The Third Wave of Democracy:Findings and Implications
Latin America contains close to twenty countries, providing a valuable laboratory to explore the big questions of political economy. Outside of Europe, it provides the only long-term political experiences useful for assessing alternative explanations of democratization and their impact on development. Peter H. Smith's regime classification suggests that dictatorships have ruled in the region for slightly less than 50 percent of "country years" during the twentieth century.1 Competitive political systems with honest elections and universal voting rights have governed for 26 percent of the time. Because of either suffrage restrictions or less-than-competitive elections, semidemocratic regimes have been the norm during the remaining 25 percent of the time.
The four books under review each contribute to an understanding of political development in the region during the twentieth century. The focus of Democracy in Latin America and of The Third Wave of Democratization in Latin America is the recent shift from authoritarian to democratic governments in Latin America and throughout the world (see Huntington 1991). In both cases, the overall assessment is that democracy is probably here to stay in the region, but the quality of democratic life in many of its regimes leaves much to be desired. [End Page 245]
La elección presidencial mediante doble vuelta en Latinoamérica provides a wealth of ideas about a major electoral reform during the third wave. Prior to the late 1970s, thirteen of eighteen countries awarded the presidency to the candidate who amassed more votes than any single rival. As of this writing, only five countries—Honduras, Mexico, Panama, Paraguay, and Venezuela—continue to use this system. Most now require that the most-voted-for candidate either obtain more than 50 percent or some threshold number of votes or face a challenger in a runoff election.
The most empirically ambitious work under review is Dieter Nohlen's compilation of electoral and political statistics, Elections in the Americas. It is or it should be an indispensable reference for any effort to make sense of the political and electoral dynamics of the region.
The Quality of Third Wave Political Systems
Neither The Third Wave nor Democracy in Latin America expects a return to authoritarian government. Yet too many political systems are incapable of providing solutions to pervasive joblessness, alarming levels of public insecurity, and widespread poverty, the three most cited problems facing the region according to the polls adduced in Frances Hagopian's insightful chapter in The Third Wave.
The latter compilation began as a conference at the Kellogg Institute, at the University of Notre Dame, in 2001 and is destined to become a standard reference book on Latin American politics in the last decades of the twentieth century. Its chapters treat most major countries in the region as well as several smaller countries that are typically and unfortunately ignored. It also includes a perceptive statistical study of the region since 1978 by Scott P. Mainwaring and Aníbal Pérez-Liñán. Using a cross-national database of economic, social, and political variables, the latter show that per capita gross domestic product (GDP) and inflation have no statistically significant relationship to the transition to and durability of democracy between 1946 and 1999. Only a regional dummy variable (measured as the number of democratic countries in any given year minus the country tested) is associated with the shift to democracy. In models of regime breakdown, the collapse of democracy is a function of the number of parties, the polarization of the party system, and the quality of democracy.
Mainwaring and Pérez-Liñán nevertheless make several useful points. First, there is an international demonstration effect in Latin America. Democracies seemingly encourage their neighbors to become democratic, a statistical finding related to a favorable ideological climate and the prodemocratic behavior of the Catholic Church, the United States, the Organization of American States, and other multilateral institutions. The authors [End Page 246] point out that the United States supported democracy in the region in part to justify its anti-Sandinista crusade during the 1980s. This interpretation of regional dummy variables requires more thorough investigation in my view, given that the United States did actively support antidemocratic forces in some countries (Guatemala in 1954 and Chile in 1973 are two good examples) and, during the Cold War, tolerated anticommunist authoritarian regimes. However, it is certainly the case that the United States promoted democracy in the region, especially after the fall of the Berlin Wall in 1989 reduced the threat—real or imagined—of Soviet- or Cuban-sponsored subversion in the Western Hemisphere.
A common theme of nearly all the chapters in The Third Wave is that democracy may be consolidated, but it is of low quality. This is a central conclusion of Kurt Weyland's study of Brazil, which in 1985 lifted the ban on illiterate voters. Although Brazil's military no longer threatens to overthrow the government and the Left has won important elections, including the presidency in 2002, Weyland notes that patronage is still important in politics (although he does not discuss its extensiveness) and that corruption remains a problem. In 2000, Brazil ranked only forty-ninth of ninety countries on the Transparency International scale. The chapters on Argentina (Steven Levitsky) and Colombia (Ana María Bejarano and Eduardo Pizarro) make similar points for these countries.
