Mano

The British Accounting Review xxx (2015) 1e17
Contents lists available at ScienceDirect
The British Accounting Review

journal homepage: www.elsevier.com/locate/bar
Audit fees, auditor choice and stakeholder inuence:

Evidence from a family-rm dominated economy
Arifur Khan a, Mohammad Badrul Muttakin a, Javed Siddiqui b, *
a
b
Department of Accounting, Faculty of Business and Law, Deakin University, Australia

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:
Received 14 January 2013
Received in revised form 10 March 2015
Accepted 11 March 2015
Available online xxx
Despite the dominance of family-owned publicly listed companies in developing economies, prior research has paid relatively little attention to this area and the socio-economic
context of these countries has been mostly ignored. This study contributes to the accounting literature by providing empirical evidence of the effects of family control and
ownership on audit pricing and auditor choice in a developing economy context. Using
1058 rm-year observations of publicly listed companies in Bangladesh, where family
rms are the most dominant form of public companies, we nd that in comparison with
non-family rms, our sample family rms pay signicantly lower audit fees and choose
lower quality auditors. However, for export-oriented industries, family rms seem to pay
signicantly higher audit fees and recruit better quality auditors compared to non-family
rms. Collectively, our ndings have important implications for audit markets in emerging
economies in which the sustainability of family rms is crucial for overall economic
development.
2015 Elsevier Ltd. All rights reserved.
Keywords:
Family control
Family ownership
Audit fees
Auditor choice
Emerging economy
1. Introduction
We explore audit fees and auditor choice in a family-rm 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 rms 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 conict of interest between owners and managers (referred to as the type I agency problem) is less of an
issue for family rms, 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 rms presents two possible scenarios. Due to lower type I agency problems, it is expected that family rms
* Corresponding author. Tel.: 44 161 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.
Please cite this article in press as: Khan, A., et al., Audit fees, auditor choice and stakeholder inuence: Evidence from a familyrm dominated economy, The British Accounting Review (2015), http://dx.doi.org/10.1016/j.bar.2015.03.002
A. Khan et al. / The British Accounting Review xxx (2015) 1e17
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 rms 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
conicting arguments makes auditor choice and audit fees in family rms an interesting research area. In an important review
paper summarising the large body of audit fee research, Hay, Knechel, and Wong (2006) identied 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 signicant 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 rms in a developing economy, using the case of Bangladesh. Bangladesh has made signicant 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 ve 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 rms, making this the dominant form
of listed companies in Bangladesh.2
The audit market in Bangladesh is characterised by poor demand for audited nancial statements (Ahmed & Goyal, 2005)
and poor perceptions regarding audit quality (Sobhan & Werner, 2003). At present, three categories of audit rms operate in
the Bangladeshi audit market: the Big 4 rm e KPMG, Big 4 afliate (B4A) rms, and local audit rms. It seems unusual that
KPMG is the only Big 4 rm with a direct presence in Bangladesh. The rest of the Big 4 rms operate through their associates
or cooperating rms; they do not have to adhere to Big 4 quality control standards. Siddiqui, Zaman, and Khan (2013) report
that unlike Big 4 rms, B4A rms do not earn a signicant audit fee premium in Bangladesh. This indicates that the quality of
B4A rms is not perceived to be the same as a Big 4 rm. 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 rms
operating in the garments sector act as supply chains to renowned international brands. Recent research (see Islam & Deegan,
2008; Khan, Muttakin, & Siddiqui, 2013) has identied these foreign stakeholders to be important pressure groups asking for
greater nancial 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 nancial statements in Bangladesh, and
inuence the choice of audit rms. Ashbaugh and Wareld (2003) document the inuence 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 rm
may stem from the audit rm's reputation regarding the quality of audit services, which mitigate some concerns regarding
the quality of nancial 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 rms more effectively. According to this approach, founding families are more likely to forgo short-term benets
from managing earnings because of the incentives to pass on their business to future generations and to protect the family's
reputation. Hence, family rms are motivated to report earnings of higher quality than non-family rms. 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 rms 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-benecial activities. In contrast, the
entrenchment effect of family ownership may motivate users of nancial statements and external shareholders to demand
high-quality earnings from family rms 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 rms (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 rms (www.dsebd.org).
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 rms 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 rms. 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 rms may forgo short term benets 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 rms in export
industries may have sufcient 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 rm-year observations of public limited companies listed on the DSE from 2005 to 2013, we
investigate auditor choice and audit fees in family rms. Our results indicate that family-owned listed public limited rms in
Bangladesh pay signicantly lower audit fees compared to non-family rms and are less interested in recruiting higher quality
auditors. However, we nd that family rms operating in export industries select better quality auditors and pay higher audit
