The British Accounting Review 47 (2015) 304e320

Contents lists available at ScienceDirect

The British Accounting Review

journal homepage: www.elsevier.com/locate/bar

Audit fees, auditor choice and stakeholder influence:

Evidence from a family-firm dominated economy
Arifur Khan a, Mohammad Badrul Muttakin a, Javed Siddiqui b, *
a
Department of Accounting, Faculty of Business and Law, Deakin University, Australia
b
Division of Accounting and Finance, Manchester Business School, University of Manchester, UK

a r t i c l e i n f o a b s t r a c t

Article history: Despite the dominance of family-owned publicly listed companies in developing econo-
Received 14 January 2013 mies, prior research has paid relatively little attention to this area and the socio-economic
Received in revised form 10 March 2015 context of these countries has been mostly ignored. This study contributes to the ac-
Accepted 11 March 2015
counting literature by providing empirical evidence of the effects of family control and
Available online 27 March 2015
ownership on audit pricing and auditor choice in a developing economy context. Using
1058 firm-year observations of publicly listed companies in Bangladesh, where family
Keywords:
firms are the most dominant form of public companies, we find that in comparison with
Family control
Family ownership
non-family firms, our sample family firms pay significantly lower audit fees and choose
Audit fees lower quality auditors. However, for export-oriented industries, family firms seem to pay
Auditor choice significantly higher audit fees and recruit better quality auditors compared to non-family
Emerging economy firms. Collectively, our findings have important implications for audit markets in emerging
economies in which the sustainability of family firms is crucial for overall economic
development.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

We explore audit fees and auditor choice in a family-firm dominated economy. The current decade has seen a phenomenal
growth in family businesses, especially in emerging economies in Asia. Credit Suisse (2011) reports that family businesses are
now the backbone of many Asian economies, accounting for 32 percent of total market capitalisation. In Southeast Asia,
where important emerging economies such as India and Malaysia are located, family businesses make up 65 percent of listed
companies, and 49 percent of market capitalisation. However, despite this remarkable contribution of family businesses,
especially in emerging economies, the topic of family business has remained relatively under researched in accounting
literature. Likewise, family firms have received little attention from audit researchers (Burkart, Panunzi, & Shleifer, 2003;
Faccio & Lang, 2002; La Porta, Lopez-de-Silanes, & Shleifer, 1999; Wang, 2006). Although the traditional agency problem,
characterised by a conflict of interest between owners and managers (referred to as the ‘type I’ agency problem) is less of an
issue for family firms, a growing body of literature (for example, Faccio & Lang, 2002; Villalonga & Amit, 2006) suggest that
families have powerful incentives to expropriate wealth from minority shareholders, and pursue their own interests at the
expense of non-controlling shareholders (referred to as the ‘type II’ agency problem). Prior research on auditor choice and
audit fees in family firms presents two possible scenarios. Due to lower type I agency problems, it is expected that family firms

* Corresponding author. Tel.: þ44 161 2750440.

E-mail addresses: arifur.khan@deakin.edu.au (A. Khan), m.muttakin@deakin.edu.au (M.B. Muttakin), javed.siddiqui@mbs.ac.uk (J. Siddiqui).

http://dx.doi.org/10.1016/j.bar.2015.03.002
0890-8389/© 2015 Elsevier Ltd. All rights reserved.
 A. Khan et al. / The British Accounting Review 47 (2015) 304e320 305

will have lower demand for audit quality and eventually pay lower audit fees (Ho & Kang, 2013). On the other hand, the
presence of strong incentives in family firms to engage in fraudulent activities may increase audit risk. To mitigate such risks,
auditors may be required to perform audits that are more extensive and will charge higher audit fees. The presence of such
conflicting arguments makes auditor choice and audit fees in family firms an interesting research area. In an important review
paper summarising the large body of audit fee research, Hay, Knechel, and Wong (2006) identified the need for further
research examining how different forms of ownership, including family ownership, have an impact on audit fees. In his
follow-up paper, Hay (2013) reported no significant further developments in this area. This presents the context for our study.
In this paper, we respond to Hay et al.'s (2006) call to explore a potential gap in the audit literature by investigating audit
fees and auditor choice in family firms in a developing economy, using the case of Bangladesh. Bangladesh has made sig-
nificant economic progress in recent years.1 However, like many other developing economies, the Bangladesh corporate
sector is characterised by relatively unsophisticated legal and regulatory framework (Farooque, Zijl, Dunstan, & Karim, 2007),
high ownership concentration, weak capital markets, lack of shareholder activism, and poor enforcement and monitoring of
regulations (Siddiqui, 2010). On average, the top five shareholders, usually belonging to the same family, hold more than 50
percent of shares in listed public limited companies (Imam & Malik, 2007). As of December 31, 2011, more than 70 percent of
the top performing companies on the Dhaka Stock Exchange (DSE) were family-owned firms, making this the dominant form
of listed companies in Bangladesh.2
The audit market in Bangladesh is characterised by poor demand for audited financial statements (Ahmed & Goyal, 2005)
and poor perceptions regarding audit quality (Sobhan & Werner, 2003). At present, three categories of audit firms operate in
the Bangladeshi audit market: the Big 4 firm e KPMG, Big 4 affiliate (B4A) firms, and local audit firms. It seems unusual that
KPMG is the only Big 4 firm with a direct presence in Bangladesh. The rest of the Big 4 firms operate through their associates
or cooperating firms; they do not have to adhere to Big 4 quality control standards. Siddiqui, Zaman, and Khan (2013) report
that unlike Big 4 firms, B4A firms do not earn a significant audit fee premium in Bangladesh. This indicates that the quality of
B4A firms is not perceived to be the same as a Big 4 firm. The presence of such quality differentiation in the audit market adds
an interesting dimension to this study.
Another major feature of the Bangladeshi corporate sector is the presence of important stakeholder groups in the form of
foreign buyers. Bangladesh is the second largest exporter of garment products in the world, and many family-owned firms
operating in the garments sector act as supply chains to renowned international brands. Recent research (see Islam & Deegan,
2008; Khan, Muttakin, & Siddiqui, 2013) has identified these foreign stakeholders to be important pressure groups asking for
greater financial disclosures, particularly in the form of corporate social responsibility. It might be reasonable to assume that
these important stakeholders would also be concerned about the quality of audited financial statements in Bangladesh, and
influence the choice of audit firms. Ashbaugh and Warfield (2003) document the influence of foreign stakeholders in the
German context. The study found that foreign stakeholders had a general preference for ‘dominant’ auditors for their capacity
to offer a wide range of services. In the context of Bangladesh, we argue that the decision to recruit a better quality audit firm
may stem from the audit firm's reputation regarding the quality of audit services, which mitigate some concerns regarding
the quality of financial statements.
The presence of such attributes makes developing economies such as Bangladesh an interesting research site. Prior
research (see, e.g., Wang, 2006) provides two competing theories on the effect of family ownership on earnings quality: the
entrenchment effect and the alignment effect. The alignment effect implies that concentrated ownership creates greater
monitoring by controlling owners (Demsetz & Lehn, 1985; Shleifer & Vishny, 1997), suggesting that controlling families might
monitor firms more effectively. According to this approach, founding families are more likely to forgo short-term benefits
from managing earnings because of the incentives to pass on their business to future generations and to protect the family's
reputation. Hence, family firms are motivated to report earnings of higher quality than non-family firms. However, the
presence of such incentives to report earnings in good faith, may, in turn, reduce the demand for audit quality, if the
stakeholders are convinced that family ownership enhances corporate governance (Wang, 2006). The entrenchment effect,
on the other hand, is consistent with the view that controlling shareholders in family firms have incentives to expropriate
wealth from other shareholders (Morck, Shleifer, & Vishny, 1988; Shleifer & Vishny, 1997). The impact of entrenchment effect
on earnings quality is not as straightforward as the alignment effect. The presence of more severe type 2 agency problem
would suggest that family ownership would be associated with the supply of lower earnings quality as family members have
the incentive and ability to manipulate accounting numbers in order to engage in self-beneficial activities. In contrast, the
entrenchment effect of family ownership may motivate users of financial statements and external shareholders to demand
high-quality earnings from family firms to better safeguard their assets and interests (Wang, 2006). Young, Peng, Ahlstrom,
Bruton, and Jiang (2008) points out that in developed countries, because minority shareholders are protected there is less
scope for expropriation. Therefore, the lower demand for audit quality in family firms (see Ho & Kang, 2013) may be due to the
alignment effect. In developing countries like Bangladesh, such institutional protection is absent (Young et al., 2008), making
these countries more susceptible to the expropriation effect. Also, whereas the market for corporate control deemed to be the

