The Effects of Recurring and Nonrecurring Tax, Audit-Related, and Other Nonaudit Services on Auditor Independence* JEFFREY S. PATERSON, Florida State University ADRIAN VALENCIA, Florida Gulf Coast University 1. Introduction For more than 30 years, legislatures and regulators have debated the effects of the joint provision of audit and nonaudit services (NAS) on auditor independence (U.S. Senate 1976; American Institute of Certified Public Accountants 1978; Levitt 2000; U.S. House of Representatives 2002; Public Company Accounting Oversight Board [PCAOB] 2005). Researchers have investigated the issue analytically and empirically. While refinements in theory and methodology have contributed to our understanding, questions about the rela- tionship between NAS and auditor independence remain. An important question is if the relationship is affected by whether auditor-provided NAS are recurring or nonrecurring. DeAngelo’s (1981) model of audit pricing describes incumbent auditors as possessing comparative cost advantages associated with audit startup and switching costs. She sug- gests that these cost advantages help incumbent auditors earn quasi-rents in future years. The intertemporal nature of this relationship results in an increase in auditor–client eco- nomic bonding and a potential decrease in auditor independence. Simunic (1984) and Beck, Frecka, and Solomon (1988a) incorporate auditor-provided NAS into the auditor– client relationship. 1 The joint provision of services increases the economic bond described by DeAngelo 1981 and potentially further impairs independence. The joint provision of NAS can also yield knowledge spillovers that generate audit cost savings and affect audi- tor independence (Simunic 1984; Beck et al. 1988a; Public Oversight Board 2000; Wu 2006). Simunic (1984) treats NAS as homogeneous in their ability to generate knowledge spillovers, but he notes that additional insights might be gained by distinguishing specific types of NAS. More recently, the Securities and Exchange Commission (SEC) began requiring public disclosure of auditor-provided NAS fees partitioned by the type of ser- vice, suggesting that such information would be useful (SEC 2000, 2003). Others suggest that it is important to distinguish NAS engagements based on whether they are recurring or nonrecurring (Beck et al. 1988a; Schneider, Church, and Ely 2006). The public disclosure of fee data led to a surge in the number of studies that examine the effect of NAS on auditor independence. Despite the SEC’s mandate that NAS fees be reported by type, most of these studies rely on single-period total NAS fee data (e.g., Frankel, Johnson, and Nelson 2002; Ashbaugh, LaFond, and Mayhew 2003; Chung and Kallapur 2003; DeFond, Raghunandan, and Subramanyam 2002; Raghunandan, Read, and Whisenant 2003; Reynolds, Deis, and Francis 2004). Research using an experimental markets methodology uses a similar approach, seeding their instruments with single-period * Accepted by K.R. Subramanyam. We gratefully acknowledge the comments of Allen Blay, Jared Delisle, Jennifer Gaver, Ayalew Lulsejed, Rick Morton, Dahlia Robinson, seminar participants at Florida State University, the 2008 KPMG Ph.D. Project Conference, and the 2010 Auditing Section Mid-Year Confer- ence. We also thank two anonymous reviewers, K.R. Subramanyam (the associate editor), and Steve Salte- rio (the editor (in chief)). 1. Some authors, including Simunic 1984 and Beck et al. 1988a, refer to management advisory services (MAS) rather than nonaudit services. We treat the two terms (MAS and NAS) as interchangeable. Contemporary Accounting Research Vol. 28 No. 5 (Winter 2011) pp. 1510–1536 Ó CAAA doi:10.1111/j.1911-3846.2010.01060.x