‘The Birth of Democracy in South America’ appears at a timely moment, returning attention to a debate that has recently been eclipsed by mounting concern over the contemporary crisis of democracy. Dense yet innovative in several respects, the volume examines the process of democratization in Latin America during the late nineteenth and early twentieth centuries. It is a significant contribution to the literature and will likely become an essential reference for future scholarship on the subject.

In the book, Raul Madrid (2025) traces the course of democratization in ten Latin American countries (Chile, Uruguay, Argentina, Colombia, Brazil, Peru, Venezuela, Bolivia, Ecuador, and Paraguay). The author’s argument – set out in the Introduction and developed in the first chapter – is clear and parsimonious: three variables account for why some countries democratized while others did not. The professionalization of the military (discussed in the third chapter) reduced incentives for opposition groups to resort to force, particularly through revolts and rebellions. Given the limited prospects of overthrowing governments through armed conflict, opposition groups increasingly turned to electoral competition. The role of political parties is examined in the fourth chapter. According to Madrid (2025), the presence of strong opposition parties pressured governing parties to enact democratic reforms, such as expanding suffrage and introducing the secret ballot. The success of these reforms, however, depended on instability within the government itself. Splits within the ruling party created opportunities for these reforms to secure parliamentary approval.

Raul Madrid (2025) tests his argument through careful research grounded in an extensive and systematic use of the secondary literature, which led to the creation of two databases (both described at length in the second chapter). The LAHED (Latin American Historical Elections Database) includes the results of 263 presidential elections, covering 95 percent of all presidential elections held in South America between independence and 1929. The LARD (Latin American Revolts Database), meanwhile, compiles data on revolts and rebellions during the same period. Taken together, the evidence indicates that the conditions necessary for democratization - the professionalization of the military, the presence of strong political parties, and divisions within governing elites that paved the way for reform - were present only in Chile, Uruguay, Colombia, and Argentina, as argued in meticulous detail in the remaining chapters of the book.

To develop a more consistent critique of the book, I believe it is useful to focus on the central issues that intersect with contemporary debates on democratization. One such issue concerns political participation. Contrary to the revisionist findings of historians, according to whom there is substantial evidence of democratic practices in nineteenth-century Latin America, Raul Madrid (2025) acknowledges that, in many of the countries examined, the persistence of literacy requirements in electoral laws constituted a barrier to participation, particularly for Indigenous and Afro-Latin populations.

The differences between the findings of Raul Madrid (2025) – and, I would argue, political scientists more broadly – and those of historians are ultimately more superficial than substantive. Political scientists generally approach political participation through the classic questions of universal suffrage and voter turnout, whereas historians tend to focus on the different forms electoral participation could take, whether constrained, coerced, violent, or at times more autonomous, emphasizing its popular and festive character. These are two distinct ways of looking at the same phenomenon. In my view, the two approaches should be read symbiotically rather than in opposition to one another. One does not invalidate the other. An example helps clarify the point. Scholarship on Latin America has shown that, in the nineteenth century, voting was a collective practice rooted in local communities, the product of the efforts of candidates and parties that coordinated and organized ‘their’ voters into groups (ROMANELLI, 1998). Similar dynamics could also be found in other contexts, such as Imperial Germany (ANDERSON, 2000) and even the United States (BENSEL, 2000). These studies clearly demonstrate that voting rarely expressed an autonomous individual preference, not least because electoral laws generally failed to ensure ballot secrecy effectively. From this perspective, the crucial question becomes not simply how many people voted, but how voting operated in practice, which shifts historians’ attention toward other issues, such as the use of the press as an arena of political struggle or the large-scale mobilization of workers and Indigenous populations.

A point that merits attention is that quantitative data on political participation and voter turnout tell us little, if anything, about these electoral practices. Raul Madrid (2025) himself acknowledges that it makes little sense to insist on universal suffrage as a defining variable. Following Scott Mainwaring, Daniel Brinks, and Aníbal Pérez-Liñán (2001), he argues that this dimension should be omitted from the definition of democracy. As Madrid (2025) repeatedly notes throughout the book, the central issue for thinking about democracy should instead be the problem of free and fair elections, “even in the absence of universal suffrage, whereas universal suffrage is of little meaning if elections are neither free nor fair” (MADRID, 2025, p. 10). The problem, however, is that political participation is never entirely bracketed out and continues to surface throughout the analysis, especially in Chapter 02. Why?