fees, suggesting that managers in export-oriented family rms 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 nancial reporting quality. Our
ndings also reveal that family rms 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 efcacy of western-styled corporate
governance mechanisms in the context of developing economies. Our ndings are robust to a number of statistical tests.
Our study is one of the rst to investigate audit fees and auditor choice in family rms in the context of developing
economies. In contrast with previous studies on family rms (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 rms 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 rms and discusses the characteristics of the audit market in Bangladesh. Based on a discussion regarding the institutional context and the role of audit rms, 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 ndings 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 metaanalysis to test the combined effect of the most commonly used variables in audit fees research. On the basis of their observations about anomalies, inconsistencies, and gaps in the previous literature, Hay et al. (2006) suggest that audit research
would benet 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) reported 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 full the minimum statutory requirements. 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 nds a
signicantly positive relationship between diffused institutional stock ownership (i.e., having less than 5 percent individual
shareholding) and audit fees, and a signicantly 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 rm size to be an important
measure of audit quality, as larger rms would have a greater reputation to lose. Francis (2004), however, notes that such
arguments do not necessarily establish that Big 4 audit rms are superior, as evidenced by high prole audit failures.
Nevertheless, the paper suggests that Big 4 rms 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 conrmatory
evidence that, on average, audit rms charging higher audit fees (i.e., Big 4 audit rms) actually provide higher quality audits,
measured in terms of observable outcomes such as the auditors' propensity to issue modied audit opinions, and the quality
of audited nancial statements.
A small number of studies have explored the choice of auditors in family rms. Using a sample of small private rms in
Finland, Niskanen et al. (2010) nd 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 rms. Ho and Kang (2013) investigate auditor choice and audit fees in family rms using data from S&P
1500 companies in the USA. They nd that family rms 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 environment differs signicantly from developing countries. However, Jaffer and Sohail (2007) report that corporate governance
structures are almost non-existent in the majority of family-owned rms in India. Consequently, only 15 percent of familyowned businesses in India and Pakistan survive into the third generation. Frequent family disputes lead to the family rm's
inability to implement a succession plan, which, in the absence of proper corporate governance mechanisms, poses a signicant threat to a family rm's survival.
We also investigate the inuence 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 rms. However, Ashbaugh and Wareld (2003) document 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 rm'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 nancial 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 rm (Siddiqui et al., 2013). Siddiqui (2010) reports that despite having a very large population, the number of qualied
auditors in Bangladesh is signicantly 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) nd that approximately 78 percent of CEOs are shareholders
of Bangladeshi rms, either as founder shareholders or as descendants of founding families. The study also nds that the ve
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 rmly based on family ownership, and in most rms, controlling families
dominate the boards, lling 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 nancial statements, had diminished the status of annual general meetings to that of a mere ritual.
Karim (2010) reports that although B4A rms are considered large rms in the context of Bangladesh, in the global context
these are not very large rms, with the number of partners ranging from 3 (for the afliated rm of Ernst and Young) to 7 (for
the afliated rm 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 rms 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 rms 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 rms in the audit market in Bangladesh.
Agency theory suggests that family rms 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 signicantly lower than its
neighbouring India and Pakistan.
gain is restricted by their long-term investment horizon, concern for their reputation and higher interest in the rm. 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 rms and
can expropriate minority shareholders (type II agency problem). Wang (2006) argues that family rms usually have weak
corporate governance as a result of ineffective monitoring by the board. This is because inuential 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 rms, audit rm size
differentiation (Big 4, B4A and local rms), and the presence of important stakeholder groups in export industries. The
interaction between these factors may signicantly affect audit fees and auditor choice in family rms.
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
rms. However, unlike many other developed countries, where previous studies have been conducted, the domination of
family rms in the Bangladeshi capital market is overwhelming. Therefore, compared to developed countries, where family
rms occupy a comparatively less important position in the capital market, rms 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 rms. Therefore, the proposed
research hypothesis is:
H1: There is no signicant difference in audit fees between family rms and non-family rms.
An interesting feature of the Bangladeshi audit market is the presence of Big 4, B4A and local audit rms. Siddiqui et al.
(2013) report that although the only Big 4 rm operating in Bangladesh (KPMG) earns an audit fee premium, the audit fees
