![Sample our Area Studies journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5922/TOC-Subject-Sample-2021-Area-Studies.jpg"})
Abstract
We examine the effect of changes in audit risk standards on the conduct of financial statement audits in a European setting. We investigate this by analysing the audit hours and audit fees for clients of Big 4 audit firms in Finland in 1996 and 2010. Our results show that audit firms became more sensitive to clients’ business risk due to the introduction of the new audit risk standards, with more audit hours allocated to owner-managed companies in 2010 than in 1996, and fewer audit hours allocated to low-risk clients in 2010 than in 1996. Also, the labour mix in the audit team changed for owner-managed companies, with a greater work load carried by junior auditors in 2010 than in 1996. Regarding the price of audit, we find an increase in audit fees for clients with high business risk, while audit fees remained at roughly the same level for low-risk clients. These findings should be of interest to the auditing profession and those involved in the development of auditing regulations.
JEL Code:
Acknowledgements
We are grateful to the editor, Paul André, and the two anonymous reviewers for their valuable feedback on this paper. We are also indebted to the participating audit firms for giving us access to the data analysed in this study. In addition, we would like to thank David Hay, Juha Kinnunen, Bill Messier and delegates at the British Accounting and Finance Association Annual Conference in Newcastle (2013) and London (2014); the Annual Congress of the European Accounting Association in Paris (2013) and Tallinn (2014); and those attending the research seminars held at Aalto University Business School and Brunel University London for their helpful comments.
Disclosure Statement
No potential conflict of interest was reported by the authors.
Notes
1 In 2014, this Directive was replaced by Directive 2014/56/EU.
2 In 2016, the AB3C was replaced by a new supervisory body under the auspices of the Finnish Patent and Registration Office.
3 Bell et al. (Citation2008) analyse changes prior the effective date of Sarbanes–Oxley Act of 2002 (SOX) and related regulations (p. 731, footnote 4).
4 The term ‘strategic systems auditing’ was also used by some of the researchers responding to the monograph published by KPMG (Bell, Marrs, Solomon, & Thomas, Citation1997, Citation2002, Citation2005).
5 A common factor of the accounting scandals in the early 2000s was fraudulent financial reporting. This wave of management-related frauds led to an increase in the auditors’ responsibility for detecting fraud by management. An example of this regulatory response was the introduction of ISA 240, The Auditor’s Responsibilities Relating to Fraud in an Audit of Financial Statements, which sensitizes auditors to the risk of fraudulent reporting and causes them to be less willing to accept management’s assertions at face value.
6 In the early 1990s, Finland experienced a deep recession with many business failures and alleged audit failures. This led to questions about the quality of the audit. The Finnish professional associations of auditors reacted to the adverse publicity by initiating quality assurance programmes based on voluntary peer-reviews.
7 The move from voluntary peer review to mandatory inspections by oversight bodies was instigated by the Statutory Audit Directive (2006/43/EC) in 2006 which required the Member States to establish an effective system of public oversight for statutory auditors and audit firms. The Statutory Audit Directive was implemented in the Finnish Auditing Act 2007.
8 This pressure seems warranted as sanctions of non-compliance are high. For example, AB3C in Finland now inspects the quality of audit work systematically and rejection leads to a re-inspection. Auditors who fail the re-inspection run the risk of having their authorization to conduct statutory audits withdrawn. In 2013, there were two cases where the Auditing Board of the State (ABS) cancelled the auditor’s right to conduct audits (TILA 5/2013; TILA 6/2013).
9 The report does not contain information about identities of auditors or the names of the firms. However, we base our conclusion to the information that almost all auditors not passing the quality inspections were second tier HTM-auditors, who typically work in small firms focusing on small clients.
10 The Auditing Act of 1994 (936/28.10.1994) was effective for all audits of 1996 year-end financial statements (the year of our initial data).
11 The Big 6 were Arthur Andersen, Coopers & Lybrand, Deloitte, EY, KMPG, and Price Waterhouse.
12 The 1996 data were collected in 1997.
13 The Big 6 were reduced to the Big 4 following the merger of Coopers & Lybrand with Price Waterhouse in 1998 and the collapse of Arthur Andersen in 2002.
14 In our additional analyses, we also employ audit fees (LNFEES) as a dependent variable to test whether additional effort is reflected in the audit price. We use the total audit fees as we do not have information on billing rates for different ranks of labour.
15 However, agency conflicts among managers (principals) and between managers and subordinates may be higher in owner-managed firms than other ownership types. We thank the anonymous reviewer for pointing this out.
16 For reasons of confidentiality, we do not know the precise ownership level.
17 In our test design, the coefficient (α1) of CLIENT_RISK shows the difference between high and low client risk in 1996 and the difference between high and low client risk firms in 2010 is given by α1 + α2.
18 The most common measure of client size is natural logarithm of total assets (Hay et al., Citation2006). However, we use net sales to allow comparison with our 1996 data. The 2010 data were collected by one audit firm according to our instructions.
19 To ensure anonymity, SUBS is the number of subsidiaries sorted into deciles (ordinal variable).
20 This allows us to conserve degrees of freedom given our small sample size. We feel this is justified because the mean and median for these variables are not statistically different between 1996 and 2010.
21 Data collected for sales and audit fees in 1996 were originally in Finnish marks. They were converted to euros using the 2010 Consumer Index and the official exchange rate. This means that any changes in sales and audit fees were not due to inflation or deflation.
22 For comparison, André, Broye, Pong, and Schatt (Citation2016) report the ratio of the audit fee/assets percentage of 0.042% for Italian companies over 2007–2011 period.
23 We also re-run the tests of the equality of variances with a sample where we drop the 2010 observations with the highest absolute values of centred LNSALES and trigger the difference variance between 1996 and 2010 for LNSALES in the full sample. We then find that the variance of LNSALES between 1996 and 2010 is statistically equal. After this resampling, we still obtain the same empirical results: the variance of junior auditor time (senior auditor time) decreased (increased) from 1996 to 2010.
24 We also calculated Spearman correlations, which are consistent with reported Pearson correlations.
25 We do not find heteroscedasticity using the White test (White, Citation1980) and the Breusch–Pagan test (Breusch & Pagan, Citation1979) in any of the models used to test the hypotheses. Nevertheless, we also calculate robust standard errors. The results remain qualitatively the same with robust standard errors.
26 The dependent variable (audit hours) and client size (sales), are natural logarithms while the test variables are in their original metric. This means that our untransformed original dependent variable changes by 100*(e^∝-1) per cent for a one unit increase in the independent variable when all other variables in the model are held constant (Woolridge, Citation2013).
27 Note that .