Scholars of democratization have tended to conflate the expansion of suffrage with the establishment of democracy. But are voting and party competition really sufficient for democracy, in the absence of credible guarantees that the act of voting actually reflects citizens’ political preferences? Isabela Mares’s important book reminds us that the question of free and fair elections cannot be set aside after the de jure expansion of suffrage. Instead, a deep and appropriately skeptical analysis of the processes by which voters actually vote needs to be undertaken if we wish to understand the origins and deepening of effective democracy. From Open Secrets to Secret Voting, which focuses on the case of Imperial Germany (1870–1918), demonstrates the pervasiveness of electoral manipulation following the introduction of universal manhood suffrage, and highlights how parties, candidates, and powerful economic interests sought to maintain such practices, or end them. The book’s subtitle emphasizes that suffrage does not necessarily entail voter autonomy—and the realization of the latter was much more halting, and in some ways even more politicized than the former.

Mares’s contribution marks a crucial shift for the democratization literature, which in recent years has pivoted toward global samples and structural theories, as in our own work. Focusing on the processes by which electoral democracy actually operates on the ground is a crucial course change. However, her book does not herald a return to the “agentic” literature on regime change. Instead it examines how structural economic factors that vary at the subnational level, along with political elites’ strategic political considerations, shape incentives to manipulate elections. By burrowing deeply into one case, the book neatly combines structural and strategic arguments to advance theoretical and empirical debates about regime change. It is sure to become an influential and well-cited contribution to our understanding of the way in which democratization plays out in practice.

Despite focusing on a few decades in one country, Mares covers an expansive empirical terrain—from qualitative accounts of ballot-box design and election intimidation to statistical analyses of subnational election manipulation, parliamentary roll calls, and voting behavior. The empirical work of the book derives cleanly from her theory, as laid out in Chapter 2, which develops an elegant account of the sources of supply and demand for electoral manipulation. Mares argues that demand for electoral manipulation came particularly from conservative politicians in marginal districts, who faced credible competition from the Social Democrats and/or from other conservative candidates. Conservative demand was met by the “suppliers” of electoral manipulation, particularly state employees (e.g., police) and/or of employees of major corporations, who each had interests in protecting the political and economic status quo. The author emphasizes that local socioeconomic factors—particularly industrial concentration, which affected the relative ease of monitoring and manipulating voters—shaped suppliers’ incentives.

There are three empirical sections that explore the implications of this theory: one that looks at the practice of electoral manipulation itself, a second that examines efforts to adopt anti-manipulation legislation, and a third that considers how curtailing manipulation affected the electoral fortunes of the Social Democratic Party and the emergence of proportional representation.

Chapters 3 and 4 explore qualitative evidence about the mechanisms of manipulation, comparing Prussia to Baden and Württemberg and Saxony to leverage some of the important socioeconomic and political differences across German regions. Building on Daniel Ziblatt’s work on electoral fraud in Imperial Germany, Mares in Chapter 5 finds strong support for her core theoretical conjectures: Districts with narrow margins, high economic concentration, moderate Socialist support, and lower right-wing fragmentation had higher levels of reported fraud.

Chapters 6 and 7 focus on efforts to reduce electoral manipulation, examining a series of legislative votes between 1870 and 1912 and demonstrating again that socioeconomic factors, particularly industrial concentration, rural employment and wage patterns, and district skill levels all exerted a strong impact on Germany legislators’ voting patterns.

Chapters 8 and 9 examine the “post-manipulation” era. Chapter 8 shows that curtailing electoral manipulation meant that areas with lower skill levels, where workers had been more easily intimidated in large factories or on farms, saw increased support for the Social Democrats. (Areas of high skill always had Social Democrat voters, because their highly specific skills protected them from retaliatory sackings by employers.)

