political connections, related party transactions, … connections, related party transactions, ......
TRANSCRIPT
Accepted Manuscript
Political connections, related party transactions, and auditor choice: Evidence
from Indonesia
Ahsan Habib, Abdul Haris Muhammad, Haiyan Jiang
PII: S1815-5669(17)30004-8
DOI: http://dx.doi.org/10.1016/j.jcae.2017.01.004
Reference: JCAE 102
To appear in: Journal of Contemporary Accounting & Economics
Received Date: 5 March 2016
Accepted Date: 12 January 2017
Please cite this article as: Habib, A., Haris Muhammad, A., Jiang, H., Political connections, related party transactions,
and auditor choice: Evidence from Indonesia, Journal of Contemporary Accounting & Economics (2017), doi: http://
dx.doi.org/10.1016/j.jcae.2017.01.004
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Political connections, related party transactions, and auditor choice:
Evidence from Indonesia
Ahsan Habib*
School of Accountancy
Massey University
Private Bag 102904
Auckland, New Zealand
Phone: +64 9 4140800
Email: [email protected]
Abdul Haris Muhammad
Staff of Directorate of Tax Regulations I, Directorate General of Taxes
Ministry of Finance, Indonesia
Jl. Gatot Subroto Kav.
40-42, Jakarta Selatan, Jakarta
Indonesia 12190 Email: [email protected]
&
Haiyan Jiang
School of Accountancy
Massey University
Private Bag 102904
Auckland, New Zealand
Email: [email protected]
*Contact author
We thank an anonymous reviewer for providing numerous constructive comments on the manuscript.
We also thank the Editor, Ferdinand Gul for comments.
2
Political connections, related party transactions, and auditor choice:
Evidence from Indonesia
Abstract
This paper investigates how political connections in concert with related party transactions
(RPTs) determine auditor choice in Indonesia. Our study is motivated by conflicting findings
in the literature on whether politically connected firms appoint reputable auditors (Big 4
auditors). On one hand, politically connected firms are less likely to appoint Big 4 auditors if
they wish to cover up RPT-related tunneling activities by providing financial statements that
fail to reflect their true economic performance. On the other hand, politically connected
insiders who refrain from self-dealing would prefer higher-quality financial reporting and,
hence, appoint Big 4 auditors. Using data from Indonesia, we find support for the former. By
documenting the role of RPTs as a motivating factor for politically connected firms to choose
non-Big 4 auditors, we enrich the political connection and auditor choice literature.
Keywords: Political connections; Related party transactions (RPTs); Indonesia; Auditor
choice
JEL classification: G3, M0, M4
3
1. Introduction
This paper investigates how political connections in concert with related party
transactions (RPTs) determine auditor choice in Indonesia. Competing arguments exist
regarding auditor choice by politically connected firms. On one hand, connected firms might
prefer to appoint non-Big 4 auditors to mask their tunneling and rent-seeking activities as
manifested in their financial statements. Such tunneling and rent seeking activities are
undertaken by politically connected firms in order to establish and maintain their political
connections (Morck et al., 2000; Ma et al., 2013). Given the value of political connections
(e.g. Fisman, 2001), politically connected firms may manipulate accounting numbers to
conceal true economic performance, in order to ensure that their diversionary practices
remain a secret (Guedhami et al., 2014). Appointing non-Big 4 auditors is a step towards that
direction.
On the other hand, connected insiders who refrain from self-dealing would prefer to
choose Big 4 auditors, so that outside investors will value their reporting transparency
positively. It is well accepted that investors value accounting transparency in order to identify
managerial misconduct detrimental to their value maximization interests (Dyck and Zingales,
2004). Following the arguments in the literature that Big 4 auditors make financial statements
more credible (DeFond and Zhang, 2014), it can be argued that connected firms would prefer
appointing Big 4 auditors. In line with this idea, Fan and Wong (2005) claim that high
quality auditors might be hired by the management to increase the credibility of their
financial reports in emerging markets.
Firms with incentives to tunnel resources from minority shareholders require channels
through which this can be achieved. We consider RPTs as one such tunneling channel, and
investigate whether connected firms with certain types of RPTs differ in terms of auditor
choice, compared to their non-connected counterparts. A related party transaction is a transfer
4
of resources, services, or obligations between related parties, regardless of whether or not a
price is charged (International Accounting Standards 24.9) (IASB, 2009), where a related
party is a person or entity related to the entity preparing its financial statements. These
transactions are diverse, and often complex, business transactions between a firm and its own
managers, directors, principal owners or affiliates, and are reasons for concern because they
violate arm’s-length market transaction principles. There is abundant empirical evidence in
the extant literature that supports the opportunistic incentive for engaging in RPTs, in
particular, RPT loan and loan guarantees (Jian and Wong, 2010; Kohlbeck and Mayhew,
2010; Chen et al., 2011; Ryngaert and Thomas, 2012; Ying and Wang, 2013).1 Since it is
easier for politically connected firms to conduct RPTs, because of their large numbers of
affiliates with complex interrelationships, we propose that politically connected firms
undertaking opportunistic RPTs would hire non-Big 4 auditors.
However, an alternative argument could be advanced from the supply side, i.e.,
auditors’ assessment of the audit risk associated with politically connected firms. Auditors’
evaluation of engagement risk is aimed at minimizing the auditor’s exposure to litigation loss,
loss of reputation, as well as regulation risk (Knechel et al., 2007). Since selection of clients
by auditors is non-random, good quality auditors might avoid auditing risky clients, in our
case, politically connected firms, because of their strong incentives for engaging in tunneling
and rent-seeking activities. This perspective suggests that auditors’ choice of clients might
drive the results, instead of the discretionary appointment of auditors by politically connected
firms. Since Big 4 auditors are of high quality, they should be more concerned about their
clients’ opportunistic business transactions, including opportunistic RPTs, and may choose to
avoid clients with political connections. This argument might also suggest a negative
association between political connections and auditor choice. However, empirical evidence
1 For example, Kohlbeck and Mayhew (2010) document a detrimental effect of disclosed RPTs on firm
valuations and subsequent returns when compared with non-RPT firms.
5
from Malaysia suggests that auditors charge higher audit fees for politically connected firms
instead of resigning from the engagement (Gul, 2006).
Guedhami et al. (2014), an important related paper published recently, find that
politically connected firms are more likely to appoint Big 4 auditors compared to their non-
connected peers. We, on the other hand, propose that politically connected firms have
incentives to appoint non-Big 4 auditors. This is due to our consideration of firm-level RPTs
that encourage politically connected firms to siphon resources from minority shareholders.
Since Guedhami et al. (2014) did not investigate the possibility that connected firms might
engage in rent-seeking using opportunistic RPTs, our prediction of a negative association is
justified. That is, Guedhami et al. (2014) considered the incentives for connected firms to be
transparent, whereas we propose that connected firms might extract rents and, hence, choose
non-Big 4 auditors.
