Abstract Corruption and anti-corruption efforts are intertwined with political and economic concerns. From an economic and political perspective, as the government strives to enhance its governance capabilities, it becomes crucial to consider the costs of anti-corruption supervision and the losses incurred from corruption. This evaluation is essential for formulating a scientifically sound anti-corruption strategy that maximizes government benefits. To address this issue, the paper presents a model that incorporates levels of supervision and associated costs. The findings reveal that in the case of homogeneous officials, the optimal level of supervisory input, which maximizes government benefit, is nearly zero when per capita income is low due to budgetary constraints on anti-corruption control. However, as per capita income reaches a certain threshold, the optimal level of supervisory input suddenly rises to its maximum and decreases as per capita income increases. Alternatively, if the government adopts a zero-tolerance approach towards corruption and provides adequate remuneration to its employees, ensuring that all competent authorities can resist corruption, then corruption can be eliminated. Moreover, when officials exhibit heterogeneity in terms of their honesty levels, certain conditions give rise to a middle per capita income range, resulting in an optimal level of supervisory input that leads to a phenomenon known as “partial corruption.” During this phase, the wages paid by the government to its employees promote honesty, preventing them from engaging in corruption. However, they are unable to curb the corrupt activities of more dishonest officials. To some extent, this model also explains the challenges associated with eradicating corruption in several middle-income countries.
Zhang et al. (Thu,) studied this question.