The chapters on Guatemala (Mitchell A. Seligson) and Venezuela (Michael Coppedge) are among the best of the entire collection because each marshals an array of country-specific and cross-national evidence to explain the phenomenon of alarmingly low-quality democracy. Seligson contends that political apathy, along with a state that does not collect more than 10 percent of the GDP in taxes, prevents the public sector from educating the population, building physical infrastructure, and resolving a multitude of conflicts that often erupt into violence. The armed left's defeat in Guatemala's civil war prevented the sort of structural reforms that have, in contrast, fundamentally realigned El Salvador's political system, as Elisabeth Jean Wood points out in her chapter. She argues that the key factor prompting political change in Salvadoran politics was the civil war: that the armed left's forcing of the right-wing government to the negotiating table led to the incorporation of the Farabundo Martí National Liberation Front (FMLN) in the political system and, most important, to the overhaul of the country's armed forces. She also points out that the traditional elite, whose power was based on the ownership of large coffee plantations, diversified its investments in the wake of extensive agrarian reform. In her chapter on Mexico, Beatrice Magaloni uses games in extensive form to explain why the Institutional Revolutionary Party (PRI) gradually reformed electoral laws in the wake of the controversial 1988 presidential elections. She persuasively argues that the threat of opposition boycotts and protest encouraged a PRI not overly concerned about [End Page 247] defeat to create an independent court system to organize elections and tally the vote.2
Coppedge does a wonderfully systematic job of comparing the decline of democratic government in Venezuela with global trends. He uses several statistical models to show why the decline in GDP per capita since the late 1980s cannot explain why Venezuela's two-party system collapsed in the 1990s. He argues that economic decline created a constituency for political change. He associates the number of households in extreme poverty, the value of the minimum wage, and the percentage of households in the informal sector with democracy in Venezuela as measured by Freedom House scores between 1973 and 1999. The oil curse, whereby citizens and parties look to a revenue-rich state to solve all of their problems, laid the groundwork for the traditional party system's disintegration. Hugo Chávez therefore found a constituency for a discourse blaming the traditional political class for long-term economic decay, which his government has done little to arrest.
Good comparative essays and careful editing prevent the chapters in The Third Wave from moving in different directions. Nevertheless, I would have liked to see each country chapter discuss similar issues. While Weyland argues that international developments—market-based reforms in the wake of the 1982 debt crisis and the fall of the Berlin Wall—induced the changes that stabilized Brazilian democracy, René Antonio Mayorga argues that behavioral changes among politicians and institutional reforms allowed democracy to endure in Bolivia. Mainwaring and Pérez-Liñán imply that Weyland is right. I would have liked to see Mayorga take a position on this crucial issue. Similarly, it would have been great to see every chapter discuss public attitudes toward third-wave developments and then identify the reasons why citizens are more or less satisfied with democracy.
Sponsored by the United Nations Development Program (UNDP), Democracy in Latin America is the outgrowth of a project directed by Dante Caputo. It is less an academic study than a critical assessment of nearly two and a half decades of neoliberal democracy in Latin America. While some of the policy recommendations strike me as wrongheaded or based on questionable empirical assumptions, the report does a wonderful job of focusing attention on dissatisfaction with democracy in the region and its shortcomings. Its key strength is a survey of citizen attitudes in every country of the region. It also has chapters on the less-than-exciting development performance of the region, the perceptions of a nonrandom [End Page 248] sample of 231 Latin-American leaders, and ideas on how to transform "democracies of voters to democracies of citizens." The main report, along with the Web site (www.democracia.undp.org) and CD, are milestones in Latin American social science.
The finding that a slight majority of Latin Americans (56.3 percent) believed that economic development is more important than democracy generated a lot of press for this publication, which builds upon this and many other survey results to develop a threefold classification of Latin American political attitudes: 26.5 percent of Latin Americans are nondemocrats, 30.5 percent are ambivalent about democracy, and 43 percent are democratic. The research team coordinated by Jorge Vargas Cullel must be complimented for using social science to speak to perennial issues that all-too-many pundits discuss without documenting.