Indonesia offers an interesting setting to explore this research due to its unique
institutional features. First, political connections affect firm value in Indonesia (Fisman,
2001). By using an event study of rumors of the health of former President Suharto during
1995 to 1997, Fisman (2001) finds that returns of shares having a close relationship with
Suharto lost value compared to those having no affiliation with the Suharto government. In
addition, Leuz and Oberholzer-Gee (2006) document that the performances of politically
connected firms in Indonesia fluctuate following the fortunes of their connections. Second,
Indonesia has a high ownership concentration (Claessens et al., 2000; Brown, 2006), with an
average 16.6% of market capitalization being confined in the hands of a single family:
Sudono Salim, a closed ally of Suharto (Brown, 2006). This number rises to a staggering
57.7% for the top ten families in Indonesia, centered on Suharto political and military
connections (Claessens et al., 2000; Brown, 2006). Such a high ownership concentration
gives rise to a type II agency problem, i.e., the risk of expropriation of minority shareholders’
6
resources by their controlling owners. Third, RPTs are significant in Indonesia, as more than
90 per cent of listed firms in Indonesia conduct various forms of RPTs (Utama and Utama,
2013).2 In addition, Indonesian listed companies are part of complex economic groups with
interlocking directorships, reciprocal ownership arrangement and excessive cross-
ownerships, which expedites RPTs. Although the Capital Market and Financial Institutions
Supervisory Agency requires listed firms to disclose RPTs, companies’ complex ownership
structures and lack of transparent disclosures makes the monitoring of RPTs difficult
(Wulandari and Rahman, 2005). Habib et al. (2016, forthcoming) find that politically
connected firms conduct more opportunistic RPTs compared to their non-connected
counterparts and further engage in income-increasing earnings management to mask such
opportunistic RPTs. This evidence corroborates politically connected firms’ incentives to
appoint non-Big 4 auditors in Indonesia.
For our study, a firm-year observation is categorized as politically connected if at
least one of its large shareholders (having control of at least 10 per cent of voters directly or
indirectly), or its board of directors or board of commissioners, is a current or former (a)
member of the parliament, (b) minister or head of a local government, or (c) closely related to
a politician or party. We further decompose political connections into one of three mutually
exclusive categories: namely, government connections, military connections, and Suharto
connections.
Using a panel data of 1,428 RPT firm-year observations from 2007 to 2013, we find
that politically connected firms with RPTs in Indonesia tend to choose non-Big 4 auditors.
2 The Indonesian Capital Market Supervisory Agency (ICMSA), requires listed firms to announce RPTs to
the public as well as to the ICMSA no later than two working days after the transactions are undertaken if
the RPT has a value larger than 0.5% of the firm’s paid capital and greater than IDR 5,000,000,000
(US$509,840). Hence, most of individual RPTs are arranged below these thresholds in order to avoid public
announcement (Utama and Utama, 2014). However, RPTs having a value less than 0.5% of the firm’s paid capital still need to be reported to the ICMSA and be disclosed in the notes to financial statements (See
Appendix II for further details on RPT regulations in Indonesia).
7
This negative association is more pronounced for firms with government connections as
opposed to firms with military connections. Firms with Suharto connections, on the other
hand, tend to choose Big 4 auditors. Our results remain robust to controls for a potential self-
selection problem.
We extend the auditor choice research by investigating how the choice of auditors in
Indonesia is systematically affected by firms’ political connections (See Appendix I for a
discussion on the external auditing environment in Indonesia). We further contribute to the
auditor choice literature by documenting the important role played by firm-level RPTs. This
latter finding is particularly insightful as this provides a contextual explanation for why an
opposite result to that of Guedhami et al. (2014) can be expected. Finally, we enrich the
political connection literature, as applied in auditing, by examining the effects of three
mutually exclusive categories of political connections separately. We find that firms having
connections with the government are more inclined to choose non Big 4 auditors.
The remainder of the paper proceeds as follows. Section 2 reviews the relevant
literature and develops hypotheses. Section 3 describes research design followed by sample
selection and descriptive statistics in Section 4. The following section explains main test
results and Section 6 concludes the paper.
2. Literature Review and hypotheses development
The classic agency problem between shareholders and managers gives rise to the
hiring of auditors, who provide independent assurance to corporate stakeholders that financial
statements prepared by corporate managers comply with generally accepted accounting
principles (GAAP) (Watts and Zimmerman, 1983). Auditing also plays a significant role in
enforcing and protecting investors’ rights, by detecting expropriation by insiders (Newman et
8
al., 2005), and benefits management by signalling the reliability of management-provided
financial information. A firm’s decision to appoint a certain type of auditor is, therefore, a
crucial element of the auditing landscape.
Previous research on the possible determinants of auditor choice (both brand name
and industry specialist auditors) include culture (Hope et al., 2008), firm-level governance,
e.g., ownership structure, country-level investor protection (Fan and Wong, 2005; Wang et
al., 2008; Guedhami et al., 2009; He et al., 2014), managerial incentives (Chen et al., 2015),
and political connections (Guedhami et al., 2014). Our study belongs to the ‘political
connection’ domain.
Many benefits accrue to politically connected firms: preferential access to lenders
(Johnson and Mitton, 2003; Khwaja and Mian, 2005; Faccio, 2006; Leuz and Oberholzer-
Gee, 2006; Boubakri et al., 2012a); low cost of debt and equity (Houston et al., 2014); high
likelihood of being bailed out (Faccio et al., 2006); profitable government contracts
(Goldman et al., 2009); favorable regulations (Goldman et al., 2009); less monitoring and
oversight (Faccio, 2006); lower taxes (Faccio, 2006; 2010); and preferential import licenses
and tariffs (Goldman et al., 2009). On the other hand, political connections are also viewed as
harmful to the minority shareholders, as these connections can lead to rent-seeking activities
(Frye and Shleifer, 1997; Faccio, 2006; Boubakri et al., 2012b), tunneling (Qian et al., 2011),
and earnings management (Chaney et al., 2011).
In considering whether to choose a Big 4 auditor, controlling shareholders will assess
potential benefits derived, and costs incurred, as a result of their choice (He et al., 2014).
Controlling shareholders might voluntarily adopt bonding mechanisms in dealing with
adverse pricing and a high cost of capital caused by asymmetric information and illiquidity
(Jensen and Meckling, 1976). Fan and Wong (2005) claim that external independent auditors
might be hired as monitors, or as one of the bonding mechanisms, designed to alleviate
9
information and agency problems. Big 4 auditors might offer high quality audit because they
have better monitoring capability (Watts and Zimmerman, 1983), are keen to maintain their
reputation, and are subject to heightened litigation (Hope et al., 2008; Guedhami et al., 2014).
In addition, Big 4 auditors might deliver a high quality of assurance consistently, owing to
their global operation (Humphrey et al., 2009; Guedhami et al., 2014). For politically
connected firms, being audited by Big 4 auditors can, therefore, reap benefits, such as better
transparency, higher valuation, lower earnings management, and cheaper cost of capital
(Guedhami et al., 2014; He et al., 2014).