charged by B4A rms are not signicantly higher than local rms. Their ndings suggest that the market does not perceive
B4A rms to be of similar quality to KPMG. However, the B4A rms are internationally linked and are relatively large in the
context of Bangladesh. Therefore, in terms of audit quality, B4A rms are most likely in between Big 4 and local audit rms. In
the context of Bangladesh, it would be interesting to explore whether family rms hire higher quality auditors. Thus the
following hypothesis is proposed:
H2: Demand for audit quality (choice of Big 4/B4A/local/specialist audit rms) is not signicantly higher in family rms than
non-family rms.
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 rms 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 nancial reporting quality. It
might be reasonable to assume that these important stakeholders would also be concerned about the quality of audited
nancial statements in Bangladesh, and inuence the choice of audit rms (scenario 2, Fig. 1). The alignment effect would
imply that in order to maintain long term and sustainable relationship with the foreign buyers, family rms would forgo the
short term benets of manipulating earnings. However, prior research documents that corporate governance mechanisms
have a largely symbolic role in family dominated rms 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 rms in the export industries may attempt to hire better
quality auditors and pay higher audit fees. However, the recent high-prole 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 inuence audit fees and auditor choice, we develop the following two hypotheses:
91 percent of our sample companies from the RMG sector are family-owned.
Fig. 1. Institutional context and hypothesis development.
H3A: There is no signicant difference in audit fees between family rms in the RMG sector and non-family rms.
H3B: Demand for audit quality (choice of Big 4/B4A/local/specialist audit rms) is not signicantly higher in family rms in the
RMG sector than non-family rms.
3. Methodology
3.1. Data and sample
The sample selection procedure is reported in Table 1. The sample consists of all 155 non-nancial 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
Table 1
Sample description.
Panel A
Year
Family
Non-family
Total
2005
2006
2007
2008
2009
2010
2011
2012
2013
84
88
84
88
76
70
67
69
42
45
46
45
48
45
46
45
42
28
129
134
129
136
121
116
112
111
70
1058
Panel B
Non-family (rmyear)
Total (rmyear)
Percent of family rms (rmyear)
Sector
Family (rmyear
Cement
Ceramics
Engineering
Food
IT
Jute
Paper & printing
Miscellaneous
Pharmaceuticals
Service & real
estate
Tannery
Textile
Total
26
14
110
90
16
16
18
37
110
18
29
10
67
73
31
9
0
51
47
28
55
24
177
163
47
25
18
88
157
46
47.3
58.3
62.1
55.2
34.0
64.0
100.0
42.0
70.1
39.1
12
194
661
32
20
397
44
214
1058
27.3
90.7
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 rm-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 nal sample of 1058 rm-years.5
Table 1 reports the sample composition by year, rm ownership and industry sector. Panel A lists the sample rms by
ownership type (family vs. nonfamily) and by year. The sample rms vary by year: 129 rms (84 family rms) in 2005 to 70
(42 family rms) in 2013. From Panel B of Table 1, it is observed that family rms represent 63 percent of the total sample. This
indicates that family rms are the dominant form of company listed on the DSE. In our sample, the paper and printing industry appears to have the highest percentage of family rms (100%), followed by the RMG industry (91%).
3.2. Measuring family ownership and control
In this study, we adopt multiple criteria to identify family rms. This is mainly because the prior literature provides no
commonly accepted measure or criterion for identifying a family rm. La Porta et al. (1999) argue that the 20 percent cut-off
point is usually enough to have effective control of a rm 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 rened denition of family rms that does not
uniquely rely on ownership concentration as major determining criterion.6 So we identify family rms as being (i) rms in
which 20 percent of a rm'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 rm is considered a family rm 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) notication, 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 Ofcer, Company Secretary, Chief Financial Ofcer, Head of Internal Audit and their spouse and minor children.
Thus in the present study, if a rm has at least one such shareholder or family member controlling 20 percent or more
shareholdings, it is considered a family rm.
Initial sample 155 rms times 9 years 1395 rm-years; Final sample (1395 337) 1058.
Cascino et al. (2010) argued that choosing a certain percentage threshold would not allow any difference in terms of family and non-family rms to be
captured because a certain percentage threshold only represents high ownership concentration rather than family ownership and management.
6
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 rm and audit fees.
AUDIT FEE b0 b1 FAMILYCON b2 FIRM SPECIFIC CONTROLS b3 INDUSTRY CONTROLS b4 YEAR CONTROLS
3
(1)
The dependent variable audit fees (AUDIT FEE) is measured by taking the natural log of audit fees. The key variable family
rm (FAMILYCON) is an indicator variable that equals 1 if the rm is a family rm 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); rm size (SIZE); leverage (LEV), auditor type
(AQ and INDSPEC); protability (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 denitions of the
variables have been provided in Appendix 1. We use two proxies to measure auditor type variable. The rst 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 auditors 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 rms 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 rms
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 rms will charge a premium for their
quality (Beattie, Goodacre, Pratt, & Stevenson, 2001; Firth, 1997). Low protability could be associated with nancial pressure
which would require more audit efforts to conrm 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 scal 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 nonaudit 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 rm and the choice of auditor.
AUDITOR CHOICE a0 a1 FAMILYCON a2 FIRM SPECIFIC CONTROLS a3 INDUSTRY CONTROLS

