The Role of Audit Firm Size, Non‐Audit Services, and Knowledge Spillovers in Mitigating Earnings Management during Large Equity Issues

• DOI: http://doi.org/10.1111/ijau.12073
• Published date: 01 November 2016
• Author: Aasmund Eilifsen,Kjell Knivsflå
• Date: 01 November 2016

The Role of Audit Firm Size, Non-Audit Services, and Knowledge Spillovers in

Mitigating Earnings Management during Large Equity Issues

Aasmund Eilifsen and Kjell Knivsflå

NHH Norwegian School of Economics

We investigate how audit firm size and large auditor-provided non-audit services (NAS) affect accruals quality

around large equity issues and acquisitions for Norwegian public companies from 1999 to 2013. We find poorer

accruals quality around large equity increases. Big 4 audit firms mitigate adverse accruals quality but only if the

provision of NAS is moderate or low. In contrast, non-Big 4 audit firms are associated with lower accruals quality

around large equity increases, butthis effect is moderated if the provision of NAS is high. The findings are consistent

with joint provision of audit and NAS as an effective learning mechanism to mitigate earnings management for

smaller auditfirms and a lesser willingness of largeraudit firms to protect earningswhen the provision of NAS is high.

The evidence contrasts with the notion of large audit firms as universally superior audit quality suppliers and

establishes new evidence on how knowledge spillovers from NAS may improve audit quality.

Key words: Abnormal accruals, audit firm size, audit quality, earnings quality, equity issues and acquisitions,

knowledge spillovers, non-audit services

Regulators and policy makers need to demonstrate that

NAS actually cause auditors to impair their inde-

pendence and this has been lacking. Likewise, the

profession needs to provide compelling evidence that

NAS do not harm but enhance audit quality and

efficiencies. Neither side has provided conclusive proof.

(Sharma, 2014, p. 85)

INTRODUCTION

This study contributes towards the resolution of the

apparent controversy between regulatory policy decisions

and research evidence on audit quality and efficiency

effects of auditor-provided non-audit services (NAS). Our

investigation extends the extant literature by examining

the combined effect of audit firm size and large provisions

of NAS on earnings qualityaround large equity issues and

acquisitions. As expected, accruals quality in periods

around large equity issues is poorer. Big 4 audit firms are

better at protecting accruals quality but only if NAS are

moderate or low. The findings indicate that large

provisions of NAS induce less willingness to protect

earnings for larger audit firms, while smaller audit firms

benefit from knowledge spillovers when NAS are large.

This supports the idea of a turning point at which NAS

harm audit quality (Knechel, Sharma & Sharma, 2012;

Sharma, 2014) and thatlarge provisions of NAS may affect

smaller audit firmsdifferently.The evidence contrasts with

the common notion of large audit firms as universally

superior audit quality suppliers.

Our investigation is motivated by the apparent tension

between regulatory policy decisions andresearch evidence

on adverse effects of auditor-provided NAS(Francis, 2011;

DeFond & Zhang, 2014; Sharma, 2014), the presumption

that NAS regulations may be motivated by concern for

low audit quality when the audited financial statements

are highly vulnerable to earnings management (Carson,

Tronnes & Wong, 2014), and the limited research evidence

on how auditor-provided NAS may create knowledge

spillover benefits (Sharma, 2014; Christensen, Olson &

Omer, 2015).