On the other hand, the appointment of a Big 4 auditor may be costly, since controlling
shareholders have fewer opportunities for expropriation owing to the significant role of Big 4
auditors in monitoring financial reporting discretion of controlling shareholders (Guedhami
et al., 2009). Politically connected firms expecting to obtain benefits from their allies, tend to
appoint non-Big 4 auditors for the following reasons. First, connected firms are inclined to
render more opaque financial statements in order to conceal their tunneling and rent seeking
activities (He et al., 2014; Chen et al., 2015). Second, any increase in transparency will
decrease the ability of controlling shareholders and their political allies to enjoy the private
benefits of control (Leuz and Oberholzer-Gee, 2006; Piotroski et al., 2015). For politically
connected firms, transparency is costly, since it might result in unwanted scrutiny, restricting
the possibility of exploiting weak corporate governance (Bona-Sánchez et al., 2014;
Piotroski et al., 2015). Further, He et al. (2014) claim that when controlling shareholders
develop political connections, their political partners need more secrecy, so that their
reputations are maintained at the expense of financial reporting transparency. Third, the lack
of transparency of non-Big 4 auditing does not reduce their chance of getting easy credit from
state-owned banks anyway (Dinç, 2005; Guedhami et al., 2014). Supporting this idea,
Bushman et al. (2004) claim that in return for bribes, nepotism, and other political supports,
10
the politicians might exercise their control over state owned banks to provide preferential
financing to their connected firms. Financial reporting transparency may have a minimal role
in this setting, diminishing the incentive for appointing a more expensive Big 4 auditor.
Moreover, Leuz and Oberholzer-Gee (2006) find that Indonesian firms with close
connections to the state avoid raising capital from abroad with more onerous disclosure
requirements that include, but are not limited to, Big 4 audits of the financial statements.3
H1: Politically connected firms are less likely to appoint Big 4 auditors.
However, a supply side argument concerning auditor choice of clients may also be
relevant in understanding the association between political connection and auditing in
Indonesia. Because of the non-random selection of clients by auditors, good quality auditors
might avoid auditing risky clients, in our case, politically connected firms, because of their
tunneling and rent-seeking activities. An extensive body of literature supports the view that
Big 4 auditors provide higher audit quality because of their exposure to greater litigation risk
3 Another perspective holds that politically connected top managers could give outside investors an
impression that the government favors the connected firms, which could conceivably be more effective
than hiring high-quality auditors. In such a case, outside investors would pay less attention to audit quality, so that the firms would be less motivated to hire high-quality auditors. Investors might also agree that
building political connections is more cost-effective than hiring reputable auditors.
11
(“deep pocket” theory) and greater reputation risk (independence concern) (DeFond and
Zhang, 2014). This argument suggests that Big 4 auditors should be more concerned about
their clients’ opportunistic business transactions, and may choose to avoid risky clients to
minimize their exposure to litigation and reputational losses. However, high quality auditors
also tend to be larger, with the capacity to diversify client risk and, thus, are better able to
bear the risk of such clients. Hence it is not obvious whether high quality auditors avoid risky
clients (DeFond and Zhang, 2014). Empirical evidence, too, finds support in favor of
retaining politically connected firms, but charging a fee premium (Gul, 2006).
In Indonesia, connected firms can be classified further into government, military, and
Suharto connections. Our sample covers two consecutive periods of Susilo Bambang
Yudhoyono (SBY)’s presidency, from 2004 to 2014. With respect to government connection,
extant literature, argues that government plays a key role in controlling and allocating key
resources (Child, 1994; Li et al., 2008). Firms willing to maintain an ongoing relationship
with government need to share the rents extracted through expropriation of minority
resources and, as well, obfuscate their financial reports to mask tunneling activities.
Appointing a non-Big 4 auditor is a proactive decision to accomplish this.
H1A: Firms having political connections to government are more likely to hire non-Big 4
auditors.
On the other hand, Suharto-connected firms are more likely to appoint Big 4 auditors.
When Suharto was in power, firms having an affiliation with his regime through his families,
friends, and military connections enjoyed ample privileges (Brown, 2006), e.g., preferential
loans from state owned banks through memo-lending and exclusive import licenses (Leuz
and Oberholzer-Gee, 2006). However, after Suharto’s resignation, firms having connections
with Suharto had reduced access to government officials. They had difficulties in establishing
12
a connection with the new government, and experienced loss of government contracts,
distributorships, and brokerage monopolies (Fukuoka, 2013).
With respect to military influence in Indonesia, it has been observed that, during the
Suharto regime both active and former military personnel held strategic posts at the national
and regional level, including managerial positions in state owned enterprises (Brown, 2006;
Bhakti et al., 2009; Sebastian and Iisgindarsah, 2013). Suharto handed over state owned
enterprises, previously seized from Dutch companies, to be managed by military personnel.
However, with the end of the Suharto era, foundations belonging to the military, Suharto’s
family and Golkar were under investigation (Brown, 2006). Therefore, Mietzner (2006)
concludes that the army have lost formal political influence considerably, and they do not
serve as a backbone for the incumbent regime anymore. With these benefits gone, firms
having Suharto as well as the military connections had less incentive to engage in tunneling
and financial report manipulation in order to obfuscate such tunneling. Based on these
arguments we hypothesize the following:
H1B: Firms having Suharto as well as military connections are more likely to hire Big 4
auditors compared to firms with government connections.
Note that H1 is opposite to what Guedhami et al. (2014) established. This is due to
our consideration of firm-level RPTs that encourage politically connected firms to siphon
resources from minority shareholders. Since Guedhami et al. (2014) did not investigate the
possibility that connected firms might engage in rent-seeking using opportunistic RPTs, our
prediction of a negative association is justified. That is, Guedhami et al. (2014) considered
the incentives for connected firms to be transparent, whereas we propose that connected firms
might extract rents and, hence, choose non-Big 4 auditors.
Our next hypothesis explicitly considers the mechanisms through which connected
firms might conduct tunneling activities, and whether that had a bearing on auditor choice by
13
connected firms. We propose that RPTs is one such channel through which politically
connected firms might conduct resource diversion. RPTs are used to structure transactions,
e.g., tunneling, propping or earnings management, among their affiliates in order for insiders
to expropriate minority resources (Thomas et al., 2004; Cheung et al., 2009). In the context of
Indonesia, Habib et al. (2016, forthcoming) document that politically connected firms
conduct more RPTs compared to their non-connected counterparts, and that this effect is
more pronounced for firms with government connections. Further analysis reveals that the
connected firms use RPTs to tunnel resources, and engage in income-increasing earnings
management to mask such tunnelling activities. Since it is easier for politically connected
firms to conduct RPTs owing to the presence of a large number of affiliates with complex
inter-relationships among them, we propose that politically connected firms with
opportunistic RPTs would hire non-Big 4 auditors.