a4 YEAR CONTROLS 3
(2)
We use two alternative proxies for auditor choice model. The rst proxy captures the impact of family rms 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
We thank an anonymous reviewer for providing us helpful suggestions to develop the categorical variable and use ordered probit model.
response to Hay's (2013) call for using alternative proxies for audit rm 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
denition of family rm (FAMILYCON) variable is similar to the denition we used in our audit fees model. The control
variables used in equation (2) are non-family insider ownership (INSIDEOWN), institutional ownership (INSTOWN), government ownership (GOVOWN), audit complexity (AUDCOMPLX), rm size (SIZE), and leverage (LEV), protability (PROF)
and international link dummy (MNC) variables. The denitions of the control variables are similar to the denitions 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 inuence 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-rm's operations. Because Since auditor change is related to its protability (see
Johnson & Lys, 1990), we include a protability (PROF) measure, dened as earnings before interest and taxes to total assets to
control for the potential effect of protability.
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 government are 15 percent and 5 percent, respectively. The mean rm 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 rms. Among
family rms 92 percent of the CEOs are family members. This is consistent with Sobhan and Werner (2003) who nd that
family rms in Bangladesh typically appoint family members as CEOs. Board independence is statistically indistinguishable
between family and non-family rms. On average, non-family rms pay higher audit fees than family rms. Institutional
ownership is more prevalent in non-family rms than in family rms although the difference is statistically insignicant.
Average government ownership of non-family rms is signicantly higher than family rms. On average, 18.4 percent of
family rms and 23.2 percent of non-family rms recruit industry specialist auditors.
In Table 3 we present the Spearman correlation matrix for the variables. Family rms have a negative relationship with
audit fees. This variable is also negatively correlated with higher quality auditors. Family rms are also positively correlated
with protability. Consistent with previous studies (see Anderson & Reeb, 2003; Setia-Atmaja et al., 2009), we nd that family
ownership is negatively related to rm 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 rms 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. Specically, 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 nd evidence that family rms pay lower audit fees than non-family rms. Specically, we nd that the coefcient
estimate for family rms (FAMILYCON) is negative and signicant (b1 0.168, p < 0.01). Therefore, H1 is rejected. This
implies that family rms are less likely to demand extensive audit and incur higher audit fees. This is consistent with the
entrenchment argument and suggests that family rms 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 efcient as probit
but does not require normality of parameter distribution (e.g., Carey, Simnett, & Tanewski, 2000).
9
As of July 1, 2014, 1US$ Tk. 78.
10
Table 2
Summary statistics.
Panel A: full sample
AUDCOMPLX
BOARDIND
INDSPEC
INSIDEOWN
FAMILYCON
FAMILYOWN
FAMCEO
INSTOWN
GOVOWN
LEV
SIZE (in million Tk.)
NAFD
PROF
AF (in Tk.)
SUBD
END
MNC
AQ
Mean
Median
Std. dev.
JarqueeBera
0.344
0.097
0.196
0.095
0.629
0.297
0.599
0.151
0.048
0.611
2450.000
0.203
0.066
116,725.900
0.297
0.518
0.056
1.159
0.332
0.111
0.000
0.020
1.000
0.311
1.000
0.130
0.000
0.502
699.000
0.000
0.063
60,000.000
0.000
1.000
0.000
1.000
0.210
0.087
0.397
0.306
0.483
2.696
0.492
0.127
0.173
0.698
6030.000
0.402
0.122
161,233.600
0.457
0.500
0.231
0.456
59.306a
25.975a
419.347a
19,827,389a
189.477a
3,075,100a
189.527a
111.344a
11,037.02a
83,323.38a
32,6676.3a
403.236a
330,203.3a
14,543.13a
219.350a
185.501a
10,397.87a
4562.534a
Panel B: family and non-family rm subsamples
AUDCOMPLX
INDSPEC
BOARDIND
INSIDEOWN
INSNTOWN
GOVOWN
LEV
SIZE (in million Tk.)
NAFD
PROF
AF (in Tk)
END
MNC
FAMCEO
SUBD
AQ
a,b
Family
Non-family
t-Test statistic
0.356
0.182
0.096
0.052
0.150
0.008
0.596
2471.756
0.137
0.069
103,982.257
0.524
0.000
0.915
0.411
1.123
0.324
0.234
0.100
0.168
0.151
0.130
0.636
2400.591
0.318
0.061
138,348.584
0.515
0.153
0.066
0.119
1.229
2.779a
1.998b
0.265
9.327a
0.463
7.412a
3.667a
1.375
7.184a
0.690
3.502a
0.289
9.641a
7.778a
6.239a
4.444a
and c represent signicance test statistic at the 1%, 5% and 10% level, respectively.
shareholders. The statistical signicance of some of the other control variables suggests that audit fees are also inuenced by
other factors. In particular, the positive and signicant coefcient for auditor quality (AQ) suggests an audit fee premium for
higher quality auditors (Beattie et al., 2001; Firth, 1997). A negative and signicant coefcient for insider ownership (INSIDEOWN) is consistent with the entrenchment argument. We also nd positive and signicant 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 signicant coefcient of rm size (SIZE) suggests that
larger rms incur higher audit fees than smaller rms (Simunic, 1980). We also nd a positive and signicant 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
