Association for Public Policy Analysis & Management

How do we turn Nigeria into Denmark? That is, how do we shift a society characterized by low honesty and high corruption into a society characterized by high honesty and low corruption? This is the central question of anti-corruption policy, and indeed, the central question of development policy generally. But after at least two decades of research and billions of dollars spent, we still do not have a satisfactory answer.

But why have these efforts failed? Anti-corruption strategies around the world have failed because their theoretical underpinnings rest on a mistake. The theoretical foundation for most anti-corruption reforms endorsed by the international community and put in place primarily in the developing world views the problem of corruption as a principal-agent problem. The idea here is that an honest principal (e.g., a public body) contracts with an agent (e.g., a politician or a bureaucrat) who turns out to abuse their public power for personal gain. Since corruption has been viewed narrowly in these terms, the policy solutions to corruption have been focused heavily on the construction of formal rules and institutions (e.g., legal rules, penal sanctions, and so on) as opposed to informal institutions (norms, expectations, and so on). Anti-corruption policy that flows from this framework aims at increasing the costs of engaging in corruption through formal mechanisms, such as increased monitoring and punishment. Call this approach the institutionalist approach.

But in developing countries where corruption is widespread, corrupt behaviour is not the result of an individual preference to maximize self-interest in the face of formal rules. Instead, corruption is best viewed as a behaviour supported by a set of informal institutions: shared beliefs and expectations, where social preferences to engage in behaviour are dependent on the expected and actual actions of others. Costs of engaging in corruption are high if you expect nobody else to engage in corruption; Costs of not engaging in corruption are high if you expect everybody else to engage in corruption. The bottom line is that anti-corruption policy must be sensitive to both the formal and informal sets of rules that structure the incentives to engage or not engage in corruption, as well as the interaction of these formal and informal rules. When viewed in these terms, it becomes clear that corruption represents more of a collective action problem instead of a principal-agent problem. Call this approach the neo-institutionalist approach.

This is the approach defended in this paper. The novel contributions of this paper are as follows: (1) I defend a rational choice neo-institutionalist collective action approach that specifies the kind of informal institutions that drive corrupt practices; (2) I present fieldwork, interview, and survey evidence from Nigeria to provide an empirical application of the theoretical account I develop; and (3) I extend critiques of the influential principal-agent approach that build upon and compliment the general critiques found in Perrson, Rothstein, and Teorrell (2013) and Mungiu-Pippidi (2011, 2013).