Working Paper: NBER ID: w30975
Authors: Martina Cuneo; Jetson Lederluis; Silvia Vannutelli
Abstract: Audits are a common mechanism used by governments to monitor public spending. In this paper, we discuss the effectiveness of auditing with theory and empirics. In our model, the value of audits depends on both the underlying presence of abuse and the government’s ability to observe it and enforce punishments, making auditing most effective in middling state-capacity environments. Consistent with this theory, we survey all the existing credibly causal studies and show that government audits seem to have positive effects mostly in middle-state-capacity environments like Brazil. We present new empirical evidence from American city governments, a high-capacity and low-impropriety environment. Using a previously unexplored threshold in federal audit rules and a dynamic regression discontinuity framework, we estimate the effects of these audits on American city finance and find no marginal effect of audits.
Keywords: audits; government; state capacity; public spending; accountability
JEL Codes: D73; H07; H83; M42
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| State Capacity (H11) | Effectiveness of Audits (M42) |
| Abuse (J12) | Effectiveness of Audits (M42) |
| State Capacity and Abuse (H11) | Effectiveness of Audits (M42) |
| High State Capacity (H11) | Low Effectiveness of Audits (M42) |
| Low State Capacity (H76) | Low Effectiveness of Audits (M42) |
| Middling State Capacity (H19) | High Effectiveness of Audits (M42) |