nd a positive and signicant coefcient of MNC dummy variable. Finally, we also nd a positive and signicant coefcient 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 nds that higher quality auditors receive higher audit fees from their clients in developing
countries (Ahmed & Goyal, 2005), we use an interaction variable for family rms (FAMILYCON) and auditor quality (AQ) to test
for this effect. Model 2 in Table 4 shows a positive and signicant coefcient for FAMILYCON*AQ implying that higher quality
auditors charge a premium to family rms. 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 rms. However, those studies typically
concentrate on private family rms whereas the family rms 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.
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
a,b
Variable
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
AUDCOMPLX
BOARDIND
INDSPEC
INSIDEOWN
FAMYLYCON
INSTOWN
GOVOWN
LEV
SIZE
NAFD
PROF
AF
SUBD
END
MNC
AQ
1.000
0.127a
0.043
0.065b
0.079c
0.001
0.031
0.005c
0.055a
0.005
0.125a
0.008b
0.044a
0.082
0.043c
0.083b
1.000
0.069b
0.008
0.041
0.038
0.165
0.221
0.133a
0.122a
0.144a
0.286a
0.027
0.122a
0.230
0.214b
1.000
0.003
0.063b
0.055
0.026
0.072
0.082b
0.137a
0.152a
0.196a
0.130a
0.171
0.117a
0.339a
1.000
0.184a
0.022
0.066b
0.012
0.003
0.026
0.006
0.029
0.122c
0.019
0.009b
0.007
1.000
0.012
0.351a
0.025
0.008
0.222a
0.001
0.120a
0.299a
0.029
0.330a
0.143a
1.000
0.194a
0.042
0.065b
0.004
0.053
0.018
0.037
0.043
0.042
0.003
1.000
0.301a
0.037
0.114b
0.150
0.062b
0.179
0.176a
0.058c
0.019
1.000
0.115
0.135a
0.347b
0.169
0.057c
0.065b
0.085a
0.053
1.000
0.057c
0.089c
0.689c
0.312a
0.239a
0.170a
0.167a
1.000
0.252
0.156a
0.050
0.175a
0.396a
0.374a
1.000
0.267a
0.105a
0.152a
0.287a
0.306a
1.000
0.342a
0.331b
0.466a
0.436a
1.000
0.200c
0.074b
0.150a
1.000
0.258b
0.308c
1.000
0.532b
1.000
and represent signicance of the coefcients at the 1%, 5% and 10% level, respectively.
11
Table 3
Spearman rank correlation matrix.
12
Table 4
Regression results: relationship between family rms and audit fees.
Variable
Intercept
FAMILYCON
AQ
FAMILYCON*AQ
INDSPEC
FAMILYCON*INDSPEC
INSIDEOWN
INSTOWN
GOVOWN
AUDCOMPLX
LEV
SIZE
BOARDIND
PROF
END
SUBD
MNC
NAFD
EXPORT
FAIMILYCON*EXPORT
Industry dummies
Year dummies
Adjusted R-squared
F-statistic
a,b
Model 1 (OLS)
Model 2 (OLS)
Model 3 (OLS)
Model 4 (OLS)
Model 5 (OLS)
AUDIT FEE (H1)
AUDIT FEE (H1)
AUDIT FEE (H3A)
AUDIT FEE (H1)
AUDIT FEE (H1)
AUDIT FEE (H3A)
Coefcient
Prob.
Coefcient
Prob.
Coefcient
Prob.
Coefcient
Prob.
Coefcient
Prob.
Coefcient
Prob.
4.951a
0.168a
0.466a
0.000
0.009
0.000
4.679a
0.508a
0.254b
0.261b
0.000
0.001
0.030
0.045
4.277a
0.035
0.416a
0.000
0.604
0.000
5.241a
0.139b
0.000
0.047
4.049a
0.124b
0.000
0.018
4.419a
0.005b
0.000
0.038
0.258a
0.000
0.113a
0.189b
0.019
0.144
0.266c
0.331a
0.050
0.341a
0.491a
1.456a
0.144a
0.190a
0.725a
0.169a
0.002
0.048
0.754
0.352
0.055
0.000
0.127
0.000
0.000
0.000
0.001
0.000
0.000
0.002
0.246a
0.000
0.609
0.014
0.001
0.325a
0.040
0.288a
0.238b
1.283a
0.053
0.165a
0.406a
0.181a
Included
Included
0.631
48.127a
0.001
0.939
0.997
0.002
0.235
0.000
0.041
0.000
0.289
0.003
0.002
0.004
0.598
0.152
0.014
0.446a
0.016
0.306a
0.887a
1.298a
0.030
0.184a
0.561a
0.182a
Included
Included
0.611
69.802a
0.002
0.381
0.935
0.000
0.638
0.000
0.001
0.000
0.530
0.001
0.001
0.004
0.581
0.040
0.011
0.382a
0.013
0.318a
0.664b
0.991a
0.044
0.103c
0.460a
0.152a
0.121b
0.252a
Included
Included
0.640
60.988a
0.002
0.814
0.948
0.000
0.695
0.000
0.025
0.000
0.349
0.056
0.000
0.011
0.047
0.001
0.628
0.258
0.066
0.371a
0.044
0.296a
0.383
1.366a
0.106b
0.164a
0.846a
0.117c
Included
Included
0.612
41.305a
0.003
0.174
0.724
0.001
0.217
0.000
0.238
0.000
0.043
0.007
0.000
0.078
Included
Included
0.647
52.332a
Model 6 (OLS)
0.647
0.257
0.062
0.408a
0.015
0.328a
0.789b
1.140a
0.103b
0.098c
0.814a
0.118c
0.161b
0.219b
0.001
0.151
0.728
0.000
0.666
0.000
0.010
0.000
0.034
0.085
0.000
0.058
0.011
0.043
Included
Included
0.631
54.625a
and c represent signicance of the coefcients 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 inuence of important
stakeholders on nancial reporting disclosures in Bangladesh. As stated before, Bangladesh is currently the second largest
exporter of textile/garment products in the world, and many family rms 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 nancial statements and this may inuence the choice of audit rms. To test this hypothesis, we include
an interaction variable based on family rms (FAMILYCON) and RMG industry variable (EXPORT). Models 3 and 6 in Table 4
show positive and signicant coefcients for FAMILYCON*EXPORT variable. Thus, H3A is rejected. Our result implies that
family rms in the export-oriented RMG sector pay signicantly higher audit fees, perhaps to allay concerns of their international 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 signicant impact of industry specialist auditor on audit fees. We also document that these audit rms obtain
fee premiums from family rms. Our overall ndings suggest the inuence of industry specialist auditors in emerging
markets.
4.3. Relationship between family rms 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 rst proxy of auditor choice, i.e., auditor quality (AQ). In
Model 1 we examine the impact of family rms on recruiting higher quality auditors. We nd that family rms are less likely
to hire higher quality auditors than non-family rms. Specically, we nd that the coefcient estimate of family rms