Chapter 9 examines the debate in Germany over the introduction of proportional representation (PR), a question that sparked Stein Rokkan’s famous hypothesis that socialist strength drove the adoption of PR across Europe. This question remains the subject of heated debate among Carles Boix, Torben Iversen and David Soskice, and others. Mares makes a valuable contribution, noting first that scholarship has been pitched at the cross-national level, impeding efforts to distinguish competing theoretical mechanisms, particularly individual legislators’ district-level incentives versus parties’ national-level incentives. She shows that Rokkan’s hypothesis involves two distinct mechanisms: At the district level, candidates facing greater right-wing fragmentation tended to support PR. Meanwhile, at the national level, parties tended to support PR if their electoral support was relatively more geographically diffuse. Mares finds no support for the argument other scholars have offered, that support for PR was greater in regions with highly skilled workers. Her findings offer a substantive advance in this important debate about the origins of modern European democracy.

In broad perspective, Mares’s argument fits neatly into but also builds clearly on well-established notions. For example, her theoretical argument echoes Robert Dahl’s core insight about the origins of democracy, that pressure for reform depends on the relative cost of repression. Her argument that more complex local economies generate greater demand for democratic reforms also echoes Seymour Lipset’s version of modernization theory. We are not, however, suggesting that one can simply boil Mares’s argument down to these nuggets, particularly given her novel theoretical focus on subnational variation in labor-market conditions. After all, her account implies that two similarly sized cities with similar levels of per capita income but different occupational structures should evince very different levels of demand for democratic reforms. Mares emphasizes that the devil of democratization is in the details of subnational variation in socioeconomic structure, rather than the national socioeconomic aggregates that cross-national research typically employs.

From Open Secrets to Secret Voting clearly demonstrates the payoff that attention to the impact of labor-market structure has on the “democratization of democracy,” so to speak, and for this reason it should have wide impact. However, we do wonder about the argument’s generalizability. On the one hand, efforts to manipulate elections remain widespread. The question is, however, whether Mares’s approach can explain the relative success of such efforts. Here, we must raise the question of the payoff of a case-study versus a cross-national approach. For example, large-n research has exposed strong time-period effects of “endogenous” factors of democratization, both per capita GDP (e.g., the work of Boix and Stokes) and land and income inequality (our own work). It is possible that the same holds for the socioeconomic factors, such as labor-market conditions, that Mares suggests are theoretically crucial. In addition, her book ignores international factors, which powerfully shaped the possibilities for regime change in the post—World War II era—both repressing and providing openings for democratization (e.g., Huntington, Levitsky, and Way), and may have had such effects in earlier eras as well, as Kurt Weyland has recently argued. Changes in the impact of international factors may help explain the varying time-period effects of “endogenous” factors.

The question of generalizability arises with any case study. Mares even acknowledges the issue of comparability, quoting Anderson (Practicing Democracy, 2000) that employer intimidation in Imperial Germany was more pronounced than elsewhere in Europe at the time. This raises the question of the utility of a within-case research design for exploring the question of regime change more broadly. It is true, as Mares argues, that explanations for the high level of voter intimidation in Germany ignore variation in intimidation within Germany (p. 70), but one might expect similar variation across the United Kingdom, and the fact that Mares reveals variation within Germany does not call into question Anderson’s cross-national point. We already know that the UK and Germany took different paths to democracy and dictatorship during this era, but Mares leaves it to future research to tell us whether the reasons can be traced to the relative costs of voter intimidation. We look forward to seeing how Germany fits within that cross-national context in her next work.

Finally, we have one specific quibble with Mares, who misinterprets how we operationalize one of the two key independent variables in our 2014 book, Inequality and Democratization: rural Inequality (p. 160). Given this mistake, her approach to exploring the rural sector in Chapter 6 actually does not diverge from ours, but rather makes the same point we do: Landholding inequality (as Boix employed it in Democracy and Retribution [2003], e.g.) is not, on its own, an optimal variable to use in statistical analysis, because it ignores the relative supply of rural labor. Like Mares, we argue (Inequality and Democratization: An Elite-Competition Approach, 2014) that what matters politically is the interrelationship between the relative inequality of landholding and the relative supply of rural labor. As such, we weight land inequality by the proportion of the population that lives in rural areas. In this sense at least, our arguments do not diverge but rather complement each other.