Accounting and Finance 48 (2008) 259–278 © The Authors Journal compilation © 2007 AFAANZ Blackwell Publishing Ltd Oxford, UK ACFI Accounting and Finance 0810-5391 © The Authors Journal compilation © 2007 AFAANZ XXX Original Articles R. J. Iyengar, E. M. Zampelli/Accounting and Finance 47 (2007) xxx–xxx R. J. Iyengar, E. M. Zampelli Auditor independence, executive pay and firm performance Raghavan J. Iyengar a , Ernest M. Zampelli b a School of Business, North Carolina Central University, Durham, 27707, USA b Department of Business and Economics, Catholic University of America, Washington, DC, 20064, USA Abstract This paper investigates whether compensation committees actively intervene to adjust accounting performance-based incentive schemes for the real, or perceived, reduced earnings credibility signalled by the purchase of non-audit services. Using a nonlinear, two-stage least-squares method that accounts for the simul- taneity of executive pay, firm performance and non-audit fees, we find a significant negative relationship between non-audit fees and the sensitivity of chief executive officer (CEO) pay to firm performance. Point estimates suggest that the reduced weight applied to accounting performance lowers the incentive component of executive pay between roughly 5 and 8 per cent for the CEO of the ‘average firm’. Key words: Executive compensation; non-audit fees; auditor independence JEL classification: M41, M42, M52 doi: 10.1111/j.1467-629x.2007.00226.x 1. Introduction In their seminal article, Jensen and Meckling (1976) demonstrate the role of the external audit as a monitoring device effective in reducing agency costs. Of course, this is predicated on the independence of the auditing firm from the The authors thank Robert Faff (Editor) and two anonymous referees for their helpful comments and suggestions. Thanks are also extended to Kofi Amoateng, Samuel Kotz, ABM Nasir, Terry Shevlin, participants at the Center for Corporate Reporting and Governance Conference, American Accounting Association (AAA) National Conference, Western Regional AAA Conference, Center for Corporate Reporting and Governance Conference, North Carolina Central University’s Statistical Modelling Seminar and Lunch Seminar Series, for their helpful comments on earlier drafts. The research assistance of Angela Ragin is gratefully acknowledged. Partial funding for this study was provided by a summer research grant from North Carolina Central University. Received 14 April 2005; accepted 5 March 2007 by Robert Faff (Editor).