1
Bangladesh has maintained more than 6 percent GDP growth over the last ten years, mainly fuelled by a booming textile sector (World Bank, 2012). The
country is now regarded as one of the ‘next eleven’ emerging economies (Goldman Sachs, 2011).
2
For this research, we use company data until 2010. As of December 31, 2011, more than 70 percent of the top performing companies in the Dhaka Stock
Exchange were family-owned firms (www.dsebd.org).
306 A. Khan et al. / The British Accounting Review 47 (2015) 304e320

governance mechanism of last resort in developed economies, it tends to be typically inactive in emerging economies (Peng,
2006), creating more scope for expropriation. Francis, Khurana, and Pereira (2003) document that countries with weak legal
framework generally demand poor quality audit. Therefore, although family firms in developing countries such as Bangladesh
are still likely to have lower demand for audit quality, this is due to the expropriation effect rather than the alignment effect.
The presence of important, albeit minority, shareholder groups in developing countries such as Bangladesh adds an
interesting dimension to the demand for audit quality in family firms. Recent studies (see Islam & Deegan, 2008; Khan et al.,
2013) have reported that in the absence of active capital markets and adequate oversight mechanisms in Bangladesh, foreign
buyers act as the most important stakeholder group. Their presence presents opportunities for risk sharing; their monitoring
activities also reduce information asymmetry. To maintain long term and sustainable relationship with the foreign buyers,
family firms may forgo short term benefits from managing earnings. However, unlike developed economies, this may not lead
to lesser demand for audit quality, the foreign buyers may not be convinced of the impact of family ownership on corporate
governance. Rather, prior studies in Bangladesh (Sobhan & Werner, 2003) suggest that family ownership actually has a
detrimental impact on good quality corporate governance. The alignment effect, therefore, implies that family firms in export
industries may have sufficient incentives to recruit better quality auditors and pay higher audit fees in order to allay concerns
of the foreign buyers, and subsequently reduce agency costs.
Using a sample of 1058 firm-year observations of public limited companies listed on the DSE from 2005 to 2013, we
investigate auditor choice and audit fees in family firms. Our results indicate that family-owned listed public limited firms in
Bangladesh pay significantly lower audit fees compared to non-family firms and are less interested in recruiting higher quality
auditors. However, we find that family firms operating in export industries select better quality auditors and pay higher audit
fees, suggesting that managers in export-oriented family firms select auditors strategically to respond to the requirements of
important stakeholder groups. We thus provide further empirical support to recent studies (see Islam & Deegan, 2008; Khan
et al., 2013) that acknowledge the role of important stakeholder groups in ensuring greater financial reporting quality. Our
findings also reveal that family firms with family CEOs tend to pay lower audit fees and recruit lower quality auditors, thus
providing empirical support to prior studies (see Siddiqui, 2010) that question the efficacy of western-styled corporate
governance mechanisms in the context of developing economies. Our findings are robust to a number of statistical tests.
Our study is one of the first to investigate audit fees and auditor choice in family firms in the context of developing
economies. In contrast with previous studies on family firms (for example, Ho & Kang, 2013; Niemi, 2005; Niskanen,
Karjalainen, & Niskanen, 2010), we provide evidence of the impact of family ownership and control on auditor choice and
audit fees in a developing economy context, where publicly held family firms constitute the majority of economic activity. By
providing evidence of the impact of ownership structure and stakeholder power on auditor choice and audit fees, we extend
the work of Siddiqui et al. (2013), Khan et al. (2013), Khan, Hossain, and Siddiqui (2011) and Islam and Deegan (2008) on audit
and disclosure quality in the context of developing economies.
The remainder of the paper is organised as follows: the next section provides a review of existing literature on auditing in
family firms and discusses the characteristics of the audit market in Bangladesh. Based on a discussion regarding the insti-
tutional context and the role of audit firms, the section then develops the hypotheses to be tested for the purpose of this study.
A subsequent section then discusses the methodology used in this paper and the models used to test the hypotheses. This is
followed by a section that discusses the results. Next, a concluding section summarises the findings and discusses possible
implications.

2. Literature review, institutional context and hypothesis development

2.1. Literature review

Hay et al. (2006) provide perhaps the most comprehensive summary of research on audit fees. Their paper uses a meta-
analysis to test the combined effect of the most commonly used variables in audit fees research. On the basis of their ob-
servations about anomalies, inconsistencies, and gaps in the previous literature, Hay et al. (2006) suggest that audit research
would benefit from studies that examine how different forms of ownership (for example, types of dominant shareholders,
such as parent/subsidiary relationships versus family-run businesses) affect audit fees. In a follow up study, Hay (2013) re-
ported that a gap still existed in the audit literature, with a very limited number of studies looking into the relationship
between ownership structure and audit fees. Chan, Ezzamel, and Gwilliam (1993) use ownership control as an explanatory
variable in identifying the determinants of audit fees in the UK. Informed by agency theory, they hypothesise that the extent
of audit services demanded is a function of ownership control variable. That is, companies with a diverse ownership structure
require more extensive and higher quality audit over and above what is necessary to fulfil the minimum statutory re-
quirements. Their paper found high ownership control to be negatively associated with audit fees, suggesting that major
stakeholders can actually monitor the activities of management through direct board membership and other informal
channels. Mitra, Hossain, and Deis (2007) investigate the relationship between ownership characteristics and audit fees. Their
paper uses a supply and demand based approach to analyse the relationship between ownership control and audit fees.
Consistent with Chan et al. (1993), Mitra et al. (2007) also acknowledge that ownership control would have an effect on audit
fees. However, Mitra et al. (2007) also argue that from the demand side perspective, when ownership interests are high,
managers may also purchase better quality audit services to signal normative behaviour to the market. Their paper finds a
significantly positive relationship between diffused institutional stock ownership (i.e., having less than 5 percent individual
 A. Khan et al. / The British Accounting Review 47 (2015) 304e320 307

shareholding) and audit fees, and a significantly negative relationship between institutional block holder ownership (i.e.,
having 5 percent or more individual shareholding) and audit fees. Additionally, they document that managerial stock
ownership is negatively associated with audit fees.
Francis (2004) notes that an important development in audit quality research is based on the premise that audit quality
differentiation exists. In many cases, such differentiation is attributed to different classes of auditors, predominantly based on
size. A number of studies (for example, DeAngelo, 1981; Simunic & Stein, 1987) consider audit firm size to be an important
measure of audit quality, as larger firms would have a greater reputation to lose. Francis (2004), however, notes that such
arguments do not necessarily establish that Big 4 audit firms are superior, as evidenced by high profile audit failures.
Nevertheless, the paper suggests that Big 4 firms generally earn an audit premium of about 20 percent around the world.
Francis (2004) indicates that a higher fee generally means higher audit quality, either through greater audit efforts or through
higher billing rates charged by the auditors due to their greater expertise and reputation. The paper provides confirmatory
evidence that, on average, audit firms charging higher audit fees (i.e., Big 4 audit firms) actually provide higher quality audits,
measured in terms of observable outcomes such as the auditors' propensity to issue modified audit opinions, and the quality
of audited financial statements.
A small number of studies have explored the choice of auditors in family firms. Using a sample of small private firms in
Finland, Niskanen et al. (2010) find that an increase in family ownership decreases the likelihood of hiring a Big 4 auditor. This
is consistent with the type I argument regarding the prevalence of lower information asymmetry between owners and
managers in family firms. Ho and Kang (2013) investigate auditor choice and audit fees in family firms using data from S&P
1500 companies in the USA. They find that family firms in the USA tend to recruit lower quality auditors and incur lower audit
fees. However, both these studies were conducted in the context of developed economies where socio-economic environ-
ment differs significantly from developing countries. However, Jaffer and Sohail (2007) report that corporate governance
structures are almost non-existent in the majority of family-owned firms in India. Consequently, only 15 percent of family-
owned businesses in India and Pakistan survive into the third generation. Frequent family disputes lead to the family firm's
inability to implement a succession plan, which, in the absence of proper corporate governance mechanisms, poses a sig-
nificant threat to a family firm's survival.
We also investigate the influence of important stakeholder groups on audit fees and auditor choice. Prior research has not
explicitly addressed the role of stakeholders in the context of family firms. However, Ashbaugh and Warfield (2003) docu-
ment that stakeholder groups such as creditors and foreign suppliers, have a preference for German listed companies to be
audited by a ‘dominant’ auditor. This is predominantly due to the dominant audit firm's ability to offer a variety of audit and
non-audit services.

2.2. Institutional context and hypothesis development

The audit market in Bangladesh is characterised by poor demand for audited financial statements (Ahmed & Goyal, 2005;
Karim & Moizer, 1996),3 poor perceptions regarding audit quality (Sobhan & Werner, 2003), and the presence of only one Big 4
audit firm (Siddiqui et al., 2013). Siddiqui (2010) reports that despite having a very large population, the number of qualified
auditors in Bangladesh is significantly low compared to neighbouring countries e India, Pakistan, and Sri Lanka. Imam and
Malik (2007) report that on average 32.33 percent of the shares of listed Bangladeshi companies are held by the top three
shareholders, usually from the same family. Farooque et al. (2007) find that approximately 78 percent of CEOs are shareholders
of Bangladeshi firms, either as founder shareholders or as descendants of founding families. The study also finds that the five
largest shareholders hold more than 50 percent of the shares in Bangladeshi companies. Farooque et al. (2007) report that
corporate governance systems in Bangladesh are firmly based on family ownership, and in most firms, controlling families
dominate the boards, filling positions of executive directors, and CEOs. A survey conducted by Sobhan and Werner (2003) found
that an overwhelming majority (73 percent) of the boards of non-bank listed companies were heavily dominated by sponsor
shareholders ‘who generally belong to a single family-the father as the chairman and the son as the managing director is the
norm’ (Sobhan & Werner, 2003, p. 34). The study reported that such overpowering family dominance, along with lack of basic
knowledge regarding financial statements, had diminished the status of annual general meetings to that of a mere ritual.
Karim (2010) reports that although B4A firms are considered large firms in the context of Bangladesh, in the global context
these are not very large firms, with the number of partners ranging from 3 (for the affiliated firm of Ernst and Young) to 7 (for
the affiliated firm of Delloitte). Another interesting feature of the Bangladesh audit market is the absence of Big 4 and B4A
market power. In Bangladesh, the Big 4 and B4A firms command only 17 percent of listed audit clients and account for only 34
percent of client assets and 45 percent of client revenue (Karim, 2010). By contrast, Big 4 audit firms have a 41 percent market
share in India (International Accounting Bulletin, 2011). The lack of market power may be due to the direct absence of most
Big 4 firms in the audit market in Bangladesh.
Agency theory suggests that family firms either mitigate (alignment effect) or exacerbate (entrenchment effect) agency
problems. Consistent with the mitigation effect, because of the lesser type I agency problem family ownership motivates
family members to maximise the wealth of all shareholders and that opportunistic behaviour of family members for personal

