The Calculus of Vote-Selling: Electoral Trust and the Value of a Vote

MPSABlog_Rau_ElectoralTrustThe following is part of a series of posts written by 2016 MPSA award recipients highlighting outstanding research presented at previous MPSA annual conferences.

Existing work has established that clientelistic parties can maximize their returns by targeting poor voters, because their votes are “cheaper” in the sense that the value of a gift is elastic and the diminishing marginal utility of income implies that a gift of a given monetary value will be worth more to a poorer voter. I examine the other side of this cost-benefit analysis, arguing that the value of the vote is elastic as well, hypothesizing that the extent of citizens’ trust in elections affects their perception of the value of a vote and thereby their reservation prices.

I consider two main interpretations of trust in elections, which respectively entail the following:

  1. Fraud-based: I believe that the outcome of this election reflects the votes cast.
  2. Value-based: I believe that it is legitimate and effective to distribute political power based on the results of an election.

According to either interpretation, higher levels of trust should be associated with higher value attached to one’s vote. If voters believe that elections are clean, then the belief that their votes will be counted accurately should cause voters to attribute greater value to their votes. A vote’s instrumental value comes from the ability it confers on the voter to express preferences and hold the government accountable. If votes are not counted, then voting no longer confers these benefits to the voter. A voter who believes that elections are wrought with fraud, then, has fewer reasons to value the vote and should consequently have a lower reservation price. In accounting for the value-based trust of elections, if a voter believes that the results of elections, based on the popular vote, are inherently valuable and preferable, this creates an additional disincentive to sell the vote.

I employed the Latin American Public Opinion Project’s AmericasBarometer survey in Argentina to test this theory of trust and vote-selling. As seen in Table 5.1, reported trust in elections significantly affects rates of vote-selling, but the effect runs in the opposite direction of what is expected, with those expressing higher levels of trust in elections also reporting selling their votes at higher rates.

Table 5.1: “Value of Elections” comes from a question asking respondents to weigh the relative importance of having a strong leader versus a democratically elected leader; “Intent to Leave” indicates intent to emigrate from the country within the next 3 years; “Sought Assistance” indicates having sought assistance from a governmental office/official within the last year.


I conclude that these results are likely a result of measurement problems: trust is measured by inquiring to what extent the respondent trusts X. The lack of an object in standard survey questions about trust (an alternative taking the form of “to what extent do you trust X to do Y?”) likely causes the questions to miss their mark. Such questions may elicit something more akin to an approval rating, due to the vagueness with regards to what trust in a given institution would entail. In that case, the positive relationship in the data would make theoretical sense – if the existence of clientelism is known to all, then those who report participation in such exchanges are more likely to approve than those who report having refused such offers.

The counterintuitive finding of this paper highlights the need for more nuanced and specific survey questions about trust in order to accurately examine its role in political behavior. In addition to reducing measurement error by articulating more specific questions, the introduction of a series of questions on trust would allow for a deeper understanding of how trust functions in the valuation of citizens’ votes. Is the belief in widespread fraud the source of the link between trust and the value of the vote? Does this relationship relate to the sense of a democratic duty to express one’s preferences through voting? Do those who believe that vote buying is widespread and pervasive, as opposed to fraud, trust elections less as a result? Such questions require new measures of trust, and their answers speak not only to the individual decision-making process of voters targeted by brokers in clientelistic systems, but also to the more general role of institutional trust in voting behavior.


About the Author:
 Eli Rau has just completed his first year as a PhD student in the Department of Political Science at Yale University. His research “The Calculus of Vote-Selling: Electoral Trust and the Value of a Vote” was awarded the Best Undergraduate Poster Award at the 2016 MPSA Conference.