Once Bitten Once Shy: Learning or Conservatism after Audit Failure?

I examine how audit firms implicated in large audit failures behave subsequently. Using a difference-in-difference approach, I find that after a large audit failure, the implicated audit firm shows conservative behavior in audit outcomes. I further examine whether this conservative behavior leads to more informative audit reports due to learning or to less accurate audit reports due to undue conservatism. I find that implicated audit firms are less accurate and are more likely to issue going-concern opinion to clients who do not subsequently fail suggesting a conservative bias that reduces informativeness of audit reports. Further, supporting the undue conservatism hypothesis, I find that the changes in audit outcomes reverse after a year. Exploring the sources of undue conservatism, I posit that this response could be over-reaction of the implicated audit firm due to availability bias. In support of availability bias, I find that within the implicated audit firm, the reaction is strongest for audit offices located geographically closest to the implicated office, and for auditors with clients in the same industry as the failed client. Overall, these results provide evidence that following large audit failures, the increased media and regulatory scrutiny can steer away audit firms from learning and result in unduly conservative behavior. This study also contributes to the literature documenting systematic biases in auditor behavior.