Publication Type

Working Paper

Publication Date



Prior studies that examine the association between excess audit fees and audit quality (measured by discretionary accruals) using US data fail to document a significant association. However, there is no convincing explanation for this result to date. In this study, we test a cancellation hypothesis, which suggests that the non-association might be due to the cancellation effect between the risk-effort effect and the bonding effect that excess fees capture. To the extent that excess fees capture the compensation for effort the auditor puts in for a risky client, which is not captured in the existing audit fee models (the risk-effort hypothesis), one would expect a negative relation between excess audit fees and discretionary accruals (measure of audit quality). However, if it reflects bonding between the client and the auditor, one would expect a positive relation between excess audit fees and discretionary accruals. In view of Ball (2001), we hypothesize that the bonding effect is likely to be stronger in US, where the capital market is most developed in the world, than in UK. However, we expect the risk-effort effect to be present in both countries. Thus, the cancellation effect from the two forces should be more pronounced in US. Consistent with this idea, we confirm the non-association between excess fees and audit quality in US but document a negative association in UK. Further analysis suggests that analyst monitoring mitigates the fee-bonding between the auditor and the client in US. Taken together, this study proposes the cancellation effect as a possible explanation for the lack of association between excess audit fees and audit quality in US.


Business Law, Public Responsibility, and Ethics | Corporate Finance | Portfolio and Security Analysis