Can corruption increase public support for redistribution?

Money exchanging hands

Many ingredients are required for the design and functioning of the social contract and the stability of welfare systems. One crucial ingredient is political support for redistribution from the rich to the poor. In practice, redistributive policies are designed and executed by public officials, who ought to ensure that these redistributive policies are properly implemented. Yet being in charge of a substantial amount of resources gives public officials incentives for corruption. Corruption can take the form of favoring special interests or miss-targeting redistributive programs. Widespread corruption in the public sector diminishes the efficiency and the functioning of the welfare state.

In the Barcelona GSE Working Paper (No. 974), “Perception of Corruption and Public Support for Redistribution in Latin America” by Esther Hauk, Mónica Oviedo and Xavier Ramos, the authors examine the relationship between the citizens’ perception of corruption by public officials and their willingness to support public policy that is directed at reducing inequality. In order to provide an answer for this relationship, the authors provide a theoretical framework which then guides the empirical analysis.

Theoretical framework

From the theoretical perspective, the authors extend the model of non-collusive corruption by Foellmi and Oechslin (2007) by introducing redistributive taxation. In order to become an entrepreneur in the model, a citizen has to make a minimum investment and needs to obtain a license from a bureaucrat. But in order to receive this license, the citizen may be forced to pay a bribe. Since citizens vary in their initial wealth and capital markets are imperfect, not every citizen can borrow a sufficient amount from the capital market to pay the bribe and the minimum investment. Consequently, not every citizen can become an entrepreneur. If not an entrepreneur, the citizen lends his or her wealth in the capital market.

These assumptions translate in the developed model in such a way that perceived corruption influences the citizens’ preferences for redistribution. Perceived corruption is the probability that a bureaucrat, who sets the amount of the bribe, goes unpunished. In particular, the model identifies two channels that affect the citizens’ preferences for redistribution caused by the perception of corruption.

First, greater redistribution policies by a government can increase the exploitation opportunities by corrupt officials. This in turn means that less of the government budget is available for redistribution. Consequently, a high level of perceived corruption affects the citizens’ trust in the government as the government appears less able or willing to redistribute in an efficient way. Put simply, high levels of perceived corruption undermine the trust in government. This channel reduces the support for redistribution as corruption increases.

Second, a corrupt government causes wealth inequality to increase in the model. The reason is that bribes demanded by corrupt officials allow less poor individuals to start their own business, as they are limited in the amount they can raise from the capital market. In addition, entrepreneurs benefit by facing lower costs from renting capital on the capital market. Hence, more corruption leads to fewer rich entrepreneurs and more poor citizens that are not entrepreneurs. That in turn likely increases the taste for redistribution. In particular, any citizen with wealth below the mean wealth level is negatively affected by higher levels of corruption and hence will favor more distribution as corruption increases. Put simply, corruption reduces the relative wealth of the disadvantaged, who in turn demand more redistribution.

Empirical analysis

Which of the two channels prevails depends on the parameter assumptions in the model and is ultimately an empirical question. In order to empirically test the theory, data for 18 countries in Latin America is employed. The authors argue that Latin America for this research question is an appropriate region because it displays weak institutions with high levels of corruption, inefficient redistributive policies, high inequality levels in wealth, and credit market imperfections.

The authors estimate the effect of perceptions of corruption on the probability of agreeing with government intervention to reduce economic inequality. A simultaneous equations model is used to account for the potential endogeneity of perceived corruption: not only do beliefs about perceived corruption shape government interventions, but government interventions also influence perceived corruption.

The empirical main result suggests that perceived corruption increases the support for redistributive policies in the case of Latin America. In other words, the channel capturing the reduced relative wealth of below-average-wealth citizens is not only positive, but also outweighs the negative trust channel caused by corruption. There is evidence that corruption undermines trust, but the evidence is weaker than the positive channel. The authors’ main result is robust to various empirical specifications and measures for favoring redistribution.


Concluding and at the highest level, the paper contributes to the identification and understanding of factors for public support of a given public policy. The authors demonstrate both theoretically and empirically how corruption (the factor) shapes preferences for redistribution (the public policy) through two channels. Lastly, the authors cautiously suggest that the experience of Latin America can provide evidence on the effects that other regions or countries in Africa and Asia can anticipate when facing corruption.