Extant literature has documented that opportunistic RPTs are primarily conducted
through RPTs involving loan guarantees and capital transfers (RPLOAN). Prior literature
reveals that RP loans and guarantees have been used by parent companies for tunnelling or
siphoning resources out of their listed subsidiaries (Berkman et al., 2009; Jiang et al., 2010).
Empirical evidence shows that, compared to those with low levels of RP loans and
guarantees, Chinese firms with high levels of RP loans and guarantees demonstrate
significantly worse future performances including sharp declines in profitability, and a higher
likelihood of entering financial distress in the future (Jiang et al., 2010). Habib et al. (2016,
forthcoming) find that connected firms use primarily RP loans and guarantees to tunnel
resources in Indonesia. RP loans are generally not made as part of the normal course of
business, and most loans do not accrue interest. If RPTs involving loans allow politically
connected firms to siphon resources, then there is an incentive for those firms to manipulate
financial reports in order to obfuscate true economic performance. This argument suggests
14
that politically connected firms with RPT loans are more likely to choose non-Big 4 auditors.
It is important to note that this argument does not imply that non-connected firms don’t use
RPT loans for opportunistic reasons. However, we do expect the effect to be more
pronounced for connected firms than for their non-connected counterparts. The following
hypothesis is developed:
H2: Firms with RP loan and guarantees are less likely to appoint Big 4 auditors and this is
more pronounced for politically connected firms.
Because we hypothesize that politically connected firms conduct opportunistic RPTs
for rent-seeking activities, the preceding hypotheses are valid in the context of such RPTs.
3. Research Design – Empirical Model
To test H1, we develop the following regression model:
)1.......(||1312
1110987
6543210
YearFEIndustryFEDACIC
ACROAFINANCESEGMENTINV
GROWTHLEVSIZEFOWNOWNCONPCONAUDITOR
where AUDITOR is dummy variable, coded 1 for Big 4 auditors and zero otherwise.
The local Indonesian audit firms that are affiliated with the Big 4 audit firms are:
Tanudiredja, Wibisana and Rekan (PWC); Purwantono, Suherman and Surja (EY); Osman
Bing Satrio and Rekan (Deloitte); and Siddharta Siddharta and Widjaja (KPMG). PCON is an
indicator variable coded 1 if the sample observations have political connections, 0 otherwise.
We expect a negative and significant coefficient on PCON to suggest that politically
connected firms will choose non-Big 4 auditors.
We include a set of control variables based on prior literature on the determinants of
auditor choice (DeFond et al., 2000; Wang et al., 2008; Guedhami et al., 2009, 2014). Larger
15
firms (SIZE), firms with growth opportunities (GROWTH), firms with more inventories in
their balance sheet (INV), multi-segment firms (SEGMENT), new insurance of finance
(FINANCE), profitable firms (ROA), firms with larger foreign ownership (FOWN), and less-
leveraged (LEV) firms are more likely to appoint a Big 4 auditor. FOWN is the total
percentage of shares owned by foreign institutional investors. Foreign institutional investors
prefer Big 4 auditors to ensure the quality and the credibility of financial statements
(Guedhami et al., 2009; He et al., 2014). The association between ownership concentration
(OWNCON) and auditor choice is also expected to be positive (Fan and Wong, 2005).
SIZE is measured as the natural log of total assets. GROWTH is the market value of
equity divided by the book value of equity. ROA is net income divided by total assets. LEV is
the total long-term debt divided by total assets. AC is dummy variable coded 1 if the firm has
established an audit committee, and zero otherwise. We expect a positive coefficient on AC,
since independent audit committees demand higher quality audits (Carcello et al., 2002;
Abbott et al., 2003). IC is dummy variable coded 1 for a firm having independent
commissioners, and zero otherwise.4 We expect a negative and significant coefficient on
discretionary accruals (DAC) to suggest that the firms having higher DAC will prefer to
appoint non-Big 4 auditors in order to convey less information on their DAC. Appendix III
provides the variable definitions.5
4 Indonesia adopts a two-tier system for board structures: the board of directors (BODs) and the board of
commissioners (BOCs). The BODs serve as firm’ executives whereas the BOCs have responsibility to
supervise management policies and to advice the BODs. The role of the BOCs in the two-tier system is
comparable to that of the BODs in a one tier system. In order to improve corporate governance in
Indonesia, the Capital Market and Financial Institutions Supervisory agency requires listed firms to have
independent commissioners and audit committees. However, those boards did not carry out their legal
duties properly because both BOCs and BODs represent the interests of, and are often selected by the15
families who control the majority of the listed companies (Wulandari and Rahman, 2005). 5 We did not include some other control variables as in Guedhami et al. (2014) for reasons outlined below.
Data on FOREIGNSALE is not reported by companies in their published annual report. We did not include
CROSS-LIST because there are only two companies that are cross-listed [Aneka Tambang (Persero) Tbk
and Telekomunikasi Indonesia (Persero) Tbk]. CONTROLRIGHT (the percentage of voting rights belonging to the ultimate owner ) and CASHFLOWRIGHT (the percentage of voting rights belonging to
the ultimate owner) are unavailable because listed firms in Indonesia are only obliged to disclose their
legal owners instead of their beneficial or ultimate owners (Utama and Utama, 2013).
16
To estimate DAC we use the cross-sectional modified Jones model, controlling for
firm performance (Kothari et al., 2005). We estimate the following model for all firms in the
same industry (using the SIC two-digit industry code) with at least eight observations in an
industry in a particular year:
)2......(..............................)()/(]/)[()/1(/
13
1211101
tt
tt tttttt ROA
TAPPE TA RECEIVABLE SALES TA TAACC
where ACC is total accruals, calculated as earnings before extraordinary items and
discontinued operations minus operating cash flows; TA is total assets in year t-1; ΔSALES is
change in sales from year t-1 to year t; ∆RECEIVABLE is change in accounts receivable from
year t-1 to year t; PPE is gross property plant and equipment; ROA is the prior year's return-
on-assets measured as earnings before extraordinary items and discontinued operations,
divided by total assets for the previous year. The coefficient estimates from Equation (2) are
used to estimate the non-discretionary component of total accruals (NDAC) for our sample
firms. The discretionary accruals are then the residual from equation (2), i.e. DAC=ACC-
NDAC.
An interesting aspect of the political connection landscape in Indonesia relates to the
different types of connection, which are not captured in a single PCON variable. We
categorize three mutually exclusive types of political connection: GCON (dummy variable, 1
for government connected firms, 0 otherwise); MCON (dummy variable, 1 for military
connected firms, 0 otherwise); and SCON (dummy variable, 1 for Suharto connected firms, 0
otherwise). The regression equation below is, therefore, estimated:
....(3).......... IndustryFEYearFEDACIC ACROAFINANCESEGMENTINVGROWTHLEVSIZE
FOWNOWNCONSCONMCONGCONAUDITOR
||1514131211
109876
543210
17
We expect the coefficient on GCON to be negative, and that on MCON and SCON to
be positive.
