Impact of Third Party Audits on Compliance and Corruption: Evidence Using a Policy Experiment

Using a policy experiment in India, we examine the impact of a third party audit on tax compliance. Employing the generalized difference-in-difference methodology, we find that growth in tax payment is 58 percentage points higher under a regime with no third-party audit. The fact that the differential growth in tax is negatively associated with differential growth in discretionary expenses and also the absence of any real growth of treated firms, suggestively indicate collusion and corruption. Our results hold up-to a battery of placebo tests, absorption of time-varying and time-invariant unobservable factors and, display heterogeneity based on expected lines.