My own reservations are threefold. First, the chapter on public opinion should have presented more disaggregated findings. Lumping countries into four subregional groupings conceals important sources of variation. Mexico, for example, is joined to Central America, a group with countries as politically divergent as Costa Rica and Guatemala. Although Mexico has, along with Costa Rica, the highest GDP per capita in this subregion, it has the highest percentage (34 percent) of nondemocrats. Guatemala has the next largest group of nondemocrats (22 percent). At the other extreme, only 9.5 percent of Costa Ricans are nondemocratic. The proportion of nondemocrats in the rest of Central America falls between Costa Rica and Guatemala. It would have been better to include these and related findings in the report itself and not just in the statistical compendium.
Second, a central implication of the work on public opinion is not explored. If 57 percent of Latin Americans are nondemocratic and ambivalent, the report might have identified the political conditions that prompt the formation of antidemocratic coalitions and the breakdown of democracy. There is, after all, a rapidly developing literature on civil wars that speaks to this issue.3 One major study concludes that the nature of the political system, more than anything else, determines whether political forces take their struggle from the legislature to the battlefield.4 Engaging this literature would have allowed the authors to identify countries that are more or less at risk of instability. [End Page 249]
Third, the final, largely normative section of the report ("Four Issues on the Agenda for Debate") does not follow from its evidence. It contains numerous unsubstantiated claims and less-than-useful ways to pose complex issues. For example, it claims that "politics has become almost entirely an activity that has little connection with the identities, interests or aspirations of society" (178). Without presenting evidence to this effect, this remains an untested hypothesis, one that requires measuring citizens' preferences over the policies and preferences (and bills and laws) of elected officials to determine whether there is a divorce between state and society. Similarly, the report claims that "the debate on economics, dodged on the grounds that it is technically complex, is increasingly removed from the arena of public discussion" (185), without demonstrating that legislatures and society at large are not debating economic issues. Hagopian's chapter in The Third Wave provides ample evidence that voters care about these issues, to which parties competing in elections must listen—and are listening. In countries like Argentina, Bolivia, and Venezuela, left-wing nationalist parties have won mandates and are changing public policy. In other countries like Chile, El Salvador, and Mexico, more centrist governments are eschewing dramatic changes in policy because most voters value economic stability over fundamental changes in state policy.
An Institutional Innovation: Runoff Elections
La elección presidencial mediante doble vuelta might strike noninstitutionalists as needlessly focused on the minutiae of electing presidents. Quite the contrary: even a rapid overview suggests that electoral systems play a pivotal role in democratic survival and citizen satisfaction. A potentially serious problem with first-past-the-post systems is that a candidate with only a minority of votes can become president, as occurred in Chile in 1970, when the candidate of the leftist Popular Unity coalition, Salvador Allende, obtained 36.6 percent of the vote, barely 1 percent more than his right-wing rival, Jorge Alessandri Rodríguez. The 2006 Mexican elections—which saw Felipe Calderón, the right-of-center candidate of the National Action Party (PAN), beat Andrés Manuel López Obrador of the Party of the Democratic Revolution (PRD) by 35.89 percent to 35.31 percent of the vote—is another example, in which Mexico almost saw the election of a candidate most distant from the preferences of the median voter.5 [End Page 250]
In one of three comparative chapters in La elección presidencial, Daniel Buquet uses a naturally occurring experiment in Uruguay to show that this formula does reduce the number of parties. Whereas the average effective number of parties at the provincial level was 3.08 in the October 1999 presidential election (first round), it fell to 2.46 in May 2000 municipal elections using plurality. Daniel Chasquetti also examines third-wave elections to argue that runoff elections do not seem to foment coalition governments; in fact, nine of fifteen presidents elected in the second round form minority governments. Finally, Pérez-Liñán uses statistical analysis to argue that runoff systems are associated with democratic stability. However, runoffs play a part in provoking military intervention if the winner was second or third runner-up in the first-round election.