Finally, we develop the following regression model to test for the effects of RPTs on
auditor choice, conditional on political connection:
)4...(||
*
151413
121110987
6543210
IndustryFEYearFEDACIC ACROAFINANCESEGMENTINVGROWTHLEV
SIZEFOWNOWNCONRPTPCONRPTPCONAUDITOR
We use three categories of RPTs. RPT_RATIO is ratio of gross RPT over total assets.
Deflation by assets is justified since RPTs vary with firm size. OPRPT_RATIO is operating
RPTs deflated by total assets. Finally, LOAN_RATIO is RPT loan deflated by total assets.
Operating RPTs, dealing with sales and purchases of goods and services, consist mainly of
trade relationships. From an efficiency-enhancing perspective, operating RPTs are argued to
be used within corporate groups as a way of optimizing internal resource allocation, reducing
transaction costs (Jian and Wong, 2010; Chen et al., 2012) and improving return-on-assets
(Khanna and Palepu, 2000). This is in contrast to RPT loan which is primarily a mechanism
for siphoning resources. Our variable of primary interest is the interactive variable
PCON*RPT. If politically connected firms make use of RPTs to siphon resources
(LOAN_RATIO), then we would expect a negative and significant coefficient on
PCON*LOAN_RATIO. The coefficient on PCON*OPRPT_RATIO is expected to be positive.
The coefficient on γ2 captures auditor choice by non-connected firms in the presence of
RPTs.
4. Sample selection and descriptive statistics
Data on the number and amount of RPT is hand-collected from audited financial
reports downloaded mainly from the website of the Indonesia Stock Exchange
18
(http://www.idx.co.id/index-En.html). If not available, the data are derived from the websites
of Indonesian listed firms. In addition, the following corporate governance data are also
manually collected from audited financial statements or annual reports: board of directors,
board of commissioners, independent commissioners, audit committee, the names of audit
firms, the names and percentage of share ownerships, and information on reportable
segments. Financial statement data are collected from the Research Insight-Global Vantage
database. Since most of the data for RPT are in Indonesian Rupiah, the data for RPT are
translated into US$ by using the exchange rate available from the DataStream. Finally,
market data are retrieved from DataStream.
The criteria for defining politically connected firms follows Faccio (2006), Chaney et
al. (2011), and Guedhami et al. (2014) with necessary modification to the Indonesian context.
A firm-year observation will be categorized as politically connected if at least one of its large
shareholders (having at least 10 per cent direct or indirect voting rights) or its board of
directors and board of commissioners is a current or former (a) member of parliament, (b)
minister or head of a local government, or (c) is closely related to a politician or party.
Connection with government ministries is extended to close relatives (spouse, sons or
daughters, and other immediate family relationship). Close relationships with politicians or
parties encompass well-known friendships as identified by The Economist, Forbes or
Fortune; share ownership or a directorship held by former ministers, former heads of
government, former member of parliament and current politicians (Faccio, 2006; Chaney et
al., 2011); well documented relationships with political parties as utilized by Johnson and
Mitton (2003); and famous connections adopted by Fisman (2001) and Leuz and Oberholzer-
Gee (2006).
To establish those political connections, the names of boards of directors, boards of
commissioners and data of ownerships, including the name and the percentage of ownership,
19
were gathered from the Indonesia Stock Exchange (http://www.idx.co.id/index-En.html),
company websites, audited financial reports and annual reports. The names of members of
parliament were collected from the website of the Indonesia House of Representatives
(http://www.dpr.go.id/id/anggota/), the names of members of cabinet were gathered from the
website of the cabinet secretariat of the Republic of Indonesia (http://setkab.go.id/en/profil-
kabinet.html). The names of heads of local government (governors) were collected from
(http://www.kemendagri.go.id/staff-directory/gubernur-dan-wakil-gubernur). The names of
members of parliament, members of cabinet, and heads of local government were matched
with the names of boards of directors, boards of commissioners and shareholders. In addition,
political connections could also be identified from the profiles of members of boards of
commissioners and directors described in the annual reports. Our primary variable of interest,
PCON, is an indicator variable coded 1 for the firm-year observation fulfilling at least one of
the abovementioned criteria, and 0 otherwise. We further categorize politically connected
firms into three mutually exclusive categories: namely, government connections, military
connections and Suharto connections.
The sample period is from 2007 until 2013, which covers two periods of the first
directly elected president, Susilo Bambang Yudhoyono (SBY)’s presidency, i.e., 2004-2014.
The first period of President SBY’s administration was from 20 October 2004 until 20
October 2009 and the second period was from 20 October 2009 until 20 0ctober 2014. The
financial year: 2014, is excluded from the observation owing to a political regime change
whereby a newly elected government took power.
Table 1, Panel A illustrates our sample selection process. The number of listed non-
financial firms on the Indonesian Stock Exchange in 2004 was 244 firms. This number
increased to 405 firms in 2013 giving us a total initial sample of 3,149 firm-year observations
for the period 2004 to 2013 inclusive. We deleted 772 firm-year observations pertaining to the
20
2004 to 2006 sample period because we could retrieve valid data for only 25 firm-year
observations.6 From the remaining sample of 2,377 firm-year observations we deleted 481
firm-year observations with unavailable audit reports. This result in a total 1,896 firm-year
observations for matching with other variables required to run a regression on RPTs. We then
deleted 113 firm-year observations with zero RPT values and 85 firm-year observations with
negative equity values. Finally, we deleted a further 269 firm-year observations with missing
data on the relevant control variables, resulting in a final usable sample of 1,429 firm-year
observations for the period 2007 to 2013.
Panel B, Table 1 presents descriptive statistics for the variables used in the
regressions. We winsorize the continuous variables at 1% and 99% of their respective
distributions to control for the effects of outliers. About 43% of the firm-year observations
are audited by a Big 4 audit firm. Thirty nine percent of the observations have political
connections, split among GCON (24%), MCON (12%) and SCON (3%). The average
RPT_RATIO (Gross RPT/Total assets) is 0.50 of total assets. The corresponding values for
OPRPT_RATIO (Gross Operating RPTs/Total assets) and LOAN_RATIO (Gross
RPT_LOAN/Total assets) are 0.28 and 0.11 respectively.
These two categories do not add up to 0.50 because there are RPTs that do not belong
to operating and loan RPTs. However, we do not include other RPTs in our regression
specification because of a lack of theoretical prediction regarding its impact on auditor choice
by politically connected firms. Sample firms have growth opportunities (mean GROWTH is
6 We contacted the Financial Services Authority (FSA), formerly known as the Indonesian Capital Market
and Financial Institution Supervisory Agency, regarding the availability of hard copy annual reports. Since
05 July 2011, the FSA requires listed firms to submit both hard copy and soft copy audited financial
reports. However, the FSA does not allow public access to those audited financial reports. The Indonesia
Stock Exchange also used to receive hard-copies of audited financial reports, but those are kept in storage
outside of Jakarta. Currently, it maintains only soft-copies that can be downloaded from its website: a
procedure that we followed in collecting more recent annual reports.
