Why most attempts to fight corruption fail

Corruption is the prototypical wicked problem, a moving target whose end is probably a utopia. The World Bank considers it a major challenge to ending extreme poverty and boosting prosperity in the poorest parts across the globe. However, corruption is like the flu, affecting all societies regardless of their affluence.

High quality institutions, of course, may help in mitigating the problem. Policy makers, on the other hand, often fail to address corruption’s root causes, designing instead interventions that only act on symptoms. Common treatments to corruption include strengthening institutions or, following the archetypical Gary Becker model of criminal behavior, increasing the certainty of punishment.

Based on my academic work modeling corruption through computational methods, I will highlight the shortcomings of such approaches, using my country (Brazil) as an example.

First, consider that corruption flourishes in points of social systems (including organizations) where ill designed incentives and inadequate governance mechanisms produce what I call pockets of temptation. Temptation accumulates when there are potentially large private benefits opaque to social monitoring. The opacity is often compounded by concentration of power in the hands of few people.

A crucial feature of human mind associated with corruption, according to scientific evidence, is the ease of rationalization. Life is full of gray areas and human beings can easily concoct comforting narratives. Researchers that have studied unethical behaviors converge on a clear point: the distance separating most people from evil is not an abyss, but something like a few steps.

In the words of Oscar Wilde, we can resist anything but temptations. It is unsurprising that pockets of temptation flourish in places such as purchasing departments in businesses and tax administrations agencies throughout the world.

Three levels. The reason most programs to combat corruption fail is their incapacity to act on the proper level of the phenomenon. My models suggest three levels.

The first one is where pockets of temptation may actually sprout. In any system, the constellation of financial and social incentives has the potential to create corruption. In fact, all systems are bound to be gamed by their agents, a majority of whom is always looking for opportunities for personal gain.

In the second level, some agents have already exploited the system, but corruption is concentrated only in few people. Often, it is a slippery road process, as opportunity for personal gains arise and personal standards are progressively relaxed.

The third level occurs when agents, often acting through networks, strive to control the system. This frequently leads to the creation of entire ecosystems of corruption. Once a system is under control, such networks cling to power channels and create mechanisms of replication. Cultures of corruption develop, normalizing deviant behaviors. Beliefs like “things are what they are, we’re not going to change the world” become part of the modus operandi, being quickly absorbed by novices. The fundamental human need to belong is too strong; no one wants to be a pariah.

Usual anti-corruption programs tend to aim at levels two and three of the phenomenon. Nonetheless, coping effectively with the problem requires redesigning our social systems at level one, where pockets of temptation sprout. Let us look at some examples in Brazil.

The country’s 1988 Constitution stimulated the creation of financially unviable municipalities. Any district of a previous city could emancipate, no matter its actual economic sustainability – municipalities in Brazil depend on taxes levied on property and services.

The result was an explosion of new cities, most of which now depends on permanent “allowances” or transfer of federal funds, all subject to lax oversight. This is the perfect recipe for corruption and evidence in this context has not been disappointing.

Another example comes from the political system. It has a dysfunctional structure, where many parties have a group of “owners” or caciques, as we call them in Portuguese. Caciques often accumulate large amounts of political capital, which they trade for influence and control over large chunks of the governmental machinery. They also act as gatekeepers for businesses wanting undue benefits from the government, reinforcing the crony capitalism still prevalent in the country.

Finally, while there are calls for a national policy to decrease red tape in Brazilian government, the underlying problem is the public management model, which is focused on control and bureaucracy, but not on results.

The model rewards weak performance while punishing harshly unintentional mistakes, which management science tells are the basis for innovation and change. The consequence is a culture of fear among public servants and the mass production of difficult-to-comply regulation, creating pockets of temptation to circumvent the profusion of norms.

In sum, effectively tackling corruption requires much more than simply punishing deviant behavior and increasing oversight. Unlike the flu, it requires open surgery in hidden societal structures.