Peter Pyne, “The Politics of Instability in Ecuador: The Overthrow of the President, 1961,” Journal of Latin American Studies 7 (May 1975): 109–33.

Richard Snyder and David Samuels

181

15. Barry Ames, Political Survival: Politicians and Public Policy in Latin America (Berkeley: University of California Press, 1987).

16. Edward L. Gibson, Ernesto Calvo, and Túlia Falleti, “Reallocative Federal-ism in the Western Hemisphere,” paper presented at the Conference on Federalism, Democracy, and Public Policy, Centro de Investigación y Docencia Económica (CIDE), Mexico City, 14–15 June 1999.

17. Alfred Stepan, “Brazil’s Decentralized Federalism,” 165. Contemporary Mexico provides additional examples of persistent subnational authoritarian en-claves. See Richard Snyder, “After the State Withdraws: Neoliberalism and Subnational Authoritarian Regimes in Mexico,” in Wayne A. Cornelius, Todd A. Eisenstadt, and Jane Hindley, eds., Subnational Politics and Democratization in Mexico (La Jolla, Calif.: Center for U.S.-Mexican Studies, University of Califor-nia–San Diego, 1999), 295–341.

18. Juan E. Méndez, Guillermo O’Donnell, and Paulo Sérgio Pinheiro, eds.,

The (Un)Rule of Law and the Underprivileged in Latin America (Notre Dame: University of Notre Dame Press, 1999).

19. Alonso Lujambio and Horacio Vives, “Nota Sobre la Redistritación,” unpubl. ms., September 2000. On IFE’s important role in Mexico’s transition to democ-racy, see Andreas Schedler, “Mexico’s Victory: The Democratic Revelation,” Journal of Democracy 11 (October 2000): 5–19.

20. Jorge Domínguez and James McCann, Democratizing Mexico (Baltimore: Johns Hopkins University Press, 1995), 118–19.

21. It should be noted, however, that in Venezuela, as in Germany, “extra” seats can be allocated to different states to make the party results more proportional.

22. One such instance is the automatic granting of a Senate seat to ex-presidents in some Latin American countries (Chile and Venezuela, for example).