Regulators continue to mandate new regulations

providing NAS to audit clients (European Union, 2014),

although there is no compelling evidence that NAS

diminish actual audit quality (DeFond & Zhang, 2014). At

the same time, to encourage more competition, legislators

and regulators seek to promote the role of non-Big 4 firms

in the market for large audits (European Commission,

2011; European Union, 2014), while research consistently

finds that large firms provide higher audit quality when

compared to smaller audit firms (Francis, 2004, 2011). One

may question why regulators are not more attuned to

insights from research in this area (Francis, 2011). One

reason may be that archival auditing research typically

investigates associations betweenthe averages of variables

in samples, while regulators find reasons to intervene to

repair problems existing in tails of sample distributions or

in their envisioned distribution of cases (Carson et al.,

2014). For example, regulatory interventions may be

motivated by concern for low audit quality when auditors

provide large NAS in circumstances where management

has strong incentives to manage earnings.A second reason

may be that regulators devote too much attention to

auditor independence and do not focus enoughon auditor

competence (Humphrey, Moizer & Turley, 2007). Insight

from the provision of NAS may improve auditors’ability

to deliver higher audit quality through knowledge

spillovers or economies of scope (Arruñada, 1999). A third

reason is that research is silent or incomplete or provides

inconsistent evidence of the many complex relationships

and variables that may affect audit quality (Francis, 2011;

DeFond & Zhang, 2014). For example, evidence is

inconclusive on how audit quality indicators such as audit

firm size and auditor industry expertise affect the

knowledge spillover benefits that may accrue from the

provision of NAS (Lim & Tan, 2008; Eilifsen & Knivsflå,

2013).

Correspondence to: Aasmund Eilifsen, NHH Norwegian School of

Economics,5045 Bergen, Norway.Email: aasmund.eilifsen@nhh.no

International Journal of Auditing doi: 10.1111/ijau.12073

Int. J. Audit. 20:239–254 (2016)

©2016 John Wiley& Sons Ltd ISSN 1090-6738

Biased financial reporting is more likely to occur in

circumstances where the audit client has strong incentives

to manage earnings (Schipper, 1989). Incentives to engage

in earnings management are arguably high when large

increases in equity take place through issues and

acquisitions. A number of studies indicate that earnings

management of audited financial numbers takes place

during initial public offerings (IPOs), seasonal equity

offerings (SEOs), and mergers and acquisitions (M&As)

(e.g., Rangan, 1998; Teoh, Welch & Wong, 1998a; 1998b;

Morsfield & Tan, 2006). Research also typically finds that

larger firm auditson average are of higher quality (Francis,

2011;DeFond & Zhang, 2014). Consistentwith this finding,

some studies of equity offerings suggest that large audit

firms reduce earnings management during IPOs, SEOs,

and M&As (e.g., Zhou & Elder, 2004; Chen, Lin & Zhou,

2005; Ahmad-Zaluki, Campbell & Goodacre,2011).

In general, research has been unable to support the

commonly held view by regulators and others that jointly

providing audit and NAS impairs auditor independence

and negatively affects earnings quality (Schneider, Church

& Ely, 2006; DeFond & Zhang, 2014; Sh arma, 2014). The

findings may reflect the existence of economies of scope

from audit-provided NAS that improve the auditor’s

ability to detect misstatement and control earnings

management (e.g., Arruñada, 1999; Kinney, Palmrose &

Scholz, 2004; Antle et al., 2006). Empirical evidence on

knowledgespillovers, however,remains mixed and elusive

(Krishnan & Yu, 2011). Extant literature fails to answer the

question of whether auditor-provided NAS trigger

knowledge spillovers that are more beneficial for smaller

firms’audits than larger firms’audits or vice versa (Louis,

2005; Lim & Tan, 2008; Krishnan & Yu, 2011; Knechel

et al., 2012; Eilifsen & Knivsflå, 2013).

Our sample consists of Norwegian public companies and

2,064 company-year observations for the period 1999–2013.

High investor protection and strong legal enforcement

characterize Norway and its financial reporting and

auditing environment (La Porta et al., 1997, 1998; Leuz,

Nanda & Wysocki, 2003; Choi & Wong, 2007; Hope et al.,

2009). The EU’s accounting and auditing regulations apply

in Norway, and all Norwegian audits should comply with

International Standards on Auditing.

1

We believe, and our

findings indicate, that Norway is a suita ble setting to gather

research evidence on the relations between audit firm size,

NAS, and earnings management around equity issues.

In this study accruals quality is measured by the

standard deviation of the residuals (Dechow & Dichev,

2002; Francis et al., 2005) obtained from the modified and

performance-adjusted Jones model (Jones, 1991; Dechow,

Sloan & Sweeney, 1995; Kothari, Leone & Wasley, 2005).