21
3.17), are low-levered (an average of 0.13), and profitable (average ROA of 7%). Eighty one
percent of the sample observations have established an audit committee.
The industry distribution of sample companies is presented in Panel C, Table 1,
revealing that materials account for 23.79% of the total sample observations, followed by
consumer discretionary and industrials with 20.08% and 17.35% of sample observations
respectively. Finally, Panel D presents a univariate test of the difference in means for the
variables between connected and non-connected firms. The proportion of firm years being
audited by Big 4 firms is much higher for connected as opposed to non-connected firms (an
average of 0.53 versus 0.37, t-stat of difference in means is 5.53). This is contrary to our
hypothesized negative association between PCON and AUDITOR. Politically connected firms
are larger, more levered, higher growth firms. There is no significant difference in gross RPT
ratios between connected and non-connected firms although OPRPT_RATIO is significantly
larger for the non-connected compared to their connected firm counterparts (t-stat of
difference in means is -1.72, p<0.10).
[TABLE 1 ABOUT HERE]
Correlations among the variables for auditor choice are presented in Table 2. The
correlation between PCON and AUDITOR is positive and significant at better than the 1%
level. This is contrary to our hypothesized negative association between PCON and
AUDITOR. Most of the independent variables are correlated positively with the dependent
variable, AUDITOR, at better than the 1% level. Untabulated correlation reveals that MCON
and SCON are positively correlated with AUDITOR, whilst GCON is negatively correlated.
[TABLE 2 ABOUT HERE]
5. Main test results
5.1 Political connections and auditor choice
22
In a multivariate regression framework, we estimate the impact of political
connections on the likelihood that firms will hire a Big 4 auditor to examine the prediction in
H1. Next, we analyze the prediction in H2 that RPTs will moderate the association between
political connections and auditor choice. Table 3, Panel A presents the regression results for
the association between political connections and auditor choice (Big 4 versus non-Big 4)
(Columns 1 and 2) and the effect on that association of gross RPTs (Columns 3 to 5). We
control for industry and year fixed effects in all the regression models. We also report all
regression results with clustered standard errors at the firm level (Petersen, 2009; Gow et al.,
2010).
Our baseline regression in Column (1) reveals that the coefficient on PCON is
negative and significant (coefficient -0.363, z-statistic -2.22, p<0.05), suggesting that
politically connected firms are less likely to hire Big 4 auditors compared to their non-
connected counterparts. In terms of economic significance, the coefficient estimate of -0.363
means that political affiliations decrease the likelihood of appointing a Big 4 auditor by 5.8%
with all other variables are assigned their mean values.7 The negative coefficient could be
consistent with the argument that politically connected firms avoid Big 4 auditors in order to
conceal tunneling activities knowing that high quality auditors are competent to detect. This
supports H1. However, this negative coefficient could also be consistent with auditor’s risk
aversion argument, i.e., Big 4 audit firms might avoid auditing risky clients to mitigate their
7
23
litigation and reputation losses. We perform additional tests in the sensitivity test section to
rule out this possibility.8
Column (2) reports this result with GCON showing a negative and significant
coefficient (coefficient -0.877, z-statistic -4.36, p<0.01) whilst that on SCON loads positively
(coefficient 2.358, z-statistic 4.05, p<0.01). In terms of economic significance, the coefficient
estimate of -0.877 on GCON means that government political affiliations decrease the
likelihood of appointing a Big 4 auditor by about 13% when all other variables are assigned
their mean values. These findings are consistent with H1A. However, we interpret the
coefficient on SCON cautiously, given the small number of observations. Among the other
firm-level determinants, we find that firm size, ownership concentration, foreign ownership,
growth, and the existence of an audit committee are related to auditor choice positively; while
leverage and profitability are related to auditor choice negatively.
γ
γ
8 We also performed a reverse regression technique whereby we regressed PCON on AUDITOR and the
remaining control variables. The untabulated result shows the coefficient on AUDITOR to be negative but
insignificant (coefficient -0.10, z-statistic -0.64).
24
(4) and (5) consider the PCON categories, i.e., GCON and MCON,
and run Equation (4) without the interaction variable, while limiting the sample to PCON=1
only. The coefficient on RPT_RATIO is negative and significant for GCON only (coefficient -
2.417, z-statistic -4.96, p<0.01). The coefficient on MCON is insignificant and that on SCON
is undetermined because of the small sample size.
25
[TABLE 3 ABOUT HERE]
5.3 Endogeneity tests
A firm’s decision to get politically connected is not random, and unobservable factors that
affect this decision may also be associated with the propensity to choose a certain type of
auditor. Selection problem, in this case, arise because one observes only the outcome of the
choice, in this case, political connection (1/0) made but not the outcomes of choices not
made. Selection bias, which is one form of endogeneity problem, can lead to inappropriate
inferences about treatment effects (Tucker, 2011). “Selection bias due to observables” arises
from a failure to control for differences researchers can observe, e.g., size, growth,
complexity, profitability. “Selection bias due to unobservable” arises because researchers use
a small set of observations. In the extant literature Heckman two-stage error correction
method has been the most popular and widely used approach for controlling the latter bias,
whereas the ‘Propensity score matching’ (PSM) has been used to control for bias due to
unobservable factors.
26
To perform the Heckman test (1979) we proceed as follows. First we model firms’
decisions to form political connections using some observable firm characteristics based on
prior research (Faccio, 2006, 2010, Boubakri et al., 2008, Bunkanwanicha and
Wiwattanakantang, 2009). Lennox et al. (2012) argue that it is important to impose exclusion
restrictions in implementing the Heckman two-stage regression, even though the inverse
Mills ratio (IMR) can be identified by its nonlinear arguments. In other words, we need at
least one variable in the first-stage model that affects auditor choice through its effects on
political connections only.
Following Kim and Zhang (2016) and Guedhami et al. (2014), we include Industry %
of connected firms (%PCON_IND) and firms’ location (HQ) respectively, as the exclusion
variables. The variable %PCON_IND is related positively to the political connections of each
individual firm within the industry. However, we have no a priori reason to believe that
industry-level political affiliation has a direct impact on auditor choice through channels
other than political connections. Prior research suggests that firms located in capital cities are
more likely to form political connections (Agrawal and Knoeber, 2001), but are unlikely to be
associated with auditor choice. Our first-stage probit model takes the following form:
..(5)……………………………… +YearFE + FEIndustry ROASEGMENTGROWTH +LEV +SIZE +FOWN +OWNCON +HQ%PCON_IND =PCON
98
7654321
%PCON_IND is the percentage of politically connected firms in a firm’s industry
group. HQ is a dummy variable coded 1 if the firm is headquartered in the capital city and
zero otherwise. Other variables are defined as before. Table 4, Column (1) reports the first
stage estimation model. As predicted, the coefficients on %PCON_IND and on HQ, are
significantly positive [(coefficients 1.311 (p<0.1) and 0.219 (p<0.05)]. We calculate IMR
from the first stage probit model, and include it as an additional independent variable in the
27
second stage regression model. We continue to find results that are consistent with results in
Panel B in Table 3. The coefficient on the interactive variable PCON*LOAN_RATIO is
negative and marginally significant (coefficient -1.51, z-statistic -1.89, p<0.10), while that on
PCON*OPRPT_RATIO is insignificant. We find the coefficient on IMR to be insignificant
suggesting that self-selection does not confound our results.
