CENTRE FOR INTERNATIONAL BUSINESS STUDIES

The Determinants of Audit Fees.

Further Evidence from the UK Charity Sector.

Emiliano Cantoni[(], Kenneth D’Silva[(] and Martin Isaacs[(]

Paper Number 2-10

Research Working Papers

Abstract

Given the lack of literature and the ever growing relevance of charities, the present paper investigates the determinants of audit fees in the UK charity sector. It proposes a theoretical framework to determine the audit fee, which is empirically tested on a dataset of 119 “largest” charities. The paper also investigates the presence of a fee premium associated with (now) Big4 audit firms and expertise in charity auditing, and compares the results with those of a leading study in the UK context (Beattie et al., 2001). Our findings basically confirm those of Beattie et al. (2001). The evidence shows that size, risk, and fees for Non-Audit Services (NAS) are the main determinants of audit fee determination, while, contrary to research in the private sector, the organisational complexity of the charity does not seem to have significant influence on audit fees. The results also show a positive association between audit fees and NAS fees, the presence of an audit fee premium between Big4 and non-Big4 auditors, as well as significant higher remuneration for the auditors’ expertise.

Keywords: audit fee determinants, charity sector, audit fee models, audit fee premium.


The Determinants of Audit Fees.

Further Evidence from the UK Charity Sector.

Emiliano Cantoni, Kenneth D’Silva and Martin Isaacs

1. Introduction and context

Trustees of charities are in law the governors of those charities. Consequently, they are responsible for all aspects of good governance in their charities, including the ethical aspects of their decisions (Kreander et al., 2009). A primary trustees’ responsibility is their duty of care to ensure that all the assets of the charity are appropriately and optimally expended (Charity Commission, 2009: 27). Moreover, charity trustees are responsible for the preparation of the annual report and financial statements, which, together, are generally recognised as key documents in discharging trustees’ accountability to external users (Connolly and Hyndman, 2004: 128). Accordingly, these should evidence the appropriate use of all funds and the level of all costs incurred. While the external audit fee is no different from other costs borne by the charity, the service received is hardly visible, with the only tangible “product” a relatively brief – and standardised – audit report (opinion). In this situation, trustees would benefit from some external assurance that their fee is not disproportionate.

Consequently, how could trustees of charities obtain an insight into the appropriateness of the audit fee charged by the charity’s external auditors? What relative yardsticks might they employ to resolve this question? The development of audit fee models should assist in providing benchmarks for assessing audit fees, which, in turn, should assist trustees both in reviewing current fee levels as well as in better informing their choice of auditors. When external audit is put out for tender, identifying those variables that are significantly associated with audit fees should help refine the tender specification, and thus improve comparison of the tenders received: over time this could increase the competitiveness of the audit market, and consequently reduce prices. Additionally, this process may well also indicate certain financial areas within the organisation where closer attention to accounting and management issues could be beneficial. Finally, charity trustees need to have confidence in their audited accounts if they are to discharge their duties effectively (Brindle, 1998 in Palmer et al., 2001: 255): using the models could enhance trustees’ accountability for, and control of, the resources expended on audit of the financial statements.

Thus, in view of the previously noted obligations of charity trustees, this paper develops explanatory models and tests them via multiple regression analysis. The final aim is to assess which determinants, stated in the financial statements of “major” charities, appear to be factors in determining the amount of the external audit fee for such entities. The paper also investigates the presence of a fee premium associated with (now) Big4[1] audit firms and expertise in charity auditing, and compares the results with those of a leading study in the UK context (Beattie et al., 2001).
The remainder of the paper is arranged as follows. The second section reviews some key literature contributions about audit fee models. The third section explains the theoretical framework (research question, objectives, and hypotheses) and the dataset employed for the study. The fourth section sets out the main findings of the research, while the fifth (and final) section considers the conclusions drawn and suggests some items for a future research agenda.

2. Literature review

The area of charity accounting was largely neglected until the 1980s. Following a research report by Bird and Morgan-Jones (1981) the standard-setting authorities turned their attention to it (Connolly and Hyndman, 2000: 95). In their seminal research into charity accounts, Bird and Morgan-Jones (1981) make an eloquent and compelling case in their section entitled “The need for competent and independent audit” (pp. 212-213). In particular they emphasize the singular features of charity audit, pointing out that audit processes are:

« […] often conducted on an honorary basis or for a nominal fee by professional accountants. There is therefore a natural tendency to try to keep down costs by using the cheapest available staff and by seeking to avoid problems that will take time to resolve. This natural tendency must be resisted. The standards of work required of a professional person are not lowered because he is doing a job for little or no fee; it would not be beneficial to anyone if they were» (Bird and Morgan-Jones, 1981: 213).