3
Investigating the determinants of audit fees in South Asia, Ahmed and Goyal (2005) report that audit fees in Bangladesh are significantly lower than its
neighbouring India and Pakistan.
308 A. Khan et al. / The British Accounting Review 47 (2015) 304e320

gain is restricted by their long-term investment horizon, concern for their reputation and higher interest in the firm. As a
consequence extant research assumes families to be better monitors of managers than other types of large shareholders.
Therefore, it can be argued that because they are better monitors, family owners are less likely to demand higher quality audit
services and are willing to pay lower audit fees given a choice of higher quality auditors. The exacerbation argument, on the
other hand, suggests that because of the concentrated ownership family members can become entrenched in family firms and
can expropriate minority shareholders (type II agency problem). Wang (2006) argues that family firms usually have weak
corporate governance as a result of ineffective monitoring by the board. This is because influential positions in management
teams and board of directors are held by family members. As a result, possible managerial entrenchment is less likely to be
constrained as the people responsible for monitoring are possibly the same people who are behaving opportunistically.
Furthermore, entrenched family members insulate themselves from different disciplinary mechanisms and are less likely to
demand higher quality audit services and pay higher audit fees. Young et al. (2008), however, point out that the presence of
strong legal protection in developed countries reduces the scope for such expropriation. Rather, the possibility of such
expropriation is higher in developing countries with weaker shareholder protection and largely inactive markets.
Fig. 1 demonstrates the institutional context of Bangladesh and the hypotheses being tested in this paper. As mentioned
before, the corporate sector in Bangladesh is characterised by the overwhelming dominance of family firms, audit firm size
differentiation (Big 4, B4A and local firms), and the presence of important stakeholder groups in export industries. The
interaction between these factors may significantly affect audit fees and auditor choice in family firms.
In the context of Bangladesh, Khan et al. (2011) report a negative relationship between ownership concentration and audit
fees. However, the study does not directly examine the relationship between family ownership and control and audit fees in
Bangladesh. The expropriation argument would suggest that the absence of strong legal framework capable of protecting the
interests of minority shareholders in Bangladesh is likely to result in low demand for audit quality in the majority of family
firms. However, unlike many other developed countries, where previous studies have been conducted, the domination of
family firms in the Bangladeshi capital market is overwhelming. Therefore, compared to developed countries, where family
firms occupy a comparatively less important position in the capital market, firms owned by families in Bangladesh, due to
their dominant presence in the capital market, may require better quality audit services. Additionally, existing literature on
family ownership provides competing and alternative predictions about audit pricing in family firms. Therefore, the proposed
research hypothesis is:

H1: There is no significant difference in audit fees between family firms and non-family firms.

An interesting feature of the Bangladeshi audit market is the presence of Big 4, B4A and local audit firms. Siddiqui et al.
(2013) report that although the only Big 4 firm operating in Bangladesh (KPMG) earns an audit fee premium, the audit fees
charged by B4A firms are not significantly higher than local firms. Their findings suggest that the market does not perceive
B4A firms to be of similar quality to KPMG. However, the B4A firms are internationally linked and are relatively large in the
context of Bangladesh. Therefore, in terms of audit quality, B4A firms are most likely in between Big 4 and local audit firms. In
the context of Bangladesh, it would be interesting to explore whether family firms hire higher quality auditors. Thus the
following hypothesis is proposed:

H2: Demand for audit quality (choice of Big 4/B4A/local/specialist audit firms) is not significantly higher in family firms than
non-family firms.

The presence of important stakeholder groups in the Bangladeshi export industries presents an interesting dimension to
this research. As mentioned before, Bangladesh is currently the second largest exporter of ready-made garment (RMG)
products in the world. Most of the firms in the RMG sector are family-owned.4 Prior research (see Islam & Deegan, 2008; Khan
et al., 2013) acknowledge the role of foreign buyers in the RMG sector, which ensures greater financial reporting quality. It
might be reasonable to assume that these important stakeholders would also be concerned about the quality of audited
financial statements in Bangladesh, and influence the choice of audit firms (scenario 2, Fig. 1). The alignment effect would
imply that in order to maintain long term and sustainable relationship with the foreign buyers, family firms would forgo the
short term benefits of manipulating earnings. However, prior research documents that corporate governance mechanisms
have a largely symbolic role in family dominated firms in Bangladesh (Sobhan & Werner, 2003). Therefore, unlike in
developed economies, foreign buyers in Bangladesh would be less convinced of the positive impact of family ownership on
corporate governance. To reduce potential agency costs, family firms in the export industries may attempt to hire better
quality auditors and pay higher audit fees. However, the recent high-profile incident in Rana plaza that claimed the lives of
1100 workers in the RMG sector in Bangladesh has raised serious questions regarding the willingness and ability of these
foreign buyers to effectively monitor the governance practices in the RMG sector. To test whether foreign buyers in the RMG
sector influence audit fees and auditor choice, we develop the following two hypotheses:

4
91 percent of our sample companies from the RMG sector are family-owned.
 A. Khan et al. / The British Accounting Review 47 (2015) 304e320 309

Fig. 1. Institutional context and hypothesis development.

H3A: There is no significant difference in audit fees between family firms in the RMG sector and non-family firms.
H3B: Demand for audit quality (choice of Big 4/B4A/local/specialist audit firms) is not significantly higher in family firms in the
RMG sector than non-family firms.

3. Methodology

3.1. Data and sample

The sample selection procedure is reported in Table 1. The sample consists of all 155 non-financial companies listed on the
DSE from 2005 to 2013. Formed in 1954, the DSE is the leading of the two stock exchanges in Bangladesh, with a market
310 A. Khan et al. / The British Accounting Review 47 (2015) 304e320

Table 1
Sample description.

Panel A

Year Family Non-family Total

2005 84 45 129
2006 88 46 134
2007 84 45 129
2008 88 48 136
2009 76 45 121
2010 70 46 116
2011 67 45 112
2012 69 42 111
2013 42 28 70
1058

Panel B

Sector Family (firm- Non-family (firm- Total (firm- Percent of family firms (firm-
year year) year) year)
Cement 26 29 55 47.3
Ceramics 14 10 24 58.3
Engineering 110 67 177 62.1
Food 90 73 163 55.2
IT 16 31 47 34.0
Jute 16 9 25 64.0
Paper & printing 18 0 18 100.0
Miscellaneous 37 51 88 42.0
Pharmaceuticals 110 47 157 70.1
Service & real 18 28 46 39.1
estate
Tannery 12 32 44 27.3
Textile 194 20 214 90.7
Total 661 397 1058

capitalisation of Tk. 2953 billion as of 30 April 2014. Our sample begins in 2005 and ends in 2013. There were 282 listed
companies on the DSE in 2005. We exclude 337 firm-years due to the unavailability of the annual reports. We further exclude
14 companies because of non-availability of corporate governance data. This produces a final sample of 1058 firm-years.5
Table 1 reports the sample composition by year, firm ownership and industry sector. Panel A lists the sample firms by
ownership type (family vs. nonfamily) and by year. The sample firms vary by year: 129 firms (84 family firms) in 2005 to 70
(42 family firms) in 2013. From Panel B of Table 1, it is observed that family firms represent 63 percent of the total sample. This
indicates that family firms are the dominant form of company listed on the DSE. In our sample, the paper and printing in-
dustry appears to have the highest percentage of family firms (100%), followed by the RMG industry (91%).

3.2. Measuring family ownership and control

In this study, we adopt multiple criteria to identify family firms. This is mainly because the prior literature provides no
commonly accepted measure or criterion for identifying a family firm. La Porta et al. (1999) argue that the 20 percent cut-off
point is usually enough to have effective control of a firm and that this cut-off point is used by a number of studies (see
Bartholomeusz & Tanewski, 2006; Setia-Atmaja, Haman, & Tanewski, 2011; Setia-Atmaja, Tanewski, & Skully, 2009).
Following Cascino, Pugliese, Mussolino, and Sansone (2010), we use a more refined definition of family firms that does not
uniquely rely on ownership concentration as major determining criterion.6 So we identify family firms as being (i) firms in
which 20 percent of a firm's share or voting rights (either direct or indirect) are held by family block holders, and (2) at least
one member of the controlling family holds a managerial position such as board member, CEO or chairman.
We use a dummy variable and set it equal to 1 if the firm is considered a family firm and 0 otherwise. Family relationships
and shareholding patterns were collected from prospectuses of the listed companies, annual reports and company websites.
Under the Bangladesh Security and Exchange Commission (SEC) notification, listed companies are required to disclose the
pattern of shareholdings. This includes a number of shares held by parent/subsidiary/associate companies, the directors, Chief
Executive Officer, Company Secretary, Chief Financial Officer, Head of Internal Audit and their spouse and minor children.
Thus in the present study, if a firm has at least one such shareholder or family member controlling 20 percent or more
shareholdings, it is considered a family firm.

5
Initial sample ¼ 155 firms times 9 years ¼ 1395 firm-years; Final sample ¼ (1395  337) ¼ 1058.
6
Cascino et al. (2010) argued that choosing a certain percentage threshold would not allow any difference in terms of family and non-family firms to be
captured because a certain percentage threshold only represents high ownership concentration rather than family ownership and management.
 A. Khan et al. / The British Accounting Review 47 (2015) 304e320 311

3.3. Model

We use two models for the purpose of identifying the determinants of audit fees and auditor choice in this study. They are
discussed below:

3.3.1. Audit fees model

We use the following an OLS regression techniques to examine the relationship between a family firm and audit fees.

