The demand determinants of non-audit services: an empirical analysis with implications for auditor independence

© 2001 Christopher C. IkinAuditor independence is being increasingly questioned in a public debate about the joint provision of audit and non-audit services (NAS) by auditors to their corporate clients. Regulators are concerned that auditor firms which provide material consultancy services to their audit clients are likely to have their independence impaired either through conflicts of interest or fee dependence or both. Public accounting firms (the auditors) deny these risks. Positive accounting theories, such as agency theory and costly contracting theory, have supported the regulators' concerns by arguing that economic bonding between auditor and client is likely to be increased by the provision of NAS by the incumbent auditor, and hence lead to an impairment (real or perceived) of auditor independence. However, academic research into the nature of the association between auditor independence and the provision of non-audit services has been inconclusive and has failed to provide a definitive resolution to the debate. A reliable, comprehensive NAS fee model is a necessary prequel to greater insights into this debate. Guided by Firth (1997) and Craswell, Guz and Francis (2000), this study hypothesises a model reflecting three demand determinants for NAS: (i) the company's ex ante need for NAS, (ii) the auditor's NAS industry expertise and (iii) the company's willingness to appoint the auditor (given a portfolio of endogenous corporate governance practices and given a set of exogenous agency cost and political cost circumstances). The data set of 432 companies is selected from the "Top 500" Australian companies listed on the ASX in 1997. OLS regression analysis strongly confirms the model's first two demand determinants. First, a company's complexity, the extent of its restructuring activities, its appointment of a new CEO and its poor performance are all shown to be significant dimensions of its "need for NAS". Secondly, the important role of a NAS industry expert is a significant new finding in the NAS fee literature. The third determinant, the "willingness to appoint" is not comprehensively tested. Nevertheless some auditor independence issues associated with the "willingness to appoint" are explored. Auditor independence appears to be of concern to companies exposed to high political costs while it does not seem to be of concern to companies facing high agency costs. However, this study argues that auditor independence needs to be considered as part of a portfolio of other corporate governance practices/positions, which taken together, influence a company's willingness to appoint an incumbent auditor to the NAS consultancy. The demand determinants for NAS are also found to be robust across various partitionings of the data and in a novel re-specification of the model focussing on periodic increases in the level of NAS fees. Some limitations are acknowledged but exciting prospects for further fruitful research are recommended