Pricing Initial Audit Engagements: Empirical Evidence Following Public Disclosure of Audit Fees

Abstract:     
Theoretical models that describe the auditor-client environment offer conflicting inferences about initial audit pricing and quality. Although not in agreement on the source of initial audit discounts, DeAngelo (1981) and Kanodia and Mukherji (1994) model audit pricing that indicates initial audits are discounted and the discount is insufficient to erode audit quality. Both studies suggest that incentives to discount initial audits would be unaffected by public disclosure of audit fees. Dye (1991), however, models a setting in which investors would perceive initial audit discounting as creating an economic bond that threatens auditor independence if audit fees are publicly disclosed. He argues that public disclosure of audit fees will eliminate the practice of discounting initial audits.

To test the competing theories of initial audit pricing, we investigate data from 661 initial and 22,117 continuing audits in the first eight years of required public disclosure of audit fees in U. S. public company audit market. We show that initial audit discounting continues even after the public disclosure of audit fees indicating initial audit pricing in the U.S. audit market is more consistent with initial audit engagement pricing theory offered by either DeAngelo (1981) or Kanodia and Mukherji (1994) than with the public disclosure argument in Dye (1991). We also find no evidence that audit quality is related to the initial audit discount. Additionally, our empirical evidence of initial audit discounting when audit fees are publicly-disclosed is in contrast with empirical evidence from the Australian audit market where audit fees are publicly disclosed and initial audits of same-tier auditor changes have been shown to not be priced at a discount (Craswell and Francis 1999). We leave to future research the question of why audit markets in the U.S. and Australia have different initial audit pricing practices. We perform various specification tests to ensure our results are not due to certain research design choices and find robust results to two measures of initial audit discounts.

Note: Previously titled "Pricing Initial Audit Engagements in U.S. Market: A Test of Competing Theories"

Number of Pages in PDF File: 33

Keywords: audit pricing, non-audit pricing, lowballing, quasi-rents, audit quality, earnings quality

JEL Classification: M14, M49, D40, M47

working papers series