Why Do People Forgive Corrupt Politicians?

The case of Hilario Ramírez, a Mexican politician who, amid a reelection campaign admitted to having stolen from the treasury, “just a bit,” he argued, is more than a simple piece of Latin American political folklore. “Layín”—as Ramírez is also known—was mayor of San Blas, a costal Mexican city, from 2008 to 2011 and a member of the conservative PAN (National Action Party). In 2014, he successfully ran for re-election, this time as an independent candidate. During that campaign he made the statement that made him famous, explaining that “what he stole with one hand, he gave to the poor with the other” and that if he had taken 150 million pesos, around 7 million dollars, as his rivals accused him, “he would have invested that money in public infrastructure.”

Layín’s electoral success goes against one of the most shared ideas about corruption: that when citizens are aware of government corruption, they punish the wrongdoers and their parties through voting. However, he is by no means an isolated case. The paradoxical fact that corruption is unpopular while some corrupt politicians are quite popular has long intrigued analysts, as noted by Oskar Kurer. One possible explanation could be a lack of information. But what about when voters have the information and vote for such politicians anyway—willingly “forgiving” these corrupt incumbents at the ballot box? Re-electing such a politician while knowing about his misdeeds turns upside down much of what we understand about electoral accountability. Research from political scientists can give us an answer to this puzzle.

Contrary to common assumptions, there is plenty of evidence that proves corruption has very modest electoral consequences. Even in contexts where there is enough information to discern who is honest and who is not, corruption is rarely punished at the ballot box (or, at least, not severely). According to research from 2015, United States congressmen involved in corruption scandals only lose between five and ten percent of their electoral support. In the United Kingdom, the vote loss is even smaller, around 1.5 percentage points. In Italy, the probability of a legislator being re-elected is practically the same whether he is linked to corruption or not. In Japan, where more than half of the legislators accused or convicted of corruption since the forties have been re-elected, some have even increased their vote share after scandals. Is it possible that voters are rational when they choose to “electorally pardon” corrupt politicians? An investigation carried out by Universitat Autònoma de Barcelona (UAB) scholars Jordi Muñoz, Eva Anduiza and Aina Gallego tested three hypotheses to explain why this is happening.

The first hypothesis this group of political scientists test is what they call “cynicism,” which suggests that voters assume that all politicians are corrupt. For this reason, even if corruption is considered reprehensible, it will not be a reason to change the direction of the vote: it is taken for granted. The second hypothesis is that of the “noise,” which has to do with voters’ interpretation of corruption accusations, very much in tune with the current climate of political polarization, the importance of negative partisanship and the epidemic of fake news. In these contexts, corruption scandals can be seen as part of a political strategy with the sole purpose of defaming rivals. This assumption brings into question the credibility of the accusations and hinders the electoral consequences of the corruption accusations. The third hypothesis proposes the existence of an “implicit exchange” mechanism, by which voters may consider that the utility generated by corrupt politicians offsets the costs of their behavior, something that would make corruption tolerable. The campaign slogan of a 20th century Brazilian politician, Ademar Pereira de Barros, sums it up well, ““Ademar rouba, mas faz”, Ademar steals, but gets things done.

To study which hypothesis is most common, UAB researchers conducted a survey experiment in which they showed over 1,000 people different scenarios involving municipal corruption, with each scenario relating to one of their hypotheses. Each participant was asked: “If the case described above referred to the municipality where you live, what would be the probability that you would vote for this mayor?”

Muñoz, Anduiza and Gallego found that “implicit exchange” is the most relevant mechanism to explain the limited electoral effects of corruption. A good administration record, they argue, increases the probability of voting for a corrupt candidate by 14 percentage points. They conclude that when a citizen considers that corrupt politicians generate benefits for her, she is likely to “forgive” the politician’s misdeeds in the following election, even if she knows about these misdeeds and finds them reprehensible.

We can link these findings on the importance of “implicit exchanges” to the work of Gonzalo Rivero, Pablo Fernández and Pablo Barberá, who found that, when punishing corrupt politicians, the type of corruption matters. After analyzing the case of Spain, a country that suffered a lot of real estate corruption during the first decade of the 2000s, these scholars found that corruption cases that generate “positive externalities” are less likely to be punished through voting. An increase in employment opportunities or in the demand for local goods or services when irregular construction permits are given would be an example of this kind of externality. However, more “predatory” corruption cases, such as fraud or embezzlement, tend to be punished more severely.

The importance of exchange mechanisms in explaining the electoral consequences of corruption has serious implications for our understanding of politics. An important point that these investigations make is that corrupt politicians’ success is not the result of cultural deficiencies or lack of democratic culture. Moreover, it could be the case that a corrupt politician’s success is not a problem of information asymmetries, either. On the contrary, this phenomenon may well exist because of the rationality of the voter. A crude rationality that we may not like, but that will not disappear by ignoring it.

The challenges posed by these findings are significant for those interested in promoting public ethics and raising the quality of democracies: if lack of information is not the main reason why citizens re-elect corrupt politicians at the ballot box, greater transparency will not be a magic solution to ensure elections really serve to “throw the rascals out,” as Carroll Quigley famously claimed.

Despite this, there is room for hope. Layín’s rising career came to an end when he was caught inappropriately touching a woman and tried to defend himself on TV. He later ran for governor of his state and lost. Soon afterwards, he was accused of embezzlement and is now a federal fugitive. Perhaps some corrupt deeds do not matter at the ballot box, but there are others misconducts that are unforgivable.