We find poorer accruals quality in periods surrounding

(i.e., two years before, during, and one year after) large

non-earned equity increases. Big 4 audit firms mitigate

the adverse effect on accruals quality but only if the

provision of NAS is moderate or low. In contrast, non-Big

4 audit firms are associated with lower accruals quality

during large equity issues, but this association is

moderated if the provision of NAS is high.

The next section presents an overview of prior research

and develops our hypotheses. This is followed by a

discussion of our research methodology. In the fourth

section, we describe our sample and our variable

measurement and report our descriptive statistics and

correlations. The following sections present our results

and those of our robustnesstests. The last section provides

a summary of conclusions.

LITERATURE AND HYPOTHESES

Issues of new equity throughIPOs, SEOs, and M&As offer

a setting with strong incentives to manage earning to

attract capital or negotiate stock swaps on favorable terms

(Schipper, 1989; Healy & Wahlen, 1999). This has en-

couraged researchers to investigate earnings management

around the times of equity issues.

Equity issuers may be motivated to adopt higher

abnormal (discretionary) accruals,for example, lower than

normal depreciations, to inflate reported earnings to

influence stock prices (Teoh et al., 1998b). Several studies

reveal evidence of higher abnormal accruals (lower

accruals quality) around equity increases such as IPOs,

SEOs, or M&As (e.g., Friedlan, 1994; Rangan, 1998; Teoh

et al., 1998a; Erickson & Wang, 1999; Shiv akumar, 2000;

DuCharme, Malatesta & Sefcik, 2001; Louis, 2004;

Morsfield & Tan, 2006; Botsari & Meeks, 2008; Cohen &

Zarowin, 2010).

2

Some studies find a weak or no asso-

ciation between equity increases and abnormal accruals

(Aharony, Lin & Loeb, 1993; Armstrong, Foster & Taylor,

2009), while Ball and Shivakumar (2008) report lower

abnormal accruals for equity-offering companies.

3

Although the evidence is mixed, most reveals that

earnings management takes place duringlarge non-earned

equity increases. We therefore advance the following

hypothesis:

H1: Large equity issues are associated with lower

accruals quality.

Audit quality through auditors’independence and

competence is a component of financial reporting quality

because the audited financial statements are the joint

product of the management and the auditor, where the

auditor assures that the financial statements are fairly

presented (DeAngelo, 1981; DeFond & Zhang, 2014). The

audit quality literature examines whether systematic

differencesin earnings quality areconditional on presumed

audit quality-related factors such as audit firm size and

provisionof NAS (Francis, 2011). Thisstudy, therefore, first

investigateswhether large equity issuesare associated with

lower accruals quality (earnings quality). If H1 is

supported by our data, we have a basis for investigating

how audit firm size and large provisions of NAS affect

accruals quality around the time of large equity increases.

The audit literature posits that large audit firms,

captured by the Big 4 (or generally Big N) membership,

positively affect audit quality. Large audit firms have

stronger incentives to deliver higher audit quality because

they are more exposed to reputation and litigation risk

and have higher competence in providing audit quality

than do smaller audit firms (DeAngelo, 1981; Arruñada,

1999). A large body of evidence supports this conjecture

(e.g., Becker et al., 1998; Francis, Maydew & Sparks, 1999;

Francis, 2004; DeFond & Zhang, 2014).

Of specific relevancefor our study, companieswith IPOs

and SEOs audited by large audit firms tend to engage in

less opportunistic earnings management (Zhou & Elder,

2002, 2004; Chen et al., 2005; Ahmad-Zaluki et al., 2011).

4,5

Also supporting the idea of more credible attestation of

assertions of financial information by large audit firms,

Big N audits are associated with reduced underpricing in

IPOs (e.g., Balvers, McDonald & Miller, 1988; Beatty, 1989;

Hogan, 1997). Pointing in the same direction, evidence

indicates that the market reaction to announcements of

240 A. Eilifsenand K. Knivsflå

©2016 John Wiley & Sons Ltd Int. J. Audit. 20:239–254 (2016)