[TABLE 4 ABOUT HERE]
Propensity-matched (PSM) technique
PSM technique is utilized to mitigate selection problem arising from observables. Matching
on firm characteristics (covariates) is ideal when the number of characteristics over which the
treated and control groups differ is limited. Rosenbaum and Rubin (1983) propose matching
by a function of covariates: the probability of an individual being selected into the program
(treatment group). This matching method is referred to as “propensity score matching”
(PSM). We use nearest neighbor (NN), and kernel technique to perform the PSM model. The
NN procedure with replacement picks a single control firm according to the closest
propensity score. Kernel matching uses the entire sample of control firms as matches, where
each unit is weighted in proportion to its closeness to the treated observation (Rosenbaum and
Rubin 1983; 1985).
Proper implementation of PSM requires both the treatment and control groups to be
similar across a number of firm characteristics excluding the main variable on which they are
expected to differ, in our case, auditor. Therefore, we first document the covariates matching,
based on the calculated propensity score
Results are reported in Panel A in Table
5. Covariate balance is achieved if both the treatment and control groups appear similar along
28
their observable dimensions, except for their choice of auditors.
Panel B, Table 5 shows the PSM regression results using the NN approach in
Columns (1) to (4) and Kernel technique in Columns (5) to (8). We find results that are
consistent with the main results: (i) politically connected firms choose non-Big 4 auditors, (ii)
firms with government and military connection choose non Big 4 auditor whereas firms with
Suharto connection choose Big 4 auditors, (iii) connected firms with RPTs choose non Big 4
auditors which is more pronounced for RP loans. Overall, our PSM analysis provides robust
evidence about the association between political connection and auditor choice and the
mediating role of firm-level RPTs.
[TABLE 5 ABOUT HERE]
6. Conclusions
This study investigates the association between political connections and auditor
choice in Indonesia. We also investigate whether firm-level RPTs moderate the association
between the two. Our study is motivated by conflicting views on whether politically
connected firms appoint reputable auditors. We find that politically connected firms in
Indonesia tend to appoint less reputable auditors. Connected firms have been found to engage
in tunneling and rent-seeking activities in order to establish and maintain their political
connections. Given the value of political connections (e.g. Fisman, 2001), politically
connected firms may manipulate accounting numbers to conceal their true economic
performance and, hence, prefer less reputable auditors. We also find that connected firms’
29
preference for appointing less reputable auditors is more pronounced when these firms
engage in significant RPTs.
The findings from this study might be generalized to other emerging countries, which
have characteristics similar to those of Indonesia, where political connections hold a
significant role, and politically connected firms conduct considerable RPTs. The findings of
the research might benefit prospective investors and minority shareholders, who have limited
information and knowledge of the true economic incentives for RPTs conducted by listed
firms that have political connections. The minority shareholders should be aware that RPTs
might be used opportunistically by controlling shareholders of politically connected firms to
carry out tunnelling that is followed by subsequent earnings management. In addition, with
regard to auditor choice, the findings of this study might benefit prospective investors and
minority shareholders regarding the rationale as to why politically connected firms appoint
non-Big 4 auditors. The research findings show that the detrimental effect to minority
interests of RPT loans conducted by politically connected firms justifies the regulatory
restrictions on RPTs and the disclosure requirements imposed on listed firms.
30
References
Abbott, L. J., Parker, S., Peters, G. F., Raghunandan, K. 2003. The association between audit
committee characteristics and audit fees. Audit. J. Pract.Th. 22, 17-32.
Agrawal, A., Knoeber, C.R. 2001. Do some outside directors play a political role? J. Law. Econ. 44, 179–98.
Armstrong, C.S., Jagolinzer, A.D., Larcker, D.F. 2010. Chief Executive Officer equity incentives and
accounting irregularities. J. Account. Res. 48, 225-271.
Basioudis, I. G., Fifi, F. 2004. The market for professional services in Indonesia. Int. J. Aud. 8, 153–164.
Berkman, H., Cole, R. A., Fu, L. J. 2009. Expropriation through loan guarantees to related parties:
Evidence from China. J. Bank. Finan. 33, 141-156. Bhakti, I. N., Yanuarti, S., Nurhasim, M. 2009. Military politics, etnnicity and conflict in Indonesia.
Centre for Research on Inequality, Human Security and Ethnicity - Working Paper, 62, 1-36.
Bona-Sánchez, C., Pérez-Alemán, J., Santana-Martín, D. J. 2014. Politically connected firms and earnings informativeness in the controlling versus minority shareholders context: European
evidence. Corp. Gov. 22, 330–346.
Boubakri, N., Cosset, J.C., Saffar, W. 2008. Political connections of newly privatized firms. J.
Corp. Fin. 14, 654–73. Boubakri, N., Cosset, J.C., Saffar, W. 2012a. The impact of political connections on firms’ operating
performance and financing decisions. J. Fin. Res. 35, 397-423.
Boubakri, N., Guedhami, O., Mishra, D., Saffar, W. 2012b. Political connections and the cost of equity capital. J. Corp. Fin. 18, 541-559.
Brown, R. A. 2006. The rise of the corporate economy in Southeast Asia. Oxon, United Kingdom,
Routledge.
Bushman, R. M., Piotroski, J. D., Smith, A. J. 2004. What determines corporate transparency? J. Account. Res. 42, 207-252.
Carcello, J. V., Hermanson, D. R., Neal, T. L., Riley, R. A. 2002. Board characteristics and audit fees.
Contem. Account. Res. 19, 365-384.
31
Chaney, P. K., Faccio, M., Parsley, D. 2011. The quality of accounting information in politically
connected firms. J. Account. Econ. 51, 58-76. Chen, C.L., Hsu, C.S., Kung, F.H. 2015. Political connections, managerial incentives and auditor
choice: Evidence from China. Pacific Account. Rev. 27, 441-465.
Chen, J. J., Chen, P., Xiao, X. 2011. Related Party transactions as a source of earning management.
App. Fin. Econ. 21, 165-181. Chen, S., Wang, K., Li, X. 2012. Product market competition, ultimate controlling structure and
related party transactions. China J. Account. Res. 5, 293-306.
Cheung, Y.L., Jing, L., Lu, T., Rau, P. R., Stouraitis, A. 2009. Tunneling and propping up: An analysis of related party transactions by Chinese listed companies. Pacific-Basin Fin. J. 17,
372-393.
