Audit M arket Structure and Reform
Explaining Audit Market Structure and Reform
Audit Independence and Provision of Non-audit Services (Part 3 of 3)
DR Tan Boon Seng
his article is the final of a series of three articles 1 in this journal analysing key aspects of the audit market reforms in major jurisdictions. There are two key changes in the industrial structure of global accounting firms in the last three decades (from mid-1980s till now) – the consolidation of the global network from Big Eight to Big Four, and the diversification into professional services for tax advisory and other advisory services ranging from sustainability reporting to information technology consulting. Business diversification makes good sense to provide a one-stop shop for the client, and may possibly reduce costs through economies of scope. Furthermore, while audit revenue grows slowly, revenue from these nonaudit services (NAS) grows rapidly to become a significant contribution. However, the increasing contribution of NAS to the revenue mix inevitably raises the question: Is the auditor’s independence threatened when the incumbent audit firm takes on NAS for the client?
CONFOUNDING ISSUES Audit market reformers appear to think so, promulgating policy that limits the types of NAS that can be provided by a company’s incumbent auditor. The United States (US) Sarbanes-Oxley Act (SOX) section 201 lists nine NAS that may impair auditor’s independence and hence, are prohibited; the European Union (EU) market reform specified the types of NAS that are prohibited and further limited the NAS that are allowed. The audit firms disagree with the policy, predictably. Audit firms are organised as networks of limited liability partnerships sharing key common assets such as brand, training and research. Audit engagement for a client that operates across several jurisdictions is usually a syndicated effort of several partnerships, but appears to be supplied by just one firm. Similarly, NAS is usually not supplied by the audit partner, but by other partners within the same network. However, as all the partnerships are
Non-audit services involving management decisions, selfreview, creating self-interest and advocacy are clear-cut threats to independence for which no safeguard would be deemed adequate by outside parties
recognised by the same well-known brand name due to the successful marketing by major audit firms, the market does not perceive these partnerships as being independent even if they are different legal entities. Moreover, depending on the specific NAS, the potential threat to audit independence varies. NAS is a heterogeneous group of services, and NAS involving management decisions, self-review, creating selfinterest and advocacy are clear-cut threats to independence for which no safeguard would be deemed adequate by outside parties. For example, the provision by the incumbent auditor of these NAS, such as internal audit outsourcing, investment advisory, bookkeeping and so on, should be prohibited. Other NAS requires weighing the risk of the threat to independence against the safeguards instituted. The effectiveness of these safeguards, depending on existing internal governance of the firm and external enforcement and oversight mechanisms, varies according to the situation. As a result, the scope of restrictions on NAS imposed by each jurisdiction also varies significantly. In a report commissioned by The Institute of Chartered Accountants in England and Wales in 2002, Beattie and Fearnley2 show that there is a set of NAS prohibited across all jurisdictions, but there are wide disagreements on prohibiting the other NAS among jurisdictions. While reformers appear to be unanimous in prohibiting selected NAS and capping allowable NAS, audit firms contest the types of NAS prohibited and the rationale for the cap. Why would audit independence be compromised if the incumbent auditor provides NAS?
The analytical arguments whether providing NAS compromises auditor’s independence apply to the set of NAS that are not clear-cut threats. There is no formal theory that links the provision of NAS, which is a diverse group of services after all, to the loss of auditor’s independence, though there are two main concerns. First, auditors become beholden to management because they wish to retain the additional income from NAS. Second, auditors identify too closely with management and lose their professional scepticism. The argument of reduced independence, that is, the independence hypothesis, means that NAS should not be allowed, or at least should be limited. The counter-argument relies on the economies of scope, which refers to the proposition that joint provision of NAS with auditing has a lower cost than the separate provision of both services. Arrunada3 elaborates on two types of economies of scope – knowledge spillover and contractual economies. Knowledge spillover arises from the transfer of information and knowledge in the course of business of the two teams. Contractual economies arise from better use of resources, such as mutual trust and reputation, already developed when contracting with the client. Whether economies of scope are significant depends on the specific NAS, and whether specialised human resource is deployed. If there are significant economies of scope, the sensible policy is not to prohibit or limit NAS. Both analytical arguments appear plausible and therefore pose a difficult policy decision. We next examine empirical evidence given the analytical context.
... as all the partnerships are recognised by the same well-known brand name due to the successful marketing by major audit firms, the market does not perceive these partnerships as being independent even if they are different legal entities
- Auditor’s Independence The empirical literature in the NASindependence relation can be divided into two streams. The older stream examines perceived independence, that is, independence-in-appearance, through mostly surveys and stock price reaction. The newer stream examines real independence, that is, independence-in-fact, using audit quality as proxy. Both aspects are required in the concept of auditor’s independence in the ethics code. Generally, survey studies and studies of market reaction to unexpected earnings news both show that providing NAS reduces perceived independence. There is good consensus for reduced perceived independence. Evidence on impairment of auditor’s real independence, which infringes ethical codes and regulations, is very hard to obtain. The study by Frankel, Johnson, and Nelson4 shows that high NAS prices reduce real independence. However, subsequent studies, such as by Ashbaugh, LaFond and Mayhew5 , show that this result is not robust and is sensitive to the sample and measure of audit quality used. Hence, there is no consensus that provision of NAS impairs real independence.
- Economies of Scope There is mixed empirical evidence on the economies of scope for the incumbent auditor to provide NAS. Starting from the seminal study by Simunic6, a series of papers show that economies of scope exist between audit services and NAS. However, a sophisticated study by Whisenant, Sankaraguruswamy and Raghunandan7 casts doubt on the earlier results. The study produces evidence for economies of scope if a statistical bias (called simultaneity bias) is not corrected, but the effect disappears if the bias is corrected. Hence, there is no current consensus whether economies of scope are significant. Empirical research for economies of scope will be increasingly difficult after SOX. This is because the data are difficult to obtain due to the prohibition of the incumbent auditor supplying most NAS. The definitive test for economies of scope is by examining the cost, but this requires proprietary cost data and being specific about the type of NAS.
CRUSHED BY THE BURDEN OF PROOF
This article discusses several issues in the debate on whether restricting the provision of NAS by the incumbent auditor is an optimal policy. We summarise the following four points before coming to a conclusion:
- NAS are not built equal in terms of threat to auditor’s independence. There is a consensus that the incumbent auditor should be prohibited from providing certain NAS. There is controversy over the prohibition outside this group of NAS.
- There is consensus that the incumbent auditor supplying NAS would lower perceived independence. This consensus is backed by empirical evidence.
- Real independence is almost impossible to prove. Using audit quality as proxy, the empirical evidence that provision of NAS lowers real independence is mixed.
- The evidence for economies of scope is mixed, and it appears that obtaining research data will be increasingly difficult after SOX. Policy choice is often made using a cost-benefit framework.
Evaluating the cost and benefit of the policy to restrict NAS should consider all the four points as a whole. The cost of a restrictive policy is the opportunity cost associated with foregoing potential economies of scope. The real benefit of the restrictive policy is the maintenance of real independence, and the apparent benefit is the maintenance of perceived independence. Empirical evidence for the cost and real benefit of the restrictive policy are both inconclusive. Many jurisdictions appear to have turned to the apparent benefit, which has empirical consensus, as the tie-breaker to favour a restrictive policy. Other jurisdictions, such as Singapore, choose to delegate the decision to the audit committees and regulate the auditors by incorporating a professional ethics code in the regulation for auditors.
Dr Tan Boon Seng is Assistant Director, Technical Research, ISCA.
Share this Page