«Auditor Independence and Non-Audit Services: A Literature Review Vivien Beattie University of Stirling and Stella Fearnley University of Portsmouth ...»
2.6.1 Experimental studies Ponemon and Gabhart (1990) use Kohlberg’s (1969) stage model of moral development and ethical cognition to examine an auditor’s implicit reasoning in the resolution of an independence conflict. This well-validated model distinguishes three levels of ethical
• pre-conventional – where the individual places self-interest well above the common interests of society and is sensitive to penalty attributes;
• conventional – where the individual conforms to the rules of society and is sensitive to affiliation attributes; and
• post-conventional – where the individual forms a judgment conforming to ethical principles and not to society’s rules.
The findings of an experimental study using 119 audit partners and managers show that a systematic relationship between auditors’ measured ethical cognition and hypothetical and audit conflict scenarios and their resolution of an independence conflict exists.4 They also found that independence judgments are significantly influenced by factors relating to penalty and are less sensitive to affiliation factors (i.e., living up to what is expected by people). There was also a systematic relationship between ethical cognition and auditors’ priority rankings of factors influencing auditor independence. In particular, subjects at the pre-conventional level of ethical cognition ranked freedom from pressure to retain client and existence of legal liability significantly higher than subjects at the conventional level.
Windsor and Ashkanasy (1995) extend Ponemon and Gabhart’s (1990) study by including economic and personal belief variables, in particular client management bargaining power and belief in a just world, in addition to the level of moral reasoning
development. Three styles of auditor decision-making emerged:
• autonomous – auditors who were responsive to personal beliefs and were more likely to resist client management power;
• accommodating – auditors who responded to both personal beliefs and client management power and who were least resistant to client management pressure;
• pragmatic – auditors who were responsive to client management power, irrespective of beliefs.
These three styles correspond to individuals with high, mid, and low levels of moral reasoning, respectively.
Other experimental studies carried out with auditors also show that the level of ethical cognition of the individual influences decision making (Ponemon, 1990, 1993).
The contingent influence of organisational culture, i.e. the moral atmosphere of the audit firm, is also being explored by researchers, although no clear results have yet emerged (Ashkanasy and Windsor, 1997; Sweeney and Roberts, 1997).
2.7 Broader based studies into auditor decision making In recent years, a number of writers have attempted to develop a broadly-based framework to explain auditors’ decisions. These frameworks covering economic, regulatory, behavioural and ethical factors.
This basic result is confirmed in a later study by Sweeney and Roberts (1997).
Johnstone, Sutton and Warfield (2001) present a framework for understanding the antecedents and consequences of independence risk. They identify as antecedents both incentives and judgment-based decisions. The incentives that create independence risk
• direct investments;
• contingent fees;
• potential employment;
• financial dependence;
• interpersonal relationships; and
• auditing work of self or firm.
The existence of a judgment-based decision that emerges from a client-auditor resolution process is a further necessary condition for independence risk to affect audit quality. This can be a difficult accounting issue, audit-conduct decisions or materiality decisions. The
factors that can mitigate these threats to independence are identified as:
• corporate governance mechanisms;
• regulatory oversight;
• auditing firm policies;
• auditing firm culture (which can range from a public duty culture, through a risk management culture to a client advocacy culture); and
• individual auditor characteristics.
Catanach and Walker (1999) offer a framework for ‘audit quality’, rather than auditor independence. They suggest that audit quality is a function of two determinants of auditors’ performance: ability and professional conduct (including independence). The impact of economic incentives, market structure and audit tenure is also included. The first four of these five components of their model each comprise a number of elements.
Kleinman and Palmon (2001) argue that ‘little can be learned from dry economic models, which assume that all of us behave in a “rational” way…we are interested in the real show…where the main actors are human beings’ (p. vii). They synthesise extant research on auditor-client relationships and use this platform, together with established theories from the field of social psychology, to construct a multi-level model of these relationships. The levels considered are those of the individual, the audit firm (together comprising the micro model) and the wider environment (the macro-model).
The individual factors impacting on the auditor’s independence decision are personality factors, values, motivation, stage of life/career, and aspiration level. Group factors relate to the auditor in his work setting (his social context) and include the impact of a variety of sub-group (office) norms. Firm-level factors focus on organisational culture, organisational structures and control issues. Finally, macro-level factors include the influence of client characteristics, the history of auditor-client interactions and general environmental factors.
Beattie, Fearnley and Brandt (2001) undertake case studies of six real-life auditor-client relationships covering 22 significant interactions, using grounded theory methods to develop a model of the contextual factors influencing the outcome of accounting interactions.
The main factors are:
• level of integrity of audit engagement partner;
• company type and situation;
• effectiveness of corporate governance;
• clarity of accounting rules on issue;
• level of audit firm support and quality control; and
• quality of primary relationship.
From the analysis of the case studies they develop a taxonomy of audit partner types:
• crusaders have extremely high professional and personal integrity and are prepared to escalate issues;
• safe hands have high professional integrity; they identify closely with the client and are also prepared to escalate issues;
• accommodators have moderate professional integrity; they will knowingly bend the rules under pressure; and
• trusters have moderate professional integrity; they may be insufficiently critical and questioning in their role as auditor and may unknowingly permit rules to be bent.
