Can politicians buy votes? The records of a Brazilian legislator may show us how.

In the APSA Public Scholarship Program, graduate students in political science produce summaries of new research in the American Political Science Review. This piece, written by Kumar Ramanathan, covers the new article by Daniel W. Gingerich, University of Virginia,  Buying Power: Electoral Strategy before the Secret Vote”


One of the greatest challenges facing candidates for political office is figuring out the most effective way to use their resources to win elections. In the US, we often hear about this challenge in terms of how much time a presidential candidate spent in a battleground state, or how much money they spent on advertising in key areas. A less well-known strategy is to rely on local “brokers”—such as local officials, employers, and traditional local leaders—who turn out voters in exchange for money, government jobs, or other benefits. While this practice was more common historically, it continues in some democracies today. In a new APSR article, political scientist Daniel W. Gingerich investigates whether this strategy is effective. He shows that it is possible for candidates to efficiently win votes through local brokers, if they can learn from and reward brokers’ performance in previous elections. Capanema relied on local brokers to secure votes. Since votes were public, brokers could accurately claim the number of voters that they had turned out for Capanema.

Gingerich draws on the records of an influential Brazilian politician, Gustavo Capanema, who served as a legislator from 1946 to 1970. Capanema ran for office in the era before the “secret ballot.” In many early democracies, parties or candidates printed ballots and distributed them to voters, which effectively made votes publicly visible. The secret ballot, where voters mark their preferences anonymously, only became commonplace in the 19th century. It was adopted in Australia in the 1850s, in the United Kingdom in 1872, and in the United States in the 1890s. Brazil was a late adopter of the fully secret ballot: some rural areas used candidate-printed ballots as late as 1970. In this context, Capanema relied on local brokers to secure votes. Since votes were public, brokers could accurately claim the number of voters that they had turned out for Capanema. He kept detailed records of his interactions with brokers, who were mostly municipal officials such as mayors. These records listed the brokers for each municipality, the number of votes they promised, and the amount of money that he sent to them.

Analyzing these records, Gingerich finds that Capanema rewarded local brokers who had done well delivering votes in previous elections. Where vote totals exceeded brokers’ promises, Capanema increased his financial contribution in the next election. On average, a broker who turned out an excess of 100 votes in the 1958 election for Capanema received an increase of 5,267 cruzeiros (about $360 in 2020 U.S. dollars) for the 1962 election. Conversely, Capanema penalized brokers who underperformed, especially if they had not successfully brought the vote out for Capanema in earlier elections. For example, a broker who had underperformed in 1958 and had not had a strong showing for Capanema in 1954 elections would receive less payment ahead of the 1962 election, but a broker who only underperformed in 1958 but otherwise had a past record of success would experience a smaller penalty. By tracking and learning from brokers’ performance across election cycles, Capanema was able to efficiently distribute resources to brokers and turn out voters across his district. When candidates are able to learn from past spending of campaign resources, they can use those resources much more effectively in the long-term.

This system was an unusually cost-effective way to spend campaign resources. Gingerich calculates that Capanema spent roughly $28 per vote (in 2020 U.S. dollars) in the 1962 election. Compared to elections elsewhere, this is a remarkably low cost per vote. For example, in the 1868 United Kingdom election, candidates spent roughly $570 (in 2020 U.S. dollars) per vote. Scholars of contemporary American politics estimate the cost of door-to-door canvassing as $47 per vote and the cost of general campaign spending as $145 per vote. In the contemporary era, political candidates in many countries (such as Mexico, Senegal, and Liberia) continue to use local brokers, but the presence of the secret ballot makes it more difficult for brokers to use aggregate voting totals to demonstrate their success. This likely means that candidates are not able to “buy” votes as efficiently.

Gingerich’s analysis shows that it is possible for politicians to use local brokers to win elections, and to use past reputation to decide how their spending should be distributed. When candidates are able to learn from past spending of campaign resources, they can use those resources much more effectively in the long-term. This finding about learning processes and localized knowledge matters for candidates today, even if they do not use local brokers


  • Kumar Ramanathan is a doctoral candidate in political science at Northwestern University and will be a doctoral fellow at the American Bar Foundation beginning in the fall of 2020. His dissertation investigates how liberal politicians in northern Democratic Party contested and constructed a civil rights legislative agenda during the 1930s-60s, and aims to explain the origins and limitations of racial liberalism as it emerged among these party elites. His research agenda also includes a set of projects on the impact of civil rights law and policy on the politics of social policy after the 1960s, and collaborative projects on immigrant political participation and urban politics. At Northwestern, Kumar is affiliated with the Chicago Democracy Project, the Comparative Historical Social Sciences Working Group, and the Program in Legal Studies. He received his B.A. in political science and philosophy from Tufts University.
  • Article details: American Political Science ReviewFirst ViewBuying Power: Electoral Strategy before the Secret Vote DataPublished online: 23 July 2020
  • About the APSA Public Scholarship Program.

Be the first to comment

Leave a Reply

Your email address will not be published.