Non-collusive corruption: theory and evidence from education sector in Bangladesh

We study non-collusive corruption in the education sector. For this purpose, we construct a simple theoretical model that captures non-collusive corruption between service providers (teachers) and service demanders (students). The model shows that the bribe paid by the service demander increases with the level of red tape and the income level, but it decrease with the improvement of the individual's social status. We also establish that with the increase in the income and the social status of the private agent (networks), the probability of paying bribes and the severity of red tape declines. Then we use a survey data set collected in 2007 by Transparency International Bangladesh, to test the predictions of the model. The estimations confirm that both the probability of being subjected to noncollusivecorruption and the cost of corruption is related to the individual characteristics of the bribe payer. Moreover, network connections are an important factor that helps to ease the burden of corruption on private agents, which is also likely ensuring the persistence of this type of corruption.