Although not found in the case studies Beattie, Fearnley and Brandt (2001) suggest the theoretical possibility of there being two more partner types, these being incompetents and rogues.
It is particularly interesting that the findings from these studies identify the wide range of influences that bear on the audit process. Apart from the more obvious influences of the regulatory framework and the governance of the client company, a vital influence is found to be the personal characteristics of the individual partner and the culture and governance of the audit firm.
These are also close links between the taxonomy developed by Beattie and Fearnley from case studies, Kohlberg’s model of ethical cognition and the findings of Ponemon and Gabhart (1990) and Windor and Ashkanasy (1995). These studies provide evidence that different levels of ethical cognition among individual auditors influence their decisions.
2.8 Summary and comments It is clear that a variety of factors influence auditors’ decisions. These factors are of four main types – economic, behavioural, regulatory and contextual. The analysis of economic factors using formal mathematical modelling allows the main economic incentives to be examined. Unfortunately, the analysis is often limited to simple single-period settings that are not representative of the usual auditor-client relationship. In addition, these models make a number of simplifying assumptions, some of which can be quite restrictive. Nevertheless, these studies generally show that the joint provision of audit and NAS by incumbent auditors does not adversely affect auditor independence. This is a significant finding, in that it allows us to see that economic incentives generally act to produce independent behaviour on the part of auditors. They also permit the impact of certain regulatory interventions to be evaluated.
Economic models do not, however, recognise the complexity of human motivation. A number of writers have recently offered more complete frameworks that identify the broad range of influences on auditors’ behaviour. However formal modelling at this level of complexity is not possible.
The findings from the experimental studies of ethical cognition and moral development are significant. They show the influence of ethical cognition and moral development on the individual auditor’s decision making. A concern about experimental studies is that participants may provide an answer which they consider is expected rather than their true beliefs. However, the findings are supported and strengthened by other empirical and theoretical studies. Also of significance are the emerging findings that the culture of the audit firm itself is also a significant influence.
PART 2 CURRENT REGULATORY FRAMEWORKS
AUDITOR INDEPENDENCE AND NAS: ANALYSIS OF CURRENT
3.1 Introduction In order to protect auditor independence, regulatory frameworks in various countries lay down regulations which auditors are expected to observe. These regulations may be set in legal or professional frameworks, and be rule or principles based depending on the country specific culture and environment.
In this chapter we review the provisions contained in key current regulatory frameworks concerning independence generally and the provisions specifically relating to non-audit services. We also briefly consider the position in relation to the audit of the smaller company. This review covers current regulatory frameworks in the UK (ICAEW, 1997, 2001b), the US (SEC, 2000), Ontario (ICAO, 2002) and Australia (ICAA, 2002) and two other significant and recently developed frameworks from the European Commission (EC, 2002) and International Federation of Accountants’ Code of Ethics for Professional Accountants (IFAC, 2001). For ease of reference these documents are henceforth referred to as, UK, SEC, Ontario, Australia, EC and IFAC frameworks.
3.2 Regulation of auditor independence An interesting conceptual framework for the regulation of auditor independence is offered in a staff report by the now defunct US Independence Standards Board (ISB, 2001),5 which also recognises that independence is a matter of degree and identifies and
discusses three basic principles for regulating auditor independence:
• considering the level of independence risk and assessing its acceptability;
• considering benefits and costs; and • considering the views of investors and other interested parties.
Independence risk is considered to be the likelihood that an auditor’s objectivity would be compromised or would appear to be compromised to well informed investors or others. Decision makers should consider the level of independence risk by analysing threats and their significance, assessing whether the level is acceptably low and applying appropriate safeguards if it is not. The benefits of reduced risk should not exceed the costs and the views of investors and others with an interest in financial reporting should be taken into account when considering the level, acceptability and cost of independence risk.
The US ISB was set up in 1997 and, rather surprisingly, disbanded in 2001. The conceptual framework was issued as a staff report but was never considered or adopted by the ISB.
3.3 Independence in professional and legal regulatory frameworks 3.3.1 Appearance and fact The main provisions in the regulatory frameworks we have reviewed in respect of independence generally and particularly about the distinctions made between independence in fact and independence in appearance are summarised in Table 3.1.
As may be observed, the Australian framework is almost identical to IFAC and was introduced in 2002. The others are recent or have been recently amended. The Council of the ICAEW in the UK, whose framework was introduced in 1997, agreed in June 2002 to adopt the 2002 EC Recommendation as best practice where it differs from the ICAEW’s current framework. (This is identical to that of the Institute of Chartered Accountants of Scotland and the Institute of Chartered Accountants in Ireland and broadly similar to that of the Association of Chartered Certified Accountants.) Minor changes were made to the UK’s framework in 2001, following a review of guidance on auditor independence (ICAEW, 2001a). The SEC’s revised rules are also very recent, taking effect from 2001.
These completely new or substantially reformed frameworks, combined with the major reviews of aspects of auditing and independence that have been carried out (e.g. the US Panel on Audit Effectiveness (POB, 2000); the Australian Ramsay Report (2001); and the Report of the Review Group on Auditing in the Republic of Ireland (2000)) reflect a considerable degree of concern among regulators in all these countries about auditor independence.