Vote buying is an especially pernicious form of corruption, as it threatens to undermine the democratic process, thereby enabling other forms of corruption and misgovernance. One of the most important institutional mechanisms designed to inhibit vote buying is the secret ballot. First adopted in Australia in 1856, the secret ballot is now the norm in liberal democracies. Yet vote buying remains a problem in many such democracies, especially in the developing world (see here, here, and here). To better understand how to fight this form of corruption, it is important to understand why the secret ballot is not sufficient to eliminate the problem. While different communities face distinct challenges, there are several common mechanisms that allow vote buying to persist, even when it is impossible for the vote-buying politicians to verify that the voters they bribed actually voted the way that they had promised.

• First, when the relevant voting unit (such as a precinct or municipality) is sufficiently small, political operatives can draw inferences about how individuals voted. For one thing, in very small units—say, a county with only a few dozen voters—it is possible to make well-informed guesses about how individuals voted—and whether they voted the way they promised—from aggregate vote counts. Additionally, in small, tight-knit communities, political operatives can use a range of vote-adjacent behavior, such as attending rallies or using party-provided transportation to the polls, to  make good guesses as to how members of their social network are voting. The close personal relationships that party operatives have with the people they bribe also helps overcome the confidentiality that the secret ballot is supposed to provide. For example, one operative in Argentina reported that he knows how his neighbors voted based on whether or not they can look him in the eye following an election.
• Second, vote buying, particularly in close-knit communities where the political operatives and the voters know each other personally, is facilitated by the strong reciprocity norms that often exist in such communities. Simply put, even if one puts aside the possibility of indirect monitoring, individuals receiving money might feel duty-bound to vote as they are paid, rather than “cheating” by taking the money and then casting a different vote.
• Third, while the term “vote-buying” implies that political operatives are paying voters to change their vote, often political operatives will use bribery to manipulate turnout. That is, party operatives may offer money or other inducements to their own supporters to go out and vote; they may also bribe voters who are likely to support the opposing candidate opponents to stay home. This strategy has an old pedigree—almost as old as the secret ballot itself. Turnout buying in New York was a reaction to the adoption of the secret ballot in the late nineteenth century: direct vote buying decreased dramatically, but voters suspected to have sympathies for one candidate were paid by his opponent to abstain. In the modern world, turnout buying continues to occur in a variety of countries, including Venezuela, Argentina, and the United States.
• Fourth, in some jurisdictions vote buying is made easier by the fact that voters cast a party-specific ballot at the ballot box, rather than filling out one ballot that lists every party’s candidates for each position. In some cases, party-specific ballots are distributed not only at polling locations on election day, but also by political operatives in the days or weeks leading up to the election. The fact that operatives can observe whether voters take a particular ballot is makes it much easier to buy votes. True, a voter could accept a bribe and take one party’s ballot, but then pick up another party’s ballot at the polling station, but the act of taking the ballot from the operative can exacerbate the pressure voters experience, especially in situations with strong reciprocity norms and tight social networks described above. In Argentina, for example, voters will often be given money and the ballot by the same political operative, and will then be given a ride in a party-owned bus to the polls, all of which can create a heightened sense of obligation.

For these and other reasons, the secret ballot, while hugely important, is not by itself sufficient to fix the enduring challenge of vote buying. But institutional design can clearly play a role in facilitating or deterring direct and indirect vote buying. Better understanding the factors listed above naturally suggests at least some straightforward reforms, including abolishing party ballots (or at least prohibiting the distribution of those ballots other than at polling stations on election day), reporting election results at a higher level of aggregation (to make it harder to infer individual votes from jurisdiction-wide vote totals), and perhaps making voting mandatory in order to deter turnout manipulation. To be sure, other problems, like social reciprocity norms, are cultural and therefore harder to address through top-down approaches or legal changes. But institutional changes can have cultural effects. Direct and indirect exposure to vote buying accustoms voters to the practice and decreases the stigma associated with it.  By making vote-buying more difficult, and therefore less common, institutional changes can prevent the normalization of vote-buying and thus impact social norms.

The integrity of elections requires that the vote, rather than just the ballot, remain truly secret. It also requires institutional safeguards against interference with the many steps to voting before the ballot is actually cast. Democratic backsliding and the rise of authoritarian and illiberal forces, even in established democracies, threaten institutional foundations that when firmly rooted and longstanding can be a bulwark against corruption. To counter these threats to democratic governance, citizens must be able to participate in the political process with the most powerful tool in their arsenal: their free and secret vote.