Power Prospects: A New Look at Bureaucratic Corruptions

Description

Corruption represents a critical social challenge, affecting both developing and developed societies and necessitating sustained attention. The significance of providing sufficient compensation to ensure an ethical civil service is broadly acknowledged in policy discussions (Myrdal, 1968; Carino,1986; Israel, 1987; Lindauer and Nunberg, 1994, among many others). According to Rahman (1986), "the (Singapore) government believed that an efficient bureaucratic system is one in which the officers are well-paid so the temptation to resort to bribes would be reduced." This insight has been echoed by the earlier theoretical work (Becker and Stigler, 1974; Shapiro and Stiglitz, 1984; Tanzi, 1994). This project is motivated by two limitations in this line of argumentation within the existing literature. First, later empirical findings linking higher salaries to reduced corruption is inconclusive (Rauch and Evans, 2000; Treisman, 2000; Van Rijckeghem and Weder, 2001; Gorodnichenko and Peter, 2007), also, these cross-country studies do not address a causal relationship (Di Tella and Schargrodsky, 2003; Swensson, 2005). Second, the literature cannot address a novel observation we find, that the bureaucrats, when contemplating corruption, prioritize more of future income growth, i.e., the slope of earnings’ profile, rather than the current or cumulative income, i.e., the intercept. The primary objective of this project is to gain new understandings of corruption, particularly in terms of how shifts in income prospects influence current corrupt practices. This understanding will guide improvements in anti-corruption mechanisms. Our project consists of four parts: Part 1. Causal Identification of how various aspects of the remuneration package, including income growth rate, promotion probability, and income level, affect corruption. Leveraging a comprehensive corruption conviction dataset from China, obtained from the online platform of court documents, we will conduct causal inferences using difference-in-difference model and instrumental variable estimation. Part 2. We will develop a novel, dynamic principal-agent model with hidden information and hidden actions where the agent privately knows his discount factor for future income flows and privately accepts bribes if any. Assuming that all types of agents are susceptible to corruption, the (benevolent) principal designs an optimal dynamic mechanism that minimizes the potential costs of corruption under the constraints of hidden information and hidden actions. We will show that agents with lower discount factors tend to be more corruptible given any payment scheme. In other words, agents who have a YOLO (or you-live-only-once) type or those who are approaching replacement or retirement are more willing to accept bribes in our model. We will show that the optimal mechanism entails a set of balanced promotion/monitoring/punishment rules that feature high salary growth, moderate monitoring, and relatively severe punishment for corruption. Part 3. Development of a structural model capturing officials' corruption decisions over their career paths. We will estimate this model and obtain the empirical distribution of the unobservables. Part 4. Drawing policy implications through a quantitative investigation into the anticorruption mechanism design. Using the structural model, we will conduct counterfactual analysis to explore how adjusting the slope of earnings profiles could reduce corruption and monitoring costs. Additionally, we will analyze how selection mechanisms can preemptively minimize corruption by recruiting individuals less susceptible to corrupt behavior into public office. Anticipated outcomes of this research project are twofold, impacting both academic discourse and policy formulation. Academically, our work advances understandings about corruption, stimulating further research and discourse in this vital area. In terms of policy, our findings will offer valuable insights for designing more corruption-resistant yet detail-free (feasible) mechanisms. For instance, we suggest a steeper compensation curve with a lower starting salary as a more impactful approach than a simple increase in flat compensation. Moreover, we propose including a question regarding time preference in the selection examination as a beneficial addition.