AUDIT FEE ¼ b0 þ b1 FAMILYCON þ b2 FIRM SPECIFIC CONTROLS þ b3 INDUSTRY CONTROLS þ b4 YEAR CONTROLS
þ3
(1)

The dependent variable audit fees (AUDIT FEE) is measured by taking the natural log of audit fees. The key variable family
firm (FAMILYCON) is an indicator variable that equals 1 if the firm is a family firm and 0 otherwise. When we examine the
relationship between family ownership and audit fees, this dummy variable is replaced by a family ownership variable
(FAMILYOWN), calculated by taking the percentage of shares owned by the family members. The control variables used in
equation (1) are: non-family insider ownership (INSIDEOWN); institutional ownership (INSTOWN); government ownership
(GOVOWN); board independence (BOARDIND); audit complexity (AUDCOMPLX); firm size (SIZE); leverage (LEV), auditor type
(AQ and INDSPEC); profitability (PROF); subsidiaries dummy (SUBD); year-end dummies (END); international link dummy
(MNC); and, non-audit fees dummy (NAFD) variables. We also introduce industry dummies (INDUM). The definitions of the
variables have been provided in Appendix 1. We use two proxies to measure auditor type variable. The first proxy captures
auditor quality (AQ) based on three different types of auditor in Bangladesh. The second proxy is an industry specialist auditor
(INDSPEC) variable. An auditor is considered an industry specialist auditor if that auditor is ranked among the top two au-
ditors in a particular industry in a given year, based on the amount of total audited assets in Bangladesh.
Insider ownership, institutional ownership and government ownership variables are controlled to address the impact of
ownership by other groups of stakeholders on audit pricing. Consistent with the alignment argument, a negative relationship
is predicted for insider ownership and audit fees. Institutional and government owners are usually perceived as monitoring
agents. Therefore, auditors assess a lower level of inherent risk and charge lower audit fees. Independent directors are likely to
be effective monitors and therefore firms with independent directors are likely to incur lower audit fees. It is also possible that
independent directors might require more rigorous audits to protect the interest of general shareholders and hence firms
incur higher audit fees. Audit complexities and the number of subsidiaries involve greater loss exposure in terms of audit risk,
which results in higher audit fees (Simon & Taylor, 1997). Previous studies suggest that size is an important determinant of
audit fees (Ahmed & Goyal, 2005; Karim & Moizer, 1996; Simunic, 1980). Auditor quality is likely to have a positive impact on
audit fees. For example, it is expected that higher quality auditors such as Big 4 audit firms will charge a premium for their
quality (Beattie, Goodacre, Pratt, & Stevenson, 2001; Firth, 1997). Low profitability could be associated with financial pressure
which would require more audit efforts to confirm that a company is not a going concern (Karim & Moizer, 1996). Ahmed and
Nicholls (1994) note that subsidiaries of multinational companies operating in Bangladesh exhibit higher accounting and
reporting standards. Consequently, these companies are expected to have higher quality audits as well as pay higher audit
fees. Therefore, a positive relationship is expected between international links and audit fees. In Bangladesh the busy period
occurs after June because of the end of fiscal year. Consequently, companies with accounting periods ending in June are
considered to be busy season clients and it is expected that they would pay a premium. So we allow for this condition.
Prior research indicates that audit fees and non-audit fees may be determined jointly (see for example, McMeeking, Pope,
& Peasnell, 2006; Simunic, 1984) and that such joint provisions may have implications on auditor independence. Unlike many
developed economies, purchase of non-audit services from auditors is not very common in Bangladesh. As only around 20% of
our sample companies purchase such services from their auditors (Table 3), a dummy variable (instead of logarithm of non-
audit service fees) is used as a proxy of non-audit fees.

3.3.2. Auditor choice model

We use the following ordered probit model to examine the relationship between family firm and the choice of auditor.

AUDITOR CHOICE ¼ a0 þ a1 FAMILYCON þ a2 FIRM SPECIFIC CONTROLS þ a3 INDUSTRY CONTROLS

þ a4 YEAR CONTROLS þ 3 (2)

We use two alternative proxies for auditor choice model. The first proxy captures the impact of family firms on recruiting
higher quality auditor (AQ) based on three types of auditors in Bangladesh: Big 4 auditor, B4A and local/national auditor.
Auditor quality is a categorical variable and is coded “3” for a big 4 auditor, “2” for a B4A auditor and “1” for a local auditor.
When we use the aforesaid categorical variable we use an ordered probit model (refer to as Models 1 and 2 in Table 5).7 In

7
We thank an anonymous reviewer for providing us helpful suggestions to develop the categorical variable and use ordered probit model.
312 A. Khan et al. / The British Accounting Review 47 (2015) 304e320

response to Hay's (2013) call for using alternative proxies for audit firm differentiation, we use the likelihood of recruiting
industry specialist auditor as another proxy of auditor choice model. Industry specialist auditor variable (INDSPEC) is a
dummy variable equals 1 if the auditor is an industry specialist auditor and otherwise 0. When we use the alternative proxy
(INDSPEC) for auditor choice we run a logit regression for the auditor choice model (refer to as Models 3 and 4 in Table 5).8 The
definition of family firm (FAMILYCON) variable is similar to the definition we used in our audit fees model. The control
variables used in equation (2) are non-family insider ownership (INSIDEOWN), institutional ownership (INSTOWN), gov-
ernment ownership (GOVOWN), audit complexity (AUDCOMPLX), firm size (SIZE), and leverage (LEV), profitability (PROF)
and international link dummy (MNC) variables. The definitions of the control variables are similar to the definitions we used
in the audit fees model. Insider ownership, institutional ownership and government ownership are controlled to address the
impact of other stakeholders on auditor choice. Because independent directors are effective monitors, boards dominated by
independent directors might require high quality auditors to protect the interests of general shareholders.
Grossman and Hart (1982) suggest that leverage can induce managers to avoid value-decreasing decisions if higher
leverage increases managers' threat of bankruptcy and loss of control. The level of leverage may influence the choice of
auditor. Following prior studies, controls for differential audit complexity are included in the analyses (Chaney, Jeter, &
Shivakumar, 2004; Piot, 2005). The variables size and complexity are included because audit effort is expected to increase
with the size and complexity of a client-firm's operations. Because Since auditor change is related to its profitability (see
Johnson & Lys, 1990), we include a profitability (PROF) measure, defined as earnings before interest and taxes to total assets to
control for the potential effect of profitability.

4. Empirical results

4.1. Summary statistics

In Table 2 we report the summary statistics of the full sample. In Panel A we present the mean, median and standard
deviation of the different variables. The mean audit fee for our sample companies is Taka. (Tk.) 116,726 (US $ 1496).9 Using
data from 2003 to 2005, Khan et al. (2011) found a mean audit fee of Tk. 67,480 (US $ 865). The mean family ownership is
approximately 30 percent. The mean ownership by non-family insiders is 9.5 percent and institutional investors and gov-
ernment are 15 percent and 5 percent, respectively. The mean firm size is Tk. 2.450 million (US $ 31 million).
Panel B of Table 2 presents the difference of means tests for key variables between family and non-family firms. Among
family firms 92 percent of the CEOs are family members. This is consistent with Sobhan and Werner (2003) who find that
family firms in Bangladesh typically appoint family members as CEOs. Board independence is statistically indistinguishable
between family and non-family firms. On average, non-family firms pay higher audit fees than family firms. Institutional
ownership is more prevalent in non-family firms than in family firms although the difference is statistically insignificant.
Average government ownership of non-family firms is significantly higher than family firms. On average, 18.4 percent of
family firms and 23.2 percent of non-family firms recruit industry specialist auditors.
In Table 3 we present the Spearman correlation matrix for the variables. Family firms have a negative relationship with
audit fees. This variable is also negatively correlated with higher quality auditors. Family firms are also positively correlated
with profitability. Consistent with previous studies (see Anderson & Reeb, 2003; Setia-Atmaja et al., 2009), we find that family
ownership is negatively related to firm size and leverage. The results also show that higher quality auditors and industry
specialist auditors are positively correlated with audit fees.

4.2. Relationship between family firms and audit fees

We now present the results for our OLS and ordered probit models. The residuals of all our models passed several tests for
heteroskedasticity and non-normality. Specifically, plots of the residuals did not show any reason for concern. Similarly, the
Huber and the White Sandwich tests did not indicate any violation of assumptions.
We report the results of the audit fee regressions in Table 4 for models 1 to 6. The adjusted-R squared for the models range
from 61.1% to 64.7%. Although this is lower compared to studies conducted in developed economies, our models compare well
with those of developing economies (see Ahmed & Goyal, 2005; Khan et al., 2011). Ahmed and Goyal (2005) note that the low
explanatory power of audit fees models in the context of developing countries may be attributed to the omission of variables.
Karim and Moizer (1996) indicate that Bangladeshi companies are smaller in size and, hence, a more relevant comparison
should be made with studies that use small company samples. The dependent variable is the natural log of audit fees (AUDIT
FEE). We find evidence that family firms pay lower audit fees than non-family firms. Specifically, we find that the coefficient
estimate for family firms (FAMILYCON) is negative and significant (b1 ¼ 0.168, p < 0.01). Therefore, H1 is rejected. This
implies that family firms are less likely to demand extensive audit and incur higher audit fees. This is consistent with the
entrenchment argument and suggests that family firms dominated by family members care very little about minority

8
Logit analysis is used because the dependent variable is a binary measure of auditor choice. Furthermore, a logit is considered to be as efficient as probit
but does not require normality of parameter distribution (e.g., Carey, Simnett, & Tanewski, 2000).
9
As of July 1, 2014, 1US$ ¼ Tk. 78.
 A. Khan et al. / The British Accounting Review 47 (2015) 304e320 313

Table 2
Summary statistics.