La elección presidencial makes it clear that Latin American countries have experimented with alternative ways of electing presidents, in part because they were aware of the potentially perverse results of plurality formulas. Bolivia, for example, continues to use a runoff system, but with an important twist: if no candidate obtains an absolute majority of the vote, it is up to the newly elected legislature to select the president. Chile also used a congressional runoff system until 1970, one also based on the norm that a congressional majority should ratify the popular election of the first runner-up. President Allende, in fact, came to power as the result of a pact between Popular Unity and the centrist Christian Democrats. Since 1990, a popular runoff system for electing the president remains in effect in Chile. The perhaps most famous qualified plurality system in Latin America developed in Costa Rica by 1936, and not in 1949 as Rivas avers. It requires the successful candidate to obtain a plurality and at least 40 percent of the valid votes to become president. This was a product of the dominant party's calculation that lowering the threshold to win the presidency by ten points would discourage intraparty rivalry.6 As of this writing, Argentina, Ecuador, and Nicaragua use variations of this system to elect their presidents.
The volume would have been stronger if it had asked authors of the country chapters to explore the origins of the shift from plurality to majority runoff systems. The coexistence of the two systems raises the question of why, for example, Brazil and Chile chose majority runoff while Argentina opted for a qualified plurality system (in which a runoff is held only if the front-runner either has less than 40 percent of the vote or does [End Page 251] not have a ten-point lead over the runner-up and obtained between 40 and 45 percent of the popular vote). Unfortunately, there is no reference in the volume to public debates, committee reports, or deliberations by constituent assemblies to determine the role played, for example, by party-system fragmentation and personal ambition in this institutional innovation. With the exception of Chasquetti, Martínez, and Pérez-Liñán, the authors of most chapters also stay away from broader debates about the origins of these systems and about implications for electoral systems debates.7
The Electoral Characteristics of the Waves
Dieter Nohlen's compilation of electoral and political statistics, Elections in the Americas, translates and updates the second edition of a work first published in German and Spanish in 1993. The latter was an immediate hit, appearing as the third wave of democratization was cresting. The present work adds twelve years of election returns, and includes Canada, the United States, and a host of Caribbean countries.
A drawback, however, is that it corrects very few of the omissions or errors of the first edition. First, there are no results for legislative races or the distribution of seats for many countries until or shortly before the inception of the third wave. The chapter on Ecuador, for example, has no legislative seat results before 1979, even though the congress played an active role in Ecuadoran politics since at least the 1920s. Much of the same holds true for El Salvador, Guatemala, Honduras, Nicaragua, Panama, Paraguay, and Peru. Other countries such as Bolivia have no electoral data whatsoever before the 1950s, although sources such as Herbert F. Klein's first book on Bolivia do mention a fair number of such data.8 Many of the chapters, in other words, do not exhaust secondary sources, much less use newspapers, official government gazettes, or the U.S. State Department and German Foreign Ministry records (which typically contain valuable reports on elections) to provide readers with relevant time-series data. Second, most chapters do not have bibliographic essays explaining why certain sources were given precedence over others with different results. Third, the tables of election dates found at the beginning of each chapter are incomplete. These also not infrequently list coups that are then not mentioned in the list of presidents at the end of each chapter. For example, Nohlen's chapter on Chile only mentions the 1973 coup in table 2.1, whereas table 2.10 ("List of Power Holders, 1920-2004") does mention that [End Page 252] there was also a coup in 1924 and two in 1932. Fourth, more than a few basic errors suggest that neither Nohlen nor reviewers fact-checked each chapter. For example, Zovatto is wrong to claim that "economic standing and educational requirements were eliminated" from the franchise in 1913 in Costa Rica (151). The 1871 constitution placed only gender and vaguely worded property requirements on the franchise so that, by the early twentieth century, nearly all men aged twenty or older were registered to vote.9 Both requirements were dropped from the 1949 constitution. Finally, it is unfortunate that Elections in the Americas does not come with a CD to make the quantitative data available in spreadsheet form.
These books take us a long way toward understanding the third wave of democratization in Latin America. In the aggregate, they generate three broad conclusions. First, if Mainwaring and Pérez-Liñán are right, neither social structural nor economic change led to the disintegration of authoritarian regimes in the region. Instead, international changes interacted with domestic political factors to generate the third wave of democracy in Latin America. How these factors promote regime change becomes clearer in the country-specific chapters in The Third Wave and La elección presidencial mediante doble vuelta. Second, the third wave has produced democracies of varying quality. Both The Third Wave and, more generally, Democracy in Latin America make the point that most Latin Americans get unaccountable governments that undersupply public services. Democracy in Latin America uses modern survey techniques to show that more than one-quarter of Latin Americans are nondemocratic and almost one-third are ambivalent about democracy. Third, institutional change, like the regionwide shift to majority runoff systems, seems to be helping to consolidate democracy. That candidates least preferred by a majority of the electorate have a harder time getting to the presidency in most countries of the region increases democracy's chances for success.