In light of the above issue, a critical review of audit fee determinants literature within the charity sector was conducted but produced poor tangible results, as confirmed in Hay et al. (2006: 156). In fact, as argued by Hyndman (1990: 295), the area of accounting and reporting by charities is relatively underdeveloped in the United Kingdom. The current meaningful corpus of directly relevant literature consists of one specific article (Beattie et al., 2001). Thus, this paper offers much needed further insights in this area.

Given the lack of such research within the charity sector, it is instructive to trace the development of relevant research within the private sector. For which there is a considerable body of literature concerning audit fee models in the private sector (Hay et al., 2006: 148-154) In general, the results generated from such models have been mixed (Chan et al., 1995: 765). Starting from the seminal paper of Simunic (1980), the research was focused on the determination of audit fee and particularly the association between audit size and audit fees. This body of literature, developed in English-speaking countries (UK, USA, Australia and New Zealand), is mainly reflected within the contributions of Taylor and Baker (1981), Francis (1984), Firth (1985), Palmrose (1986a), Francis and Stoke (1986), Francis and Simon (1987), Simon and Francis (1988), Ettredge and Greenberg (1990), Chan et al. (1993), Brinn et al. (1994).

Prior research based on the private sector suggests that any objective model of audit fee determination will depend on the time spent by the audit firm on the work (complexity of audit process), and/or the underlying risk exposure of the auditor in carrying out the work (audit risk). The evidence has shown that auditee size (measured in balance sheet and/or income/expense terms), and auditee complexity (assessed in terms of scope of operations) are the most significant determinants that impact on the level of audit fees (Table 1). The presence of a price premium between (now) Big4 audit firms and non-Big4 audit firms was investigated in audit fee models literature. While, in general, a price premium was recognized, the conclusions are not consistent. On this topic, Ireland and Lennox (2002) find that the big audit firm premium is related to a significant selectivity bias. They report that large auditors attract clients that are above average quality and require less than average audit effort (Ireland and Lennox, 2002: 89). Moreover, Craswell et al. (1995: 319) report that the fee premium of big auditors is related both to general brand name and industry specialization.

The relationships between size, complexity, risk, and audit fees has been investigated, as reported in Jubb et al. (1996) and Pong (2004). A wide discussion about audit fee models set-up is reported in Pong and Whittington (1994). Recently, the presence of internal audit board and the relation with audit fees amount have been investigated (Hay et al., 2008; Hogan and Wilkins, 2008). Table 1 – at the end of the paper - summarises the main key findings from a selection of relevant comparable studies.

Haskin and Williams (1988) presented a cross-country comparison among English-speaking countries (UK, USA, Australia and New Zealand). The results were generally positive and significant, both across countries and across firms, although the level of applicability varied from country to country, and firm to firm. The researchers considered that this demonstrated consistency in the model across the ranges of firms and countries.

Researches into audit fee models were carried out for countries other than UK, USA, Australia and New Zealand, with audit fee models developed and tested in India (Simon et al., 1986), India and Pakistan (Ahmed and Goyal, 2005), Singapore (Low at al., 1990), Hong Kong and Malaysia (Simon et al., 1992), Bangladesh (Karim and Moizer, 1996), Bahrein (Joshi and Al-Bataski, 2000), Canada (Chung and Lindsay, 1988; Anderson and Zeghall, 1994), South Africa (Simon, 1995; Firer and Swartz, 2006), the Netherlands (Langedijk, 1997), Norway (Firth, 1997), Ireland (Simon and Taylor, 2002), Italy (Cameran, 2005), France (Gonthier and Scatt, 2007), and Denmark (Thinggaard and Klertzner, 2008). Cobbin (2002), in a comparative review of the literature, reports the presence of a model of core audit fee determinants that is used and adapted in a limited way, to reflect market specific circumstances and to address market specific issues. The review also indicates some consistency across markets in respect of generic variables identified as core determinants of the level of audit fees (i.e. auditee size, complexity, risk profile, and auditor size).