Panel A: full sample

Mean Median Std. dev. JarqueeBera

AUDCOMPLX 0.344 0.332 0.210 59.306a
BOARDIND 0.097 0.111 0.087 25.975a
INDSPEC 0.196 0.000 0.397 419.347a
INSIDEOWN 0.095 0.020 0.306 19,827,389a
FAMILYCON 0.629 1.000 0.483 189.477a
FAMILYOWN 0.297 0.311 2.696 3,075,100a
FAMCEO 0.599 1.000 0.492 189.527a
INSTOWN 0.151 0.130 0.127 111.344a
GOVOWN 0.048 0.000 0.173 11,037.02a
LEV 0.611 0.502 0.698 83,323.38a
SIZE (in million Tk.) 2450.000 699.000 6030.000 32,6676.3a
NAFD 0.203 0.000 0.402 403.236a
PROF 0.066 0.063 0.122 330,203.3a
AF (in Tk.) 116,725.900 60,000.000 161,233.600 14,543.13a
SUBD 0.297 0.000 0.457 219.350a
END 0.518 1.000 0.500 185.501a
MNC 0.056 0.000 0.231 10,397.87a
AQ 1.159 1.000 0.456 4562.534a

Panel B: family and non-family firm subsamples

Family Non-family t-Test statistic

AUDCOMPLX 0.356 0.324 2.779a
INDSPEC 0.182 0.234 1.998b
BOARDIND 0.096 0.100 0.265
INSIDEOWN 0.052 0.168 9.327a
INSNTOWN 0.150 0.151 0.463
GOVOWN 0.008 0.130 7.412a
LEV 0.596 0.636 3.667a
SIZE (in million Tk.) 2471.756 2400.591 1.375
NAFD 0.137 0.318 7.184a
PROF 0.069 0.061 0.690
AF (in Tk) 103,982.257 138,348.584 3.502a
END 0.524 0.515 0.289
MNC 0.000 0.153 9.641a
FAMCEO 0.915 0.066 7.778a
SUBD 0.411 0.119 6.239a
AQ 1.123 1.229 4.444a
a,b
and c represent significance test statistic at the 1%, 5% and 10% level, respectively.

shareholders. The statistical significance of some of the other control variables suggests that audit fees are also influenced by
other factors. In particular, the positive and significant coefficient for auditor quality (AQ) suggests an audit fee premium for
higher quality auditors (Beattie et al., 2001; Firth, 1997). A negative and significant coefficient for insider ownership (INSI-
DEOWN) is consistent with the entrenchment argument. We also find positive and significant relationships for audit
complexity (AUDCOMPLX) and subsidiary dummy (SUBD) variables, implying that auditors charge higher audit fees because
of greater client complexity (Simon & Taylor, 1997). The positive and significant coefficient of firm size (SIZE) suggests that
larger firms incur higher audit fees than smaller firms (Simunic, 1980). We also find a positive and significant relationship
between board independence (BOARDIND) and audit fees, suggesting that despite concerns over board independence in
Bangladesh, independent boards actually tend to employ better quality auditors and pay higher audit fees. As expected, we
find a positive and significant coefficient of MNC dummy variable. Finally, we also find a positive and significant coefficient for
non-audit fees (NAFD). This indicates that companies purchasing non-audit services from incumbent auditors pay higher
audit fees than those that do not use such services.
Because previous research finds that higher quality auditors receive higher audit fees from their clients in developing
countries (Ahmed & Goyal, 2005), we use an interaction variable for family firms (FAMILYCON) and auditor quality (AQ) to test
for this effect. Model 2 in Table 4 shows a positive and significant coefficient for FAMILYCON*AQ implying that higher quality
auditors charge a premium to family firms. The results contradict prior UK and USA studies (see Chaney et al., 2004; Fortin &
Pittman, 2007) that report the absence of Big 4 audit fee premium for family firms. However, those studies typically
concentrate on private family firms whereas the family firms used in our sample are comparatively large in terms of market
capitalisation and are publicly listed companies. A further explanation for the contrasting result is that high quality auditors in
developing countries may face less competition for their services (compared to the competition in developed countries),
thereby being able to charge higher fees in developing countries.
 314
A. Khan et al. / The British Accounting Review 47 (2015) 304e320
Table 3
Spearman rank correlation matrix.

Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)
(1) AUDCOMPLX 1.000
(2) BOARDIND 0.127a 1.000
(3) INDSPEC 0.043 0.069b 1.000
(4) INSIDEOWN 0.065b 0.008 0.003 1.000
(5) FAMYLYCON 0.079c 0.041 0.063b 0.184a 1.000
(6) INSTOWN 0.001 0.038 0.055 0.022 0.012 1.000
(7) GOVOWN 0.031 0.165 0.026 0.066b 0.351a 0.194a 1.000
(8) LEV 0.005c 0.221 0.072 0.012 0.025 0.042 0.301a 1.000
(9) SIZE 0.055a 0.133a 0.082b 0.003 0.008 0.065b 0.037 0.115 1.000
(10) NAFD 0.005 0.122a 0.137a 0.026 0.222a 0.004 0.114b 0.135a 0.057c 1.000
(11) PROF 0.125a 0.144a 0.152a 0.006 0.001 0.053 0.150 0.347b 0.089c 0.252 1.000
(12) AF 0.008b 0.286a 0.196a 0.029 0.120a 0.018 0.062b 0.169 0.689c 0.156a 0.267a 1.000
(13) SUBD 0.044a 0.027 0.130a 0.122c 0.299a 0.037 0.179 0.057c 0.312a 0.050 0.105a 0.342a 1.000
(14) END 0.082 0.122a 0.171 0.019 0.029 0.043 0.176a 0.065b 0.239a 0.175a 0.152a 0.331b 0.200c 1.000
(15) MNC 0.043c 0.230 0.117a 0.009b 0.330a 0.042 0.058c 0.085a 0.170a 0.396a 0.287a 0.466a 0.074b 0.258b 1.000
(16) AQ 0.083b 0.214b 0.339a 0.007 0.143a 0.003 0.019 0.053 0.167a 0.374a 0.306a 0.436a 0.150a 0.308c 0.532b 1.000
a,b c
and represent significance of the coefficients at the 1%, 5% and 10% level, respectively.
 A. Khan et al. / The British Accounting Review 47 (2015) 304e320 315

Table 4
Regression results: relationship between family firms and audit fees.

Variable Model 1 (OLS) Model 2 (OLS) Model 3 (OLS) Model 4 (OLS) Model 5 (OLS) Model 6 (OLS)

AUDIT FEE (H1) AUDIT FEE (H1) AUDIT FEE (H3A) AUDIT FEE (H1) AUDIT FEE (H1) AUDIT FEE (H3A)

Coefficient Prob. Coefficient Prob. Coefficient Prob. Coefficient Prob. Coefficient Prob. Coefficient Prob.
Intercept 4.951a 0.000 4.679a 0.000 4.277a 0.000 5.241a 0.000 4.049a 0.000 4.419a 0.000
FAMILYCON 0.168a 0.009 0.508a 0.001 0.035 0.604 0.139b 0.047 0.124b 0.018 0.005b 0.038
AQ 0.466a 0.000 0.254b 0.030 0.416a 0.000
FAMILYCON*AQ 0.261b 0.045
INDSPEC 0.258a 0.000 0.113a 0.002 0.246a 0.000
FAMILYCON*INDSPEC 0.189b 0.048
a a a a a
INSIDEOWN 0.609 0.001 0.598 0.002 0.581 0.002 0.628 0.003 0.019 0.754 0.647 0.001
INSTOWN 0.014 0.939 0.152 0.381 0.040 0.814 0.258 0.174 0.144 0.352 0.257 0.151
GOVOWN 0.001 0.997 0.014 0.935 0.011 0.948 0.066 0.724 0.266c 0.055 0.062 0.728
AUDCOMPLX 0.325a 0.002 0.446a 0.000 0.382a 0.000 0.371a 0.001 0.331a 0.000 0.408a 0.000
LEV 0.040 0.235 0.016 0.638 0.013 0.695 0.044 0.217 0.050 0.127 0.015 0.666
SIZE 0.288a 0.000 0.306a 0.000 0.318a 0.000 0.296a 0.000 0.341a 0.000 0.328a 0.000
BOARDIND 0.238b 0.041 0.887a 0.001 0.664b 0.025 0.383 0.238 0.491a 0.000 0.789b 0.010
PROF 1.283a 0.000 1.298a 0.000 0.991a 0.000 1.366a 0.000 1.456a 0.000 1.140a 0.000
END 0.053 0.289 0.030 0.530 0.044 0.349 0.106b 0.043 0.144a 0.001 0.103b 0.034
SUBD 0.165a 0.003 0.184a 0.001 0.103c 0.056 0.164a 0.007 0.190a 0.000 0.098c 0.085
MNC 0.406a 0.002 0.561a 0.001 0.460a 0.000 0.846a 0.000 0.725a 0.000 0.814a 0.000
NAFD 0.181a 0.004 0.182a 0.004 0.152a 0.011 0.117c 0.078 0.169a 0.002 0.118c 0.058
EXPORT 0.121b 0.047 0.161b 0.011
FAIMILYCON*EXPORT 0.252a 0.001 0.219b 0.043

Industry dummies Included Included Included Included Included Included

Year dummies Included Included Included Included Included Included
Adjusted R-squared 0.631 0.611 0.640 0.612 0.647 0.631
F-statistic 48.127a 69.802a 60.988a 41.305a 52.332a 54.625a
a,b
and c represent significance of the coefficients at the 1%, 5% and 10% level respectively using two-tailed tests.

A number of recent studies (see, Islam & Deegan, 2008; Khan et al., 2013) have highlighted the influence of important
stakeholders on financial reporting disclosures in Bangladesh. As stated before, Bangladesh is currently the second largest
exporter of textile/garment products in the world, and many family firms operating in the garments sector act as supply
chains for renowned international brands. It is therefore possible that international buyers could be concerned about the
quality of the audited financial statements and this may influence the choice of audit firms. To test this hypothesis, we include
an interaction variable based on family firms (FAMILYCON) and RMG industry variable (EXPORT). Models 3 and 6 in Table 4
show positive and significant coefficients for FAMILYCON*EXPORT variable. Thus, H3A is rejected. Our result implies that
family firms in the export-oriented RMG sector pay significantly higher audit fees, perhaps to allay concerns of their inter-
national buyers.10
We also explore the impact of auditor type on audit fees by using an alternate proxy. In particular, we create an industry
specialist dummy variable (INDSPEC) and re-run the original models. The results are presented in Models 4e6. We document
a positive and significant impact of industry specialist auditor on audit fees. We also document that these audit firms obtain
fee premiums from family firms. Our overall findings suggest the influence of industry specialist auditors in emerging
markets.

