Biased Bureaucracy: Inequality and Corruption in Public Administration

by Paul Lagunes

On Election Day, the modern State treats every citizen in the same terms—we all get one vote. This political fact serves as a reminder that the principle of democracy is intimately related to another principle, that of equality.

Vote buying, electoral intimidation, and other undesired practices remain a challenge for young democracies. That said, the principle of equality before government faces its biggest challenge during election periods. On a regular basis, phrases like “who do you know?” determine the way that public servants treat citizens.

As an example of this biased treatment, Mexico’s top newspaper reported (PDF: 76 KB) the case of a woman who was mistreated and then arrested by the police. The mistreatment occurred after she took the parking place of a government employee in the state of Oaxaca. Turning to social media, people used the hashtag “Lady” or “Lord” on Twitter to highlight cases of abuse, including one where the daughter of a high-level government official had a restaurant closed after they failed to seat her.

The problem with a biased state is that it lacks the necessary autonomy to regulate the behavior of elites. Social scientists argue that State capture is a common occurrence in contexts of high inequality. That is, in contexts in which wealth is concentrated in a small number of individuals who can more easily solve their collective action problem and in which re-distributive pressures encourage the wealthy to lobby for formal and informal institutions that benefit them. Still, there is much that social science has yet to explain about the relationship between corruption and inequality.

For example, what mechanisms do the elite use to secure preferential treatment from the State? Are these mechanisms always a form of corruption? The way I see it, the elite sometimes recurs to bribes as a way to reward some public servants. However, what happens when the elite recurs to coercion or when privileges are automatically triggered in reaction to implicit biases deeply ingrained in society? Do public servants bear all the guilt when they unconsciously – and in keeping with tradition and social mores – treats the citizen with a fancy last name better than an average citizen?

There is evidence that in highly unequal societies with weak institutions, elites do not have to recur to bribes to influence state actors. Two of my collaborators and I have examined this question. My collaborators are Brian Fried and Atheendar Venkataramani.

Perusing history books, we found a key anecdote about former president Adolfo Ruiz Cortines. During his mandate, Ruiz Cortines pushed for bringing back probity to public office. Such was his drive for public integrity that once he ordered his chauffeur to make an illegal turn on a city avenue so that he could then demand police officers to apply the law and give him a ticket. The then president sought that the law was applied regardless of who had committed the infraction.

Following the example of the former president, several years ago in 2006, my collaborators and I took on the task of performing a sociological experiment in Mexico City.

For this study, we hired four car drivers, all men and in the same age group, but from different socioeconomic backgrounds, high and low. Their socioeconomic background was apparent in the way they dressed, the vehicle they drove, and other characteristics like skin color.

The four drivers committed minor, but visible traffic infractions while they drove through a random sequence of traffic crossings. We then observed if traffic officers stopped the drivers for committing the infraction, and if they did, we checked if officers issued a ticket, asked for a bribe, or simply warned them. We also interviewed a number of police officers to contextualize our findings.

Once the study was complete, we found that officers were more likely to ask for a bribe from drivers from a low socioeconomic background. This said, the size of the bribes they asked from the rich drivers was larger. For this reason, we analyzed the global results of the study and we found that officers expected to receive, on average, the same economic compensation from the bribes obtained from high class and low class drivers. So, according to our data, corruption is regressive: it puts a relatively higher burden on the poor drivers.

The interviews we carried out with police officers gave us additional insights. The majority of agents thought that richer individuals are better connected, such that in some cases, issuing them a ticket could come back and hurt them later on even if the ticket was warranted. The higher perceived cost (in terms of potential repercussions) probably explains why, in general, police officers refrain from bribing individuals with greater economic resources.

Note, however, that neither bribes nor explicit threats are necessary for better off individuals to enjoy preferential treatment from behalf of public servants. This raises important questions. We all agree that bribes are clear examples of corruption. There is no debate about this. However, can we say that only those without as many resources who pay bribes participate in corruption? If we say yes, then we are not seeing the full picture. When public servants do not comply with established rules and guidelines when dealing with privileged individuals, they are also engaging in a form of corruption. Through their negligence, they sustain a system of privileges in an environment in which democracy is said to be valued.

In another study, my co-author Oscar Pocasangre and I, took on the task of testing if there is preferential treatment in the world of access to government information. As is common knowledge, freedom of information laws have the objective of providing timely and relevant content without discriminating based on the identity of the person requesting the information.

Mexico passed its transparency law in 2002. Nonetheless, given the inequality in the country, there is a risk that such a law – just like the transit laws – are applied in a biased manner. If that is the case, those who have a privileged status would be able to obtain valuable information while those less privileged would remain ignorant.

Faced with this risk, my collaborator and I created two identities: one for a middle-class man and another for a high-class man with alleged wealth and political connections. These two characters – making use of the online system provided by the Instituto Nacional de Acceso a la Información (INAI; National Institute for Access to Information), submitted a set of questions to over 200 government entities at the federal level.

To carry out the analysis, we counted the number of requests that received a reply. We then calculated how long the entities took to provide a reply, the amount of information provided, and if entities charged a fee to provide an answer. We also identified the number of questions in the information requests that were answered and the quality of the information provided. In the end, we wanted to find a sign of preferential treatment based on the alleged economic and political influence of the person requesting the information.

It was a welcomed surprise to find that, overall, the access to information system in Mexico guarantees equal treatment, without regard to the apparent socioeconomic status of solicitors. Both identities were treated the same in terms of time to reply. Likewise, both enjoyed an equal rate of reply to their information requests.

What could explain the different findings in these studies? Why is it that in one study, police officers behaved in a biased way and in the other, bureaucrats in charge of replying to information requests exhibited equal treatment? I think the most convincing hypothesis is that in the case of information requests, the interaction between citizens and public servants is done through an impersonal medium that has many automated components. Meanwhile, for police, it is much easier to distinguish with their own eyes if a citizen is high or low class; the public servant in charge of reading and replying to information requests does not have as many tools to determine if the citizen has much clout or not.

Hence, I propose that one way to reduce favoritism in Mexico is through the automation of some bureaucratic processes and certain government actions, taking care of offering alternatives and support to those with difficulties obtaining access to technology. Hence, I propose that one way to reduce favoritism in Mexico is through the automation of some bureaucratic processes and certain government actions. The longer-term solution will require addressing the issue of inequity head-on.

Paul Lagunes is an Assistant Professor at Columbia University’s School of International and Public Affairs.


The views, opinions and positions expressed within all posts are those of the author alone and do not represent those of the Program on Corporate Compliance and Enforcement (PCCE) or of New York University School of Law.  PCCE makes no representations as to the accuracy, completeness and validity of any statements made on this site and will not be liable for any errors, omissions or representations. The copyright of this content belongs to the author and any liability with regards to infringement of intellectual property rights remains with the author.