A second variable of particular importance in political science approaches to democratization is political competition. The data presented in the book indicate that political competition was limited in many Latin American countries, as “elections generally became less competitive as the nineteenth century progressed” (MADRID, 2025, p. 64), a pattern that contrasts with the European experience. According to the evidence assembled by the author, 70 percent of presidential elections were noncompetitive, and at least one-third were entirely uncontested. The data are convincing, but they tell only half the story. For one thing, the lack of competition in presidential elections cannot simply be assumed to reflect the same situation in elections for other offices, such as seats in the lower house. Although access to this type of data remains limited, we know that within legislatures – particularly in Congress – the president-elect rarely enjoyed unanimous support, meaning that governing often required negotiation and coalition-building with other parties. It is high time that studies on democratization began to incorporate legislative elections more systematically, without which our understanding of the dynamics of political competition remains necessarily limited.

We must also bear in mind that elections with a single candidate were not necessarily devoid of political competition. Brazil offers a particularly revealing example. Following the collapse of the Empire, between 1889 and 1930, only four presidential elections can be classified as competitive (1910, 1919, 1922, and 1930), according to the criteria employed by the author himself (RICCI and ZULINI, 2023, see table 8.1 to 8.12). Yet the absence of competition in the remaining elections masks a widespread practice among political elites at the time: negotiations over the choice of the future president often took place months before the election itself (VISCARDI, 2012). Once an agreement had been reached, the election served largely to ratify a decision that had already been settled in advance. Political competition, in other words, did not take place at the ballot box but beforehand.

Bolivia in 1908 offers another revealing case. The election was uncontested because the Liberal Party and the Constitutional Party had reached a prior agreement over the distribution of candidacies: the Liberal Fernando Eloy Guachalla would assume the presidency, while the offices of first and second vice president would be assigned to two Constitutionalist politicians (VICTORIANO, 2000). Examples such as these suggest that political competition cannot be reduced to electoral competition alone, but must also include the pre-electoral agreements that shaped political dynamics in nineteenth-century oligarchic regimes.

A final point concerns the question of free and fair elections. When Raul Madrid (2025) turns his attention to the countries that democratized in the early twentieth century – Chile, Uruguay, Colombia, and Argentina – he argues that elections in these cases became more ‘free and fair’ following the adoption of democratic reforms. The same claim is made with respect to several nineteenth-century presidential elections as well (see Table 6.4 for a detailed breakdown). Elections are classified as free and fair because they were competitive, featured narrow margins of victory, and resulted in opposition victories. The difficulty, however, is that there is no direct evidence demonstrating that these outcomes resulted from an electoral process free from interference. For instance, we do not know whether opposition forces became electorally competitive insofar as they gained access to resources previously monopolized by the government, such as voter registration commissions. As scholarship on elections has shown, democratic reforms – such as the introduction of the secret ballot or the appointment of electoral judges to supervise elections – may also lead to new forms of fraudulent practices. As institutions change, political actors adjust their strategies and adapt to new rules (SJOBERG, 2014). The same may have occurred with the establishment of electoral courts in Chile and Uruguay during the 1920s (MOZAFFAR and SCHEDLER, 2002), or with the adoption of measures intended to secure ballot secrecy (KUO and TEORELL, 2017; MARES, 2015).

Here, it is important to stress that the author himself acknowledges that “it is not easy to assess how free and fair elections were during this period, given the limited data available and the potential biases of the sources” (MADRID, 2025, p. 66). This point, however, is crucial. Future research should devote greater attention to how the democratic reforms of those years were actually implemented in practice. Without confronting this issue directly, there is a risk of building interpretations of the region’s democratic trajectory on uncertain evidence. Could the short-lived nature of democratic experiences in Latin America be explained less by the strong presence of the military than by the persistence of fraudulent and manipulative electoral practices?

In sum, Raul Madrid’s book (2025) stands as an important contribution to comparative politics. It is a genuine effort at synthesis and analysis that encourages dialogue with neighboring fields, especially history. On the one hand, the book advances debate over the different dimensions of democratization; on the other, it pushes researchers to seek out new sources, data, and evidence from the period. This is a difficult but necessary task, and one that, I hope, may provide fertile ground for future collaboration between political scientists and historians.