4.3. Relationship between family firms and quality auditor choice

We report the results of ordered probit and logit regressions with respect to auditor choice in Table 5. For the ordered
probit regression (Models 1 and 2) our dependent variable is the first proxy of auditor choice, i.e., auditor quality (AQ). In
Model 1 we examine the impact of family firms on recruiting higher quality auditors. We find that family firms are less likely
to hire higher quality auditors than non-family firms. Specifically, we find that the coefficient estimate of family firms
(FAMILYCON) is negative and significant (a1 ¼ 0.520, p < 0.01). Therefore, H2 is rejected. Consistent with the entrenchment
argument this finding also implies that family firms are less likely to demand quality audit services. So they are more likely to
recruit lower quality auditors compared to their non-family counterparts. The statistically significant coefficients of the other
variables suggest that the choice of a firm's higher quality auditor is also influenced by other factors. In particular, insider
ownership (INSIDEOWN) is negative and significant for H2. So firms with a higher percentage of insider ownership are less
likely to have higher quality auditors. Our results also show positive and significant relationships for firm size (SIZE) and

10
We also use a three way interaction between family firms (FAMILYCON), auditor quality (AQ) and RMG industry variable. We discuss the result as a part
of additional analysis.
316 A. Khan et al. / The British Accounting Review 47 (2015) 304e320

Table 5
Regression results: relationship between family firms and auditor choice.

Variable AQ (H2) AQ (H3B) INDSPEC (H2) INDSPEC (H3B)

Model 1(ordered probit) Model 2 (ordered Model 3 (logit) Model 4 (logit)

probit)

Coefficient Prob. Coefficient Prob. Coefficient Prob. Coefficient Prob.

Intercept 9.797 0.000 8.631 0.000 4.255 0.000 3.435 0.004
FAMILYCON 0.520a 0.000 0.436a 0.003 0.354b 0.039 0.643a 0.006
INSIDEOWN 0.021c 0.056 0.006 0.957 0.099 0.713 0.095 0.726
INSTOWN 1.129 0.228 1.170 0.329 1.277 0.164 1.129 0.170
GOVOWN 0.734 0.238 0.895 0.313 0.747 0.214 0.867 0.153
AUDCOMPLX 0.435 0.165 0.367 0.256 0.332 0.405 0.395 0.325
LEV 0.407a 0.000 0.412a 0.000 0.102 0.592 0.117 0.540
SIZE 0.434a 0.000 0.424a 0.000 0.145a 0.009 0.118b 0.035
BOARDIND 1.061b 0.041 0.901b 0.030 0.325b 0.040 0.604c 0.059
PROF 4.743a 0.000 4.817a 0.000 3.269a 0.001 3.203a 0.001
MNC 1.404a 0.000 1.613a 0.000 0.047c 0.094 0.034c 0.070
EXPORT 0.434b 0.070 0.761b 0.045
FAMILYCON*EXPORT 0.857 0.001 0.822a 0.007

Industry dummies Included Included Included Included

Year dummies Included Included Included Included
Pseudo R-squared 0.334 0.343 0.184 0.172
a,b
and c represent significance of the coefficient at the 1%, 5% and 10% level respectively using two-tailed tests.

profitability (PROF), implying that the choice of higher quality auditors is significantly related to size and profitability. We also
find that multinational companies (MNC) are more likely to recruit higher quality auditors.
Because of international stakeholders' pressure family firms could produce higher quality audited financial statements, we
explore whether family firms in the garments export-oriented sector are more likely to recruit higher quality auditors. To test
this prediction, we include the interaction variable FAMILYCON*EXPORT. Model 2 of Table 5 shows a positive and significant
coefficient for FAMILYCON*EXPORT. This indicates that family firms in the export-oriented sector are more likely to recruit
higher quality auditors. Hence, H3B is rejected.
We then explore the impact of family firms on auditor choice by using alternate proxy, i.e., the likelihood of recruiting
industry specialist auditors using logit regressions (Models 3 and 4). We use an industry specialist dummy (INDSPEC) as our
dependent. Models 3 and 4 show that family firms are generally less likely to recruit industry specialist auditors. However,
when these firms belong to the export-oriented sector, they are more likely to recruit industry specialist auditors.

4.4. Additional analysis

4.4.1. Family ownership, family CEO and audit fees and auditor choice
Previous studies have examined the impact of corporate governance on audit fees. Hay, Knechel, and Ling (2008) docu-
ment a positive relationship between measures of corporate governance and audit fees for New Zealand firms. Like many
other developing countries, Bangladesh has also adopted western-styled corporate governance codes, requiring the sepa-
ration of the chair and the CEO. However, Sobhan and Werner (2003) report that in the majority of listed companies in
Bangladesh, the chair and the CEO are recruited from the same family, significantly reducing the effectiveness of an important
corporate governance scheme such as CEO-chair separation. 59.9 percent of our sample family firms had CEOs recruited from
the same family. We test the impact of family ownership and family CEO on audit fees and auditor choice and report the
results in Table 6.
We re-run our audit fees model by replacing the FAMILYCON variable with FAMILYOWN. We also incorporate a family CEO
dummy variable (FAMILYCEO) if the CEO is a family CEO. Model 1 of Table 6 shows a negative and significant coefficient for
each of the FAMILYOWN and FAMILYCEO variables. A negative and significant coefficient of FAMILYOWN variable suggests
that as the family ownership increases families tend to incur lower audit fees. A negative and significant coefficient of
FAMILYCEO variable implies that family firms incur lower audit fees when a family member serves as a CEO. This is consistent
with the notion that unskilled and inefficient family CEOs representing the interest of the families do not want to spend
higher audit fees to ensure better monitoring through an extensive audit. From a socio-economic context, family firm CEOs
may also be reluctant to pay higher audit fees if minority shareholders can be overpowered in the annual general meetings. In
Model 2 we re-run the same regression replacing our auditor quality (AQ) variable by an industry specialist auditor dummy
variable (INDSPEC). Our results in regard to family ownership (FAMILYOWN) and family CEO variables (FAMILYCEO) remain
unchanged.
Panel B of Table 6 shows that results for the auditor choice model after replacing FAMILYCON with the FAMILYOWN
variable. We also incorporate a family CEO dummy (FAMILYCEO) variable in the same model. Model 3 shows a negative and
significant coefficient for FAMILYOWN. This implies that firms with a larger percentage of family ownership are less likely to
 A. Khan et al. / The British Accounting Review 47 (2015) 304e320 317

Table 6
Regression results: relationship between family ownership and family CEO and audit fees and auditor choice.

Variable Panel A Panel B

AUDIT FEE (H1) AUDIT FEE (H1) AQ (H2) INDSPEC (H2)

Model 1 (OLS) Model 2 (OLS) Model 3(ordered probit) Model 4 (logit)

Coefficient Prob. Coefficient Prob. Coefficient Prob. Coefficient Prob.

Intercept 4.927a 0.000 5.219a 0.000 9.919a 0.000 4.063a 0.001
FAMILYOWN 0.436b 0.010 0.480a 0.008 0.013b 0.032 0.065b 0.016
FAMCEO 0.293a 0.000 0.287a 0.000 0.427b 0.011 0.448b 0.022
AQ 0.476a 0.000
INDSPEC 0.259a 0.000
INSIDEOWN 0.407b 0.025 0.435b 0.028 0.033 0.769 0.104 0.706
INSTOWN 0.064 0.715 0.312 0.194 1.044 0.239 1.283 0.162
GOVOWN 0.115 0.494 0.179 0.330 0.835 0.427 0.857 0.147
AUDCOMPLX 0.291a 0.005 0.339a 0.003 0.414 0.190 0.460 0.259
LEV 0.028 0.405 0.033 0.352 0.429 0.179 0.100 0.596
SIZE 0.288a 0.000 0.297a 0.000 0.438a 0.000 0.138a 0.012
BOARDIND 0.210 0.495 0.365 0.258 1.222b 0.088 0.328b 0.039
PROF 1.108a 0.000 1.199a 0.000 4.821a 0.000 3.105a 0.002
END 0.061 0.214 0.114b 0.028
SUBD 0.139b 0.013 0.138b 0.021
MNC 0.476a 0.000 0.919a 0.000 1.474a 0.000 0.032 0.196
NAFD 0.188a 0.003 0.123c 0.062

Industry dummies Included Included Included Included

Year dummies Included Included Included Included
Adjusted R-squared 0.636 0.618
F-statistic 47.246a 40.773a
Pseudo R-squared 0.327 0.132
a,b
and c represent significance of the coefficients at the 1%, 5% and 10% level respectively using two-tailed tests.

undergo a rigorous audit and are thus likely to select lower quality auditors. This is also consistent with our main findings in
regards to family firms. We also document a negative significant coefficient of FAMILYCEO variable. This is consistent with the
notion that entrenched family CEOs are less likely to recruit higher quality auditors.
Model 4 shows the results of the logit regression using the industry specialist auditor dummy variable (INDSPEC) as the
dependent variable. Our results with regards to family ownership (FAMILYOWN) and family CEO (FAMILYCEO) variables
suggest that firms with a greater percentage of ownership by family members and family CEOs are less likely to recruit in-
dustry specialist auditors. Overall, our results suggest that the efficacy of a western-styled corporate governance mechanism
such as the separation of CEO and chair is significantly compromised due to the tendency of selecting the chair and CEO from
the same family in Bangladesh.