(FAMILYCON) is negative and signicant (a1 0.520, p < 0.01). Therefore, H2 is rejected. Consistent with the entrenchment
argument this nding also implies that family rms 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 signicant coefcients of the other
variables suggest that the choice of a rm's higher quality auditor is also inuenced by other factors. In particular, insider
ownership (INSIDEOWN) is negative and signicant for H2. So rms with a higher percentage of insider ownership are less
likely to have higher quality auditors. Our results also show positive and signicant relationships for rm size (SIZE) and
10
We also use a three way interaction between family rms (FAMILYCON), auditor quality (AQ) and RMG industry variable. We discuss the result as a part
of additional analysis.
13
Table 5
Regression results: relationship between family rms and auditor choice.
Variable
AQ (H2)
AQ (H3B)
INDSPEC (H2)
INDSPEC (H3B)
Model 1(ordered probit)
Model 2 (ordered
probit)
Model 3 (logit)
Model 4 (logit)
Coefcient
Prob.
Coefcient
Prob.
Coefcient
Prob.
Coefcient
Prob.
Intercept
FAMILYCON
INSIDEOWN
INSTOWN
GOVOWN
AUDCOMPLX
LEV
SIZE
BOARDIND
PROF
MNC
EXPORT
FAMILYCON*EXPORT
9.797
0.520a
0.021c
1.129
0.734
0.435
0.407a
0.434a
1.061b
4.743a
1.404a
0.000
0.000
0.056
0.228
0.238
0.165
0.000
0.000
0.041
0.000
0.000
8.631
0.436a
0.006
1.170
0.895
0.367
0.412a
0.424a
0.901b
4.817a
1.613a
0.434b
0.857
0.000
0.003
0.957
0.329
0.313
0.256
0.000
0.000
0.030
0.000
0.000
0.070
0.001
4.255
0.354b
0.099
1.277
0.747
0.332
0.102
0.145a
0.325b
3.269a
0.047c
0.000
0.039
0.713
0.164
0.214
0.405
0.592
0.009
0.040
0.001
0.094
3.435
0.643a
0.095
1.129
0.867
0.395
0.117
0.118b
0.604c
3.203a
0.034c
0.761b
0.822a
0.004
0.006
0.726
0.170
0.153
0.325
0.540
0.035
0.059
0.001
0.070
0.045
0.007
Industry dummies
Year dummies
Pseudo R-squared
Included
Included
0.334
a,b
Included
Included
0.343
Included
Included
0.184
Included
Included
0.172
and c represent signicance of the coefcient at the 1%, 5% and 10% level respectively using two-tailed tests.
protability (PROF), implying that the choice of higher quality auditors is signicantly related to size and protability. We also
nd that multinational companies (MNC) are more likely to recruit higher quality auditors.
Because of international stakeholders' pressure family rms could produce higher quality audited nancial statements, we
explore whether family rms 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 signicant
coefcient for FAMILYCON*EXPORT. This indicates that family rms in the export-oriented sector are more likely to recruit
higher quality auditors. Hence, H3B is rejected.
We then explore the impact of family rms 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 rms are generally less likely to recruit industry specialist auditors. However,
when these rms 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) document a positive relationship between measures of corporate governance and audit fees for New Zealand rms. Like many
other developing countries, Bangladesh has also adopted western-styled corporate governance codes, requiring the separation 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, signicantly reducing the effectiveness of an important
corporate governance scheme such as CEO-chair separation. 59.9 percent of our sample family rms 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 signicant coefcient for
each of the FAMILYOWN and FAMILYCEO variables. A negative and signicant coefcient of FAMILYOWN variable suggests
that as the family ownership increases families tend to incur lower audit fees. A negative and signicant coefcient of
FAMILYCEO variable implies that family rms incur lower audit fees when a family member serves as a CEO. This is consistent
with the notion that unskilled and inefcient 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 rm 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
signicant coefcient for FAMILYOWN. This implies that rms with a larger percentage of family ownership are less likely to
14
Table 6
Regression results: relationship between family ownership and family CEO and audit fees and auditor choice.
Variable
Intercept
FAMILYOWN
FAMCEO
AQ
INDSPEC
INSIDEOWN
INSTOWN
GOVOWN
AUDCOMPLX
LEV
SIZE
BOARDIND
PROF
END
SUBD
MNC
NAFD
Industry dummies
Year dummies
Adjusted R-squared
F-statistic
Pseudo R-squared
a,b
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)
Coefcient
Prob.
Coefcient
Prob.
Coefcient
Prob.
Coefcient
Prob.
4.927a
0.436b
0.293a
0.476a
0.000
0.010
0.000
0.000
5.219a
0.480a
0.287a
0.000
0.008
0.000
9.919a
0.013b
0.427b
0.000
0.032
0.011
4.063a
0.065b
0.448b
0.001
0.016
0.022
0.407b
0.064
0.115
0.291a
0.028
0.288a
0.210
1.108a
0.061
0.139b
0.476a
0.188a
0.025
0.715
0.494
0.005
0.405
0.000
0.495
0.000
0.214
0.013
0.000
0.003
0.259a
0.435b
0.312
0.179
0.339a
0.033
0.297a
0.365
1.199a
0.114b
0.138b
0.919a
0.123c
0.000
0.028
0.194
0.330
0.003
0.352
0.000
0.258
0.000
0.028
0.021
0.000
0.062
0.033
1.044
0.835
0.414
0.429
0.438a
1.222b
4.821a
0.769
0.239
0.427
0.190
0.179
0.000
0.088
0.000
0.104
1.283
0.857
0.460
0.100
0.138a
0.328b
3.105a
0.706
0.162
0.147
0.259
0.596
0.012
0.039
0.002
1.474a
0.000
0.032
0.196
Included
Included
0.636
47.246a
Included
Included
0.618
40.773a
Included
Included
Included
Included
0.327
0.132
and c represent signicance of the coefcients 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 ndings in