A cross issue in audit fee models concerns the relation between audit fee and fee for non-audit services (Houghton and Jubb, 1999), especially after the financial scandals of the early years of this century (Ewert, 2004). The findings generally report a significant positive association between these two variables (Simunic, 1984; Palmrose, 1986b; Ezzamel et al., 1996; Firth, 1997; Ezzamel et al., 2002), even in the absence of a plausible reason for this association (Firth, 1997: 524). However, one potential explanation could be that non-audit fees are proxies for size attributes or organisational complexity (e.g. accounting and control systems) which require both increased audit and non-audit work, and which are not adequately controlled for (Simunic, 1984; Palmrose, 1986b).

The current literature suggests that, in the public sector and, particularly, in the charity sector, audit fee models have not been widely investigated by researchers. Basically, in the public sector context, a few studies examined the determinants of audit fees in the US municipal audit market Baber, 1983; Beck and Barefield, 1986; Baber et al., 1987; Rubin, 1988; Ward et al., 1994; Copley et al., 1995). The results report a similarity with audit fees determinants in the private sector. The extensive study of Beattie et al. (2001) tries to fill this gap, in the context of a growing demand for accountability in the public sector, and in particular in the charity sector. This study analyses a sample of 210 UK charities extracted from the Barings list of the top 500 UK charities, developing multiple regression models of audit fee determinants for the UK charity sector. The main findings are as follows:

1.  As in previous private sector company studies, size and organisational complexity are the major determinants of audit fee

2.  Charity sector factors of empirical significance include the nature of the charity (i.e. grant-making or fund-raising), its area of activity and the significance of trading income

3.  The audit of fund-raising charities seems to be more complex than that of grant-making ones

4.  Auditors’ provision of non-audit services is much less significant in the charity sector than in the UK private sector

5.  There is a positive association between audit fees and fees for non-audit services

6.  The results in general show that (now) Big4 audit firms receive higher remuneration than non-Big4 ones. Moreover, non-Big4 audit firms with charity expertise are rewarded with a fee premium over other non-Big4 ones

7.  Finally, the study demonstrates that the charity audit fee rate (e.g. audit fee/sales) is significantly lower than that of private sector companies: in fact it is approximately half (Beattie et al., 2001: 271). The authors argue that the lower audit fees in the charity sector could be consistent with an inferior perception of risk by auditors, with auditors’ altruism but also with lower audit quality[2].

3. Research design and methodology

3.1. Theoretical framework

The present paper aims to contribute to the knowledge of audit fee determinants, with particular emphasis on non-profit organisations. The objective of the study is to explore and test the determinants affecting audit fees in the UK charity sector, and compare the findings with those of a leading research (Beattie et al., 2001) in this Country. The research question is therefore focused on the determinants of audit fees for the charity sector. In fact, as argued in Beattie et al. (2001: 249):

«Much of the logic of previous work on private sector companies is relevant in deriving [the] model but it is also necessary to consider additional potential explanatory variables to capture the unique aspects of charities.»

The determinants (and their proxies) of audit fee for charities may be identified from a priori theoretical relevance, from analysis of the literature , and by using statistical technique (e.g. factor analysis) in order to reduce and summarize the information content of variables (i.e. empirical analysis).

The selection of the variables hinged on arriving at a consensus of what would be the primary or key factor(s) underlying any rational model of audit fee determination, and then deciding what components of a set of charity accounts would give some indication of those factors, bearing in mind that a different selection of variables could modify the structure of the models and their reliability. In this paper a specific theoretical framework for the UK charity sector is introduced. It derives from the relevance of factors discussed in literature and considers the main results in Beattie et al. (2001). This process was also informed by extensive discussion of the working paper in interviews with acknowledged charity experts heading specialist charity audit teams[3]. The initial assumption was that the primary factor affecting the amount of audit fees in charities would be time spent by the audit firm on the audit process. There is broad agreement on this issue, which then progresses to explore the underlying factors determining time spent (Pong and Whittington, 1994: 1075). It emerges that size alone is likely to be relatively minor: in fact, it could be argued that increased organisational size might result in relative reduction in time spent on audit, if it brought with it improved systems and relatively larger resources in the finance department. Again, there is general agreement (Hay et al., 2006: 169-170 and 179) that the organisational complexity of the audited organisation, the risk associated with the audit activity and the amount of fees for non-audit services are primary determinants of the level of audit fees: this is also argued to apply to the charity sector (Beattie et al., 2001: 261). In light of these considerations, the following areas/variables are proposed as audit fee determinants.