4.4.2. Family firms, auditor quality and garment/textile industry variables

In our audit fees model we introduce a three-way interaction between family firms, auditor quality and RMG industry
variables (FAMILYCON*AQ*EXPORT). Our objective is to explore the impact of family firms in the RMG sector with higher
quality auditors on audit fees. We re-run the audit fees model using an OLS regression. Our unreported results suggest a
positive and significant coefficient of the interaction variable indicating that family firms in the export sector with higher
quality auditors pay higher audit fees. We run another regression replacing the previous three way interaction variable by
FAMILYCON*INDSPEC*EXPORT interaction variable to test the impact of family firms in RMG sector with industry specialist
auditor on audit fees. Once again, we find a positive and significant coefficient for the three way interaction variable, which
implies that family firms in the RMG industry with industry specialist auditors incur higher audit fees.

4.5. Sensitivity analysis

We conducted a number of tests to check robustness of our results. First, we use an alternative definition of family firms. In
particular, we define a firm as family controlled when an individual, or group of family members, hold at least 50 percent of a
firm's share (voting rights). We use a dummy variable to identify the firms and are set equal to 1 if the firm is considered to be
family firm and 0 otherwise. When we use this alternative ownership threshold the number of family firms is reduced to 283.
The results for the FAMILYCON variable remain unchanged both in audit fees and auditor choice model. Second, we take into
account the possibility that the determinants of audit fees and auditor choice may be different in family firms as opposed to
non-family firms. The results for family-subsamples are generally consistent with our results for the full sample. This is
possibly due to the dominance of family firms in the Bangladeshi corporate sector. Third, motivated by the findings of family
ownership-performance literature we test for a nonlinear impact of family ownership on audit fees and the auditor choice
model. In particular, we use a quadratic specification of family ownership variable. However, we fail to document any
318 A. Khan et al. / The British Accounting Review 47 (2015) 304e320

significant non-linear impacts of family ownership variable on our key variables of interest. Finally, we repeat all the analyses
using a fixed effect model. We do not find any qualitative difference to our original results.

5. Conclusions

The paper investigates audit fees and auditor choice in family firms in Bangladesh. Unlike many developed countries,
family firms are the dominant form of listed public companies in Bangladesh, and the majority of leading companies listed in
the DSE are actually family firms. We find that compared to non-family firms, family firms in Bangladesh tend to pay lower
audit fees and are less likely to hire top quality audit firms. However, family firms in export oriented industries are inclined to
pay higher audit fees and recruit better quality auditors. Our results also reveal that firms with CEOs11 from the same family
tend to pay significantly less audit fees compared to firms with hired CEOs. In addition, we identify a number of important
determinants of audit pricing and auditor choice and provide first-hand evidence of the joint determination of audit and non-
audit service fees in an emerging economy context.
Our study is among the first to investigated audit fees and auditor choice in family firms in the context of developing
economies. Previously, Ho and Kang (2013) reported that family firms in the USA had a tendency to recruit lower quality
auditors and incur lower audit fees. This suggests that higher audit efforts in family firms in the US (in order to avoid litigation
risks) are offset by lower demand for quality audit (due to alignment effect). We should note however that firms, whether
family or non-family listed, are operating in a more heavily regulated environment in developed countries. Audit firms are
also operating under more competitive business conditions in developed countries. The absence of these conditions in
developing countries, may lead to greater litigation risk and high fees for family firms in developing countries. However, audit
markets in developing countries such as Bangladesh are characterised by a relatively unsophisticated legal and regulatory
environment (Farooque et al., 2007) and lack of investor (including minority shareholders) protection (Ashraf & Ghani, 2005).
In the entire history of Bangladesh, a legal action has never been taken against an auditor for professional negligence (World
Bank, 2002). Young et al. (2008) argue that in such a context, poor demand for audit quality in family firms is likely to be
driven by the expropriation effect rather than the alignment effect, as demonstrated in earlier studies conducted in developed
markets. Our findings contradict with Fan and Wong (2005), who report a positive association between higher ownership
concentration and auditor choice in emerging economies. However, Fan and Wong (2005) use a sample of companies from
East Asian countries where ‘legal institutions provide some degree of help in reinforcing and monitoring auditors' actions’ (p.
42). In case of Bangladesh, the legal institutions are much weaker, as evidenced by the sheer absence of legal actions taken
against auditors. In such an environment, external auditors would not be expected to perform a corporate governance
function (Francis et al., 2003).
The presence of important stakeholders in export-oriented industries presents an important dimension to the study. Our
results indicate that, consistent with the alignment argument, managers in export-oriented family firms appear to be more
inclined towards paying higher audit fees and appointing better quality auditors. We hence provide further empirical support
to prior studies (such as, Islam & Deegan, 2008; Khan et al., 2013) suggesting that in the absence of an active capital market
and strong legal protection, important stakeholder groups in the form of foreign buyers play a corporate governance role in
Bangladesh. In addition, we extend prior studies such as Siddiqui (2010) by providing empirical evidence that the efficacy of
important western styled corporate governance mechanisms such as CEO-chair separation can be significantly compromised
in Bangladesh.
Our findings have a number of important implications for the audit market in developing countries. As mentioned before,
there has been recent phenomenal growth in the number family firms, especially in emerging economies in Asia. For some of
the emerging economies in Asia, family firms now provide the backbone for the economy, making up to 65 percent of the
listed companies, and 49 percent of market capitalisation. Therefore, the sustainability of family firms is crucial for economic
growth in these regions. However, due to weaknesses in corporate governance and internal control mechanisms, the majority
of these family firms tend to perish early. As our findings indicate, the majority of the family firms are reluctant to recruit
quality audit firms and pay higher audit fees, hence increasing the risk of poor financial management. For long-term survival
of these family firms, it is important that these companies address important issues surrounding their corporate governance
structures and incorporate stronger monitoring mechanisms in the form of external audit. As demonstrated by our study,
family firms belonging to the export industry do appear to realise the benefits of recruiting quality auditors. Our findings may
be of interest for policymakers in developing countries as they demonstrate the power of strong stakeholder groups in
ensuring audit quality and eventually protect broader public interest.
This study has a number of potential limitations. First, despite using a reasonably large sample, the explanatory power of
the models used in our study has remained low, suggesting that this study could suffer from an omitted variables problem.
Future research may be able to incorporate more variables (such as, expertise of audit committees) and use different proxies
for audit quality (for example, local audit office) that would explain variations of audit fees in the context of developing
countries. Also, cross-country studies may be useful to distinguish country effects from family firm effects. Secondly, although
B4A firms are relatively large compared to local audit firms, in the global context, they are still considered small audit firms.
Therefore, further research can look into the quality differentiation between B4A and local audit firms by using other variables

11
92 percent of the family firms in our sample employ CEOs from the same family.
 A. Khan et al. / The British Accounting Review 47 (2015) 304e320 319

such as number of partners, quality assurance mechanisms etc. It is also possible that the results of this study may have been
affected by inaccurate definition and measurement of variables. The use of publicly available information only allows us to be
able to use proxies that can be generated from the annual reports of the firms. As such, some proxies, such as family CEOs etc.
might be difficult to capture. Future research in this area may use interview techniques to identify the informal relationship
between family members and the auditors that might have an impact on auditor choice and audit fees.

Acknowledgements

The authors would like to thank the two anonymous reviewers for their constructive comments. The kind assistance and
helpful guidance received from Professor Nathan Lael Joseph, Joint-editor of The British Accounting Review is gratefully
acknowledged and highly appreciated.

Appendix 1

Variable definition

Variable name Definitions

AF Audit fees are measured as the fees paid by a firm to the auditor for audit services
AUDUT FEE Logged audit fees are measured as the natural log of fees paid by a firm to the auditor for audit services
AQ Auditor quality is measured as a categorical variable and coded “3” if the auditor is a Big 4 auditor, “2” if the auditor is a Big 4
representative in Bangladesh and “1” if the auditor is a local auditor.
INDSPEC Industry specialist auditor is measured as a dummy variable equal to 1 if the auditor is an industry specialist or otherwise 0. An
auditor has been considered as an industry specialist auditor if that auditor is ranked among the top two auditors in a particular
industry in a given year based on the amount of audited total assets in Bangladesh.
FAMILYCON Family firm is measured as a dummy variable equal to 1 if a firm is owned and controlled by family or otherwise 0.
FAMILYOWN Family ownership is measured as the percentage of shares owned by the family members of a firm.
FAMCEO Family CEO is measured as a dummy variable equals 1 if the CEO of a firm is a family member or otherwise 0.
INSIDEOWN Board ownership is measured by the percentage of shares held by non-family inside owners.
INSTOWN Institutional ownership is measured by the percentage of shares held by institutional investors.
GOVOWN Government ownership is measured by the percentage of shares held by the government.
AUDCOMPLX Audit complexity is measured by taking the sum of accounts receivables and inventories scaled by total assets.
LEV Leverage is measured by the ratio of total debt to total assets.
SIZE Firm size is measured by the natural log of total assets.
BOARDIND Board independence is measured by the proportion of independent directors in the board.
PROF Profitability is measured by the ratio of net income and total assets.
END Financial year end is measured as a dummy variable equal to 1 if the year end of the firm is 30 June or otherwise 0.
SUBD Subsidiary is measured as a dummy variable equal to 1 if a firm has above median subsidiaries in a particular year or otherwise 0.
MNC International link is measured as a dummy variable equal to 1 if the firm is a multinational firm or otherwise 0.
NAFD Non-audit fees is measured as a dummy variable equal to 1 if the firm purchases non-audit services from incumbent auditor o
otherwise 0.
EXPORT Export sector is measured as a dummy variable equal to 1 if the firm belongs to the garments/textile sector or otherwise 0.
INDUM Dummy variables for different industrial sectors

