Working Paper: CEPR ID: DP5252
Authors: Jennifer Hunt
Abstract: Using detailed Peruvian data measuring bribery, I assess which types of public official are most corrupt and why. I distinguish between the bribery rate and the size of bribes received, and seek to explain the variation in each across public institutions. The characteristics of officials' clients explain most of the variation for bribery rates, but none for bribe amounts. A measure of the speed of honest service at the institution explains much of the remaining variation for both bribery rates and amounts. The results indicate that the bribery rate is higher at institutions with bribe-prone clients, and that bribery rates and bribe amounts are higher where clients are frustrated at slow service. Faster and better service would reduce corruption. Overall, the judiciary and the police are by far the most corrupt institutions.
Keywords: corruption; governance
JEL Codes: H4; K4; O1
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| client frustration with slow service (L86) | bribery rates (H57) |
| adjusted conclusion rate (C29) | bribery rates (H57) |
| client characteristics (I11) | service efficiency (D61) |
| judiciary conclusion rate (K41) | bribery rates (H57) |
| judiciary conclusion rate (K41) | bribery amounts (H57) |
| police conclusion rate (K42) | bribery rates (H57) |
| client characteristics (I11) | bribery rates (H57) |
| service efficiency (D61) | bribery rates (H57) |