regards to family rms. We also document a negative signicant coefcient 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 rms with a greater percentage of ownership by family members and family CEOs are less likely to recruit industry specialist auditors. Overall, our results suggest that the efcacy of a western-styled corporate governance mechanism
such as the separation of CEO and chair is signicantly compromised due to the tendency of selecting the chair and CEO from
the same family in Bangladesh.
4.4.2. Family rms, auditor quality and garment/textile industry variables
In our audit fees model we introduce a three-way interaction between family rms, auditor quality and RMG industry
variables (FAMILYCON*AQ*EXPORT). Our objective is to explore the impact of family rms 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 signicant coefcient of the interaction variable indicating that family rms 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 rms in RMG sector with industry specialist
auditor on audit fees. Once again, we nd a positive and signicant coefcient for the three way interaction variable, which
implies that family rms 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 denition of family rms. In
particular, we dene a rm as family controlled when an individual, or group of family members, hold at least 50 percent of a
rm's share (voting rights). We use a dummy variable to identify the rms and are set equal to 1 if the rm is considered to be
family rm and 0 otherwise. When we use this alternative ownership threshold the number of family rms 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 rms as opposed to
non-family rms. The results for family-subsamples are generally consistent with our results for the full sample. This is
possibly due to the dominance of family rms in the Bangladeshi corporate sector. Third, motivated by the ndings 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 specication of family ownership variable. However, we fail to document any
15
signicant non-linear impacts of family ownership variable on our key variables of interest. Finally, we repeat all the analyses
using a xed effect model. We do not nd any qualitative difference to our original results.
5. Conclusions
The paper investigates audit fees and auditor choice in family rms in Bangladesh. Unlike many developed countries,
family rms are the dominant form of listed public companies in Bangladesh, and the majority of leading companies listed in
the DSE are actually family rms. We nd that compared to non-family rms, family rms in Bangladesh tend to pay lower
audit fees and are less likely to hire top quality audit rms. However, family rms in export oriented industries are inclined to
pay higher audit fees and recruit better quality auditors. Our results also reveal that rms with CEOs11 from the same family
tend to pay signicantly less audit fees compared to rms with hired CEOs. In addition, we identify a number of important
determinants of audit pricing and auditor choice and provide rst-hand evidence of the joint determination of audit and nonaudit service fees in an emerging economy context.
Our study is among the rst to investigated audit fees and auditor choice in family rms in the context of developing
economies. Previously, Ho and Kang (2013) reported that family rms in the USA had a tendency to recruit lower quality
auditors and incur lower audit fees. This suggests that higher audit efforts in family rms 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 rms, whether
family or non-family listed, are operating in a more heavily regulated environment in developed countries. Audit rms 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 rms 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 rms is likely to be
driven by the expropriation effect rather than the alignment effect, as demonstrated in earlier studies conducted in developed
markets. Our ndings 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 rms 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 efcacy of
important western styled corporate governance mechanisms such as CEO-chair separation can be signicantly compromised
in Bangladesh.
Our ndings 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 rms, especially in emerging economies in Asia. For some of
the emerging economies in Asia, family rms 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 rms is crucial for economic
growth in these regions. However, due to weaknesses in corporate governance and internal control mechanisms, the majority
of these family rms tend to perish early. As our ndings indicate, the majority of the family rms are reluctant to recruit
quality audit rms and pay higher audit fees, hence increasing the risk of poor nancial management. For long-term survival
of these family rms, 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 rms belonging to the export industry do appear to realise the benets of recruiting quality auditors. Our ndings 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 ofce) 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 rm effects. Secondly, although
B4A rms are relatively large compared to local audit rms, in the global context, they are still considered small audit rms.
Therefore, further research can look into the quality differentiation between B4A and local audit rms by using other variables
11
92 percent of the family rms in our sample employ CEOs from the same family.
16