References

Ahmed, K., & Goyal, M. K. (2005). A comparative study of pricing of audit services in emerging economies. International Journal of Auditing, 9(2), 103e116.
Ahmed, K., & Nicholls, D. (1994). The impact of non-financial company characteristics on mandatory disclosure compliance in development countries: the
case of Bangladesh. International Journal of Accounting, 29(1), 62e77.
Anderson, R. C., & Reeb, D. M. (2003). Founding-family ownership and firm performance: evidence from the S&P 500. Journal of Finance, 58(3), 1301e1328.
Ashbaugh, H., & Warfield, D. T. (2003). Audits as a corporate governance mechanism: evidence from the German market. Journal of International Accounting
Research, 2(1), 1e21.
Ashraf, J., & Ghani, W. I. (2005). Accounting development in Pakistan. The International Journal of Accounting, 40, 175e201.
Bartholomeusz, S., & Tanewski, G. A. (2006). The relationship between family firms and corporate governance. Journal of Small Business Management, 44(2),
245e267.
Beattie, V. A., Goodacre, A., Pratt, K. C., & Stevenson, J. E. (2001). The determinants of audit fees e evidence from the voluntary sector. Accounting and
Business Research, 31(4), 243e274.
Burkart, M., Panunzi, F., & Shleifer, A. (2003). Family firms. Journal of Finance, 58(2), 2167e2202.
Carey, P., Simnett, R., & Tanewski, G. (2000). Voluntary demand for internal and external auditing by family businesses. Auditing: A Journal of Practice and
Theory, 19(s-1), 37e51.
Cascino, S., Pugliese, A., Mussolino, D., & Sansone, C. (2010). The influence of family ownership on the quality of accounting information. Family Business
Review, 23(3), 246e265.
Chan, P., Ezzamel, M., & Gwilliam, D. (1993). Determinants of audit fees for quoted UK companies. Journal of Business Finance and Accounting, 20(6),
765e786.
Chaney, P. K., Jeter, D. C., & Shivakumar, L. (2004). Self-selection of auditors and audit pricing in private firms. The Accounting Review, 79(1), 51e72.
Credit Suisse. (2011). Asian family business report: Key trends, economic contribution and performance. Switzerland: Emerging Market Research Institute.
Available at http://www.scribd.com/doc/71428527/CS-Asian-Family-Report-2011 Accessed on 16.06.14.
320 A. Khan et al. / The British Accounting Review 47 (2015) 304e320

DeAngelo, L. E. (1981). Auditor size and audit quality. Journal of Accounting and Economics, 3(3), 183e199.
Demsetz, H., & Lehn, K. (1985). The structure of corporate ownership: causes and consequences. Journal of Political Economy, 93, 1155e1177.
Faccio, M., & Lang, L. H. P. (2002). The ultimate ownership of Western European corporations. Journal of Financial Economics, 65(3), 365e395.
Fan, J. P. H., & Wong, T. J. (2005). Do external auditors perform a corporate governance role in emerging economies? Evidence from East Asia. Journal of
Accounting Research, 43(1), 35e72.
Farooque, O. A., Zijl, T. V., Dunstan, K., & Karim, A. K. M. W. (2007). Corporate governance in Bangladesh: link between ownership concentration and
financial performance. Corporate Governance: An International Review, 15(6), 1453e1468.
Firth, M. (1997). The provision of non-audit services and the pricing of audit fees. Journal of Business Finance and Accounting, 24(3), 511e525.
Fortin, S., & Pittman, J. A. (2007). The role of auditor choice in debt pricing in private firms. Contemporary Accounting Research, 24(3), 859e896.
Francis, J. (2004). What do we know about audit quality? British Accounting Review, 36(4), 345e368.
Francis, J., Khurana, I. K., & Pereira, R. (2003). The role of accounting and auditing in corporate governance and development in financial markets around the
world. Asia-Pacific Journal of Accounting and Economics, 10(1), 1e30.
Goldman Sachs. (2011). It is time to re-define emerging markets. In Goldman Sachs strategy series. 31 January 2011. Available at www.ivci.com.tr/Uploads/
GoldmanSachsTurkeyBRIC.pdf Accessed on 30.05.14.
Grossman, S. J., & Hart, O. D. (1982). Corporate financial structure and managerial incentives. In Anonymous (Ed.), The economics of information and un-
certainty (pp. 107e140). Cambridge, MA: National Bureau of Economic Research.
Hay, D. (2013). Further evidence from meta-analysis of audit fee research. International Journal of Auditing, 17(2), 162e176.
Hay, D., Knechel, W. R., & Ling, H. (2008). Evidence on the impact of internal control and corporate governance on audit fees. International Journal of
Auditing, 12(1), 9e24.
Hay, D., Knechel, W. R., & Wong, N. (2006). Audit fees: a meta-analysis of the effect of demand and supply attributes. Contemporary Accounting Research,
23(1), 141e191.
Ho, J. L., & Kang, F. (2013). Auditor choice and audit fees in family firms: evidence from the S&P 1500. Auditing: A Journal of Practice & Theory, 32(4), 71e93.
Imam, M. O., & Malik, M. (2007). Firm performance and corporate governance through ownership structure: evidence from Bangladesh stock market.
International Review of Business Research Papers, 3(4), 88e110.
International Accounting Bulletin. (2011). Special report: Market concentration. Available at www.InternationalAccountingBulletin.com, 06.06.14.
Islam, M. A., & Deegan, C. (2008). Motivations for an organisation within a developing country to report social responsibility information: evidence from
Bangladesh. Accounting, Auditing & Accountability Journal, 21(6), 850e874.
Jaffer, M. J., & Sohail, S. B. (2007). Corporate governance issues in family-owned enterprises. Center for International Private Enterprise (CIPE), Economic
Reform Feature Service. Available at http://www.cipe.org/publications/fs/pdf/042007.pdf Accessed on 12.06.14.
Johnson, W. B., & Lys, T. (1990). The market for audit services: evidence from voluntary auditor changes. Journal of Accounting and Economics, 12(1e3),
281e308.
Karim, A. K. M. W. (2010). Audit pricing, audit concentration and big 4 premium in Bangladesh. Available at http://papers.ssrn.com/sol3/papers.cfm?
abstract_id¼1613454 Accessed on 26.05.14.
Karim, A. K. M. W., & Moizer, P. (1996). Determinants of audit fees in Bangladesh. The International Journal of Accounting, 31(4), 497e509.
Khan, A., Hossain, D., & Siddiqui, J. (2011). Corporate ownership concentration and audit fees: the case of an emerging economy. Advances in International
Accounting, 27(1), 125e131.
Khan, A., Muttakin, M. B., & Siddiqui, J. (2013). Corporate governance and corporate social responsibility disclosures: evidence from an emerging economy.
Journal of Business Ethics, 114(2), 207e223.
La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (1999). Corporate ownership around the world. Journal of Finance, 54(2), 471e517.
McMeeking, K. P., Pope, P. F., & Peasnell, K. V. (2006). The determinants of the UK big six premium. Accounting and Business Research, 36(3), 207e231.
Mitra, S., Hossain, M., & Deis, D. R. (2007). The empirical relationship between ownership characteristics and audit fees. Review of Quantitative Finance and
Accounting, 28(3), 257e285.
Morck, R., Shleifer, A., & Vishny, R. (1988). Management ownership and market valuation: an empirical analysis. Journal of Financial Economics, 20, 293e315.
Niemi, L. (2005). Audit effort and fees under concentrated client ownership: evidence from four international audit firms. The International Journal of
Accounting, 40(4), 303e323.
Niskanen, M., Karjalainen, J., & Niskanen, J. (2010). The role of auditing in small, private family firms: is it about quality and credibility? Family Business
Review, 23(3), 230e245.
Peng, M. W. (2006). Global strategy. Cincinnati, OH: Thomson South-Western.
Piot, C. (2005). Auditor reputation and model of governance: a comparison of France, Germany and Canada. International Journal of Auditing, 9(1), 21e44.
Setia-Atmaja, L., Haman, J., & Tanewski, G. (2011). The role of board independence in mitigating agency problem II in Australian family firms. British Ac-
counting Review, 43(3), 230e246.
Setia-Atmaja, L., Tanewski, G. A., & Skully, M. (2009). The role of dividends, debt and board structure in the governance of family controlled firms. Journal of
Business Finance and Accounting, 36(7e8), 863e898.
Shleifer, A., & Vishny, R. (1997). A survey of corporate governance. Journal of Finance, 52, 737e783.
Siddiqui, J. (2010). Development of corporate governance regulations: the case of an emerging economy. Journal of Business Ethics, 91(2), 253e274.
Siddiqui, J., Zaman, M., & Khan, A. (2013). Do big-four affiliates earn audit fee premiums in emerging markets? Advances in Accounting, 29(2), 332e342.
Simon, D., & Taylor, M. H. (1997). The market for audit services in Pakistan. Advances in International Accounting, 10(1), 87e101.
Simunic, D. A. (1980). The pricing of audit services: theory and evidence. Journal of Accounting Research, 18(1), 161e190.
Simunic, D. A. (1984). Auditing, consulting and auditor independence. Journal of Accounting Research, 22(2), 679e702.
Simunic, D., & Stein, M. T. (1987). Product differentiation in auditing: Auditor choice in the market for unseasoned new issues. In Research monograph no. 13.
Vancouver, British Columbia: The Canadian Certified General Accountants' Research Foundation.
Sobhan, F., & Werner, W. (2003). Diagnostic study of existing corporate governance scenario in Bangladesh. In F. Sobhan, & W. Werner (Eds.), A comparative
analysis of corporate governance in South Asia. Bangladesh: Bangladesh Enterprise Institute.
Villalonga, B., & Amit, R. (2006). How does family ownership, control and management affect firm value? Journal of Financial Economics, 80(2), 385e417.
Wang, D. (2006). Founding family ownership and earning quality. Journal of Accounting Research, 44(3), 619e656.
World Bank. (2002). Bangladesh: Financial accountability for good governance. Washington, D.C.: The World Bank.
World Bank. (2012). Bangladesh: Corporate governance country assessment. Washington, D.C.: The World Bank.
Young, M. N., Peng, M. W., Ahlstrom, D., Bruton, G. D., & Jiang, Y. (2008). Corporate governance in emerging economies: a review of principaleprincipal
perspective. Journal of Management Studies, 45(1), 196e219.