Subscribe to E-Bulletin Donate to CEGA

Vote Buying and Reciprocity

Development Challenge

Vote buying, or the exchange of in-kind goods or money for a vote for a specified politician, is a common occurrence in many developing countries. It is widely acknowledged that the practice detrimentally affects the democratic process and can have widespread, negative socioeconomic consequences. This practice persists despite the introduction of secret ballots, which cannot be viewed by the “buyer” of the vote, and therefore are often expected to serve as a solution to vote buying. Strong cultural norms of reciprocity – ie. treating others the way you have been treated – may enable vote-buying to flourish even when there is no binding commitment; however, evidence of this phenomenon remains limited.


In 2006 the Paraguayan political system was dominated by two major parties: Colorado and Liberal. Ideological differences between the two parties were minimal, and Transparencia Paraguay found vote-buying to be common practice leading up to the 2006 elections. Within this context, researchers sought to identify a relationship between reciprocal preferences and the incidence of vote buying.

Evaluation Strategy

Researchers analyzed the data from Transparencia Paraguay and the Land Tenure Center household surveys to determine whether and how reciprocity affected targeted vote-buying in the 2006 national elections using multivariate regression. The Land Tenure Center surveys included thirty households in each of 15 villages were randomly selected for a total sample size of 450 households throughout rural Paraguay. To measure reciprocity, the survey asked if an individual would always, sometimes, or never put somebody in a difficult situation if that person put him in a difficult situation. People who answered “always” were classified as having a preference for reciprocity. To measure vote-buying, participants were asked if any political party had offered them money, food, payment of bills, medicine, or other goods during the run-up to the previous election. They were also asked if a party had offered to solve a non-monetary problem for them.

An additional measure of reciprocity was created from a trust game that researchers conducted with 140 participants. One individual would give a second, anonymous individual money in the hopes of receiving money in return in the next round. The correlation between the outcomes of the two measures was positive indicating that the survey could be used to measure reciprocity on a wider scale. Finally, the researchers identified and tested several possible alternative theories to determine the robustness of the relationship between reciprocity preferences and targeted vote-buying.

Results and Policy Implications

Researchers found that 33 percent of those surveyed had been offered something during the election campaign with significant variations in the total value offered as well as regional concentrations of vote-buying. There was a strong correlation between individuals’ preference for reciprocity and likelihood of being targeted for vote-buying in the 2006 election.

Those that showed a preference for reciprocal behavior were 15.4 percentage points more likely to be targeted for vote-buying and 15.9 percentage points more likely to vote for the candidate that offered them something. Politicians often targeted reciprocal individuals for vote-buying through their local political operators, often community leaders that could have known people’s preferences. The results suggest that internalized norms of reciprocity, or the unseen social pressure to treat others as they have treated you, allowed politicians to buy votes in modern election models where individual votes are secret. Politicians also target voters that are generally poorer, less educated, and attend rallies but do not strongly support a particular party. This indicates that while strong social preferences for norms like reciprocity exist vote-buying may be difficult to eliminate.