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 denition 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 rms. As such, some proxies, such as family CEOs etc.
might be difcult 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 denition
Variable name
Denitions
AF
AUDUT FEE
AQ
Audit fees are measured as the fees paid by a rm to the auditor for audit services
Logged audit fees are measured as the natural log of fees paid by a rm to the auditor for audit services
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.
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.
Family rm is measured as a dummy variable equal to 1 if a rm is owned and controlled by family or otherwise 0.
Family ownership is measured as the percentage of shares owned by the family members of a rm.
Family CEO is measured as a dummy variable equals 1 if the CEO of a rm is a family member or otherwise 0.
Board ownership is measured by the percentage of shares held by non-family inside owners.
Institutional ownership is measured by the percentage of shares held by institutional investors.
Government ownership is measured by the percentage of shares held by the government.
Audit complexity is measured by taking the sum of accounts receivables and inventories scaled by total assets.
Leverage is measured by the ratio of total debt to total assets.
Firm size is measured by the natural log of total assets.
Board independence is measured by the proportion of independent directors in the board.
Protability is measured by the ratio of net income and total assets.
Financial year end is measured as a dummy variable equal to 1 if the year end of the rm is 30 June or otherwise 0.
Subsidiary is measured as a dummy variable equal to 1 if a rm has above median subsidiaries in a particular year or otherwise 0.
International link is measured as a dummy variable equal to 1 if the rm is a multinational rm or otherwise 0.
Non-audit fees is measured as a dummy variable equal to 1 if the rm purchases non-audit services from incumbent auditor o
otherwise 0.
Export sector is measured as a dummy variable equal to 1 if the rm belongs to the garments/textile sector or otherwise 0.
Dummy variables for different industrial sectors
INDSPEC
FAMILYCON
FAMILYOWN
FAMCEO
INSIDEOWN
INSTOWN
GOVOWN
AUDCOMPLX
LEV
SIZE
BOARDIND
PROF
END
SUBD
MNC
NAFD
EXPORT
INDUM
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-nancial 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 rm performance: evidence from the S&P 500. Journal of Finance, 58(3), 1301e1328.
Ashbaugh, H., & Wareld, 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 rms 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 rms. 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 inuence 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 rms. 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.
17
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
nancial 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 rms. 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 nancial markets around the
world. Asia-Pacic Journal of Accounting and Economics, 10(1), 1e30.
Goldman Sachs. (2011). It is time to re-dene 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 nancial structure and managerial incentives. In Anonymous (Ed.), The economics of information and uncertainty (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 rms: 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_id1613454 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
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 rms. The International Journal of
Niskanen, M., Karjalainen, J., & Niskanen, J. (2010). The role of auditing in small, private family rms: 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 rms. British Accounting 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 rms. 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 afliates 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 Certied 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 rm 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.

Mano

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Mano

Hochgeladen von

Copyright:

Verfügbare Formate

The British Accounting Review xxx (2015) 1e17

Contents lists available at ScienceDirect

The British Accounting Review

Audit fees, auditor choice and stakeholder inuence:

Department of Accounting, Faculty of Business and Law, Deakin University, Australia

* Corresponding author. Tel.: 44 161 2750440.

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

Fig. 1. Institutional context and hypothesis development.

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

Percent of family rms (rmyear)

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

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

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

Panel B: family and non-family rm subsamples

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

AUDIT FEE (H1)

AUDIT FEE (H1)

AUDIT FEE (H3A)

AUDIT FEE (H1)

AUDIT FEE (H1)

AUDIT FEE (H3A)

4.3. Relationship between family rms and quality auditor choice

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

Model 1(ordered probit)

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

AUDIT FEE (H1)

AUDIT FEE (H1)

Model 3(ordered probit)

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

A. Khan et al. / The British Accounting Review xxx (2015) 1e17

Das könnte Ihnen auch gefallen