Bribery inhibits private-enterprise development in emerging-market economies. Solving this problem requires a thorough understanding of how corruption works and little is currently known about the role of individual bureaucrats' incentives. In particular, a first-order question remains unanswered: how does a bureaucrat's term of office affect firms' bribery costs and, in turn, economic outcomes? Term of office is an important research focus because it may drive, not just the frequency and magnitude of bribes that firms pay, but the type of bribes they pay. On the one hand, shorter terms may limit collusive bribes by preventing bureaucrats and firms from building trustworthy relationships. On the other hand, shorter terms may limit coercive bribes by shortening bureaucrats' time horizons.
To resolve this uncertainty, Samuel Leone will study Tunisia's customs bureau leveraging detailed high-frequency data on the interactions clearing agents have with customs officers. Tunisian clearing agents work on behalf of client firms to shuttle goods through ports, and in the process they often pay customs officers both coercive bribes (e.g. to avoid red tape) and collusive bribes (e.g. to evade taxes). The study will leverage two natural experiments. First, the random matching of firms’ clearing agents with one of 32 customs officers and second, the cycling of customs officers between ports every two to three years. This creates plausibly exogenous variation in customs officers’ tenure, and also creates sampling variation in the length of relationships between customs officers and clearing agents. Through these natural experiments, the researcher will study whether customs officers affect firms’ bribery costs and trade costs, and how these officers' time horizons and professional relationships operate as mechanisms.
By shedding light on how individual bureaucrats affect bribery, this study has the potential to inform new and innovative anti-corruption strategies. In addition, the research design amounts to a policy impact evaluation of two existing anti-corruption strategies. First, the time-horizon results will speak to the efficacy of frequently rotating bureaucrats - i.e. whether it is better to have long tenures and risk collusive bribes or to have short tenures and risk coercive bribes. Second, the professional-relationship results will speak to the efficacy of e-governance reforms that minimise interactions between citizens and their government officials.