The four books reviewed in this essay also underscore the importance of strengthening efforts to collect and analyze good time-series data on politics. Although Elections in the Americas remains an indispensable source of electoral information, future editions must do a more thorough job charting changes in electoral law and providing basic data on democracy and elections. The field must also build on the public opinion work in Democracy in Latin America, both by replicating its findings and by [End Page 253] improving on its methods. The UNDP report and La elección presidencial mediante doble vuelta also jointly suggest that surveys can be used to determine the congruence between the preferences of citizens and elected offi cials. How democratic political systems are and how they perform should become priority topics for future research. [End Page 254]
Fabrice Lehoucq holds a Ph.D. in political science from Duke University and is a specialist in institutional analysis, electoral politics, and political economy. Lehoucq is the author of several books, including lead author of Stuffing the Ballot Box: Fraud, Democratization, and Electoral Reform in Costa Rica (Cambridge University Press, 2002) and articles in comparative political studies, comparative politics, and electoral studies. He has received support for his research from the Inter-American Development Bank, Kellogg Institute (University of Notre Dame), the Alexander von Humboldt Foundation, the National Endowment for the Humanities, the Social Science Research Council, and the World Bank. At present, he is at work on a book called Political Institutions, Instability, and Democratic Performance in Latin America. Lehoucq is on the faculty at the Centro de Investigación y Docencia Económicas (CIDE) in Mexico City and, in August 2005, will become an associate professor, Department of Political Science, University of North Carolina, Greensboro.
1. Peter H. Smith, "Los ciclos de la democracia en América Latina," Política y Gobierno 11 (2004): 189-228.
2. I make much the same argument about why incumbents reform electoral laws in a paper that Magaloni does not cite. See my article "Institutional Change and Political Conflict: Evaluating Alternative Explanations of Electoral Reform in Costa Rica," Electoral Studies 14 (1995): 23-45.
3. See, e.g., Paul Collier and Anke Hoeffler, "Greed and Grievance during Civil Wars," Oxford Economic Papers 54 (2004): 563-595; James D. Faeron and David D. Laitin, "Ethnicity, Insurgency, and Civil War," American Political Science Review 97 (2003): 75-90; and Stathis N. Kalyvas, "The Ontology of 'Political Violence': Action and Identity in Civil Wars," Perspectives on Politics 1 (2003): 475-494.
4. Jack A. Goldstone, et al., "A Global Forecasting Model of Political Instability," paper presented at the Annual Meeting of the American Political Science Association, Washington, DC (September 1-4, 2005).
5. Kenneth A. Greene, "El votante mediano y la elección presidencial por mayoría relativa en México," Política y Gobierno 14 (2007): 203-214. See also Cynthia M. McClintock, "Plurality versus Majority Runoff Rules for the Election of the President in Latin America: Insights from the 2006 Peruvian and Mexican Elections," paper prepared for presentation at the meeting of the American Political Science Association, Philadelphia, PA (August 31-September 3). Mathew Søberg Shugart, "Mayoría relativa vs. segunda vuelta: La elección presidencial mexicana del 2006 en perspectiva comparada," Política y Gobierno 14 (2007):175-202.
6. Fabrice Lehoucq, "Costa Rica: Modifying Majoritarianism with a 40 Percent Threshold," in Josep M. Colomer, ed., Handbook of Electoral System Choice (London: Palgrave Macmillan, 2007).
7. Gabriel L. Negretto, "Choosing How to Choose Presidents: Parties, Military Rulers, and Presidential Elections in Latin America," Journal of Politics 68 (2006): 421-433.
8. Herbert F. Klein, Parties and Political Change in Bolivia, 1880-1962 (Cambridge: Cambridge University Press, 1969).
9. Fabrice Lehoucq and Iván Molina, Stuffing the Ballot Box: Fraud, Electoral Reform, and Democratization in Costa Rica (New York: Cambridge University Press, 2002), 41-43.