Child, J. 1994. Management in China during the age of reform. Cambridge, Cambridge University Press.
Claessens, S., Djankov, S., Lang, L. H. P. 2000. The separation of ownership and control in East
Asian corporations. J. Fin. Econ. 58, 81-112.
Davis, L. R., Ricchiute, D. N., Trompeter, G. 1993. Audit effort, audit fees, and the provision of nonaudit services to audit clients. Account. Rev. 68, 135-150.
DeFond, M. L., Wong, T. J., Li, S. 2000. The impact of improved auditor independence on audit
market concentration in China. J. Account. Econ. 28, 269-305. DeFond, M. L., Zhang, J. 2014. A review of archival auditing research. J. Account. Econ. 58, 275-
326.
Dinç, I. S. 2005. Politicians and banks: Political influences on government-owned banks in emerging markets. J. Fin. Econ. 77, 453–479.
Dyck, A., Zingales, L. 2004. Private benefits of control: An international comparison. J. Fin. 59.
Faccio, M. 2006. Politically connected firms. Am. Econ. Rev. 96, 369-386.
Faccio, M. 2010. Differences between politically connected and nonconnected firms: A cross-country analysis. Fin. Mgt. Autumn, 905-927.
Faccio, M., Masulis, R. W., Mcconnell, J. J. 2006. Political connections and corporate bailouts. J. Fin.
61. 2597-2635. Fan, J. P. H., Wong, T. J. 2005. Do external auditors perform a corporate governance role in emerging
market? Evidence from East Asia. J. Account. Res. 43, 35-72.
Financial Conduct Authority. 2015. Asia Resource Minerals plc (formerly Bumi plc) fined £4.65m for
breaching the Listing Rules. Financial Conduct Authority. Available: https://www.fca.org.uk/news/asia-resource-minerals-plc-formerly-bumi-plc-fined-4-65m-for-
breaching-the-listing-rules# [Accessed 19 October 2015].
Fisman, R. 2001. Estimating the value of political connections. Am. Econ. Rev. 91, 1095-1102. Frye, T., Shleifer, A. 1997. The invisible hand and the grabbing hand. Am. Econ. Rev. 87, 354-358.
Fukuoka, Y. 2013. Indonesia's 'democratic transition' revisited: A clientelist model of political
transition. Democratization 20, 991-1013. Goldman, E., Rocholl, J., So, J. 2009. Do politically connected boards affect firm value? Rev. Fin. St.
22, 2331-2360.
Gow, I. D., Ormazabal, G., Taylor, D. J. 2010. Correcting for cross‐sectional and time‐series
dependence in accounting research. Account. Rev. 85, 483-512. Guedhami, O., Pittman, J. A., Saffar, W. 2009. Auditor choice in privatized firms: Empirical evidence
on the role of state and foreign owners. J. Account. Econ. 48, 151-171.
Guedhami, O., Pittman, J. A., Saffar, W. 2014. Auditor choice in politically connected firms. J. Account. Res. 52, 107-162.
Gul, F.A. 2006. Auditors’ response to political connections and cronyism in Malaysia. J. Account.
Res. 44(5), 931-963.
Habib, A., Haris, A., Jiang, H. 2016. Political connections and related party transactions: Evidence from Indonesia. Int. J. Account. Forthcoming.
He, X., Rui, O., Zheng, L., Zhu, H. 2014. Foreign ownership and auditor choice. J. Account. Public
Pol. 33, 401-418. Heckman, J. J. 1979. Sample-selection bias as a specification error. Econometrica 47, 153-62.
32
Hope, O.K., Kang, T., Thomas, W., Yoo, Y. K. 2008. Culture and auditor choice: A test of the
secrecy hypothesis. J. Account. Public Pol. 27, 357–373. Houston, J. F., Jiang, L., Lin, C., Ma, Y. 2014. Political connections and the cost of bank loans. J.
Account. Res. 52, 193-243.
Humphrey, C., Loft, A., Woods, M. 2009. The global audit profession and the international financial
architecture: Understanding regulatory relationships at a time of financial crisis. Account. Org. Society. 34, 810–825.
IASB 2009. Related Party Disclosures. International Accounting Standard 24. Available at:
http://www.iasplus.com/en/standards/ias/ias24. Jensen, M. C., Meckling, W. H. 1976. Theory of the firm: Managerial behavior, agency costs and
ownership structure. J. Fin. Econ. 3, 305–360.
Jian, M., Wong, T. J. 2010. Propping through related party transactions. Rev. Account. St. 15, 70-105. Jiang, G., Lee, C. M. C., Yue, H. 2010. Tunnelling through intercorporate loans: The China
experience. J. Fin. Econ. 98, 1-20.
Johnson, S., Mitton, T. 2003. Cronyism and capital controls: Evidence from Malaysia. J. Fin. Econ.
67, 351–382. Khanna, T., Palepu, K. 2000. Is group affiliation profitable in emerging markets? An analysis of
diversified Indian business groups. J. Fin. 55, 867-891.
Khwaja, A. I., Mian, A. 2005. Do lenders favor politically connected firms? Rent provision in an emerging financial market. Quart. J. Econ. 120, 1371-1411.
Kim, C.F., Zhang, L. 2016. Corporate political connections and tax aggressiveness. Contem. Account.
Res. 33, 78-114. Knechel, R.W., Salterio, S., Ballou, B., 2007. Auditing: Assurance and Risk. South-Western College
Pub.
Kothari, S. P., Leone, A. J., Wasley, C. E. 2005. Performance matched discretionary accrual
measures. J. Account. Econ. 39, 163-197. Kohlbeck, M., Mayhew, B. W. 2010. Valuation of firms that disclose related-party transactions. J.
Account. Public Pol. 29, 115–137.
Lennox, C.S., Francis, J.R., Wang, Z. 2012. Selection models in accounting research. Account. Rev. 87(2), 589–616
Leuz, C., Oberholzer-Gee, F. 2006. Political relationships, global financing, and corporate
transparency: Evidence from Indonesia. J. Fin. Econ. 81, 411-439.
Li, H., Meng, L., Wang, Q., Zhou, L.A. 2008. Political connections, financing and firm performance: Evidence from Chinese private firms. J. Dev. Econ. 87, 283–299.
Ma, L., Ma, S., Tian, G. 2013. Political connections, founder-managers, and their impact on tunneling
in China's listed firms. Pacific-Basin Fin. J. 24, 312-339. Mietzner, M. 2006. The politics of military reform in post-Suharto Indonesia: Elite conflict,
nationalism, and institutional resistance. Washington, D.C., East-West Center Washington.
Morck, R. K., Stangeland, D. A., Yeung, B. 2000. Inherited wealt, corporate control, and economic growth: The Canadian disease. In: Morck, R. K. (ed.) Concentrated Corporate Ownership.
Chicago: University of Chicago Press.
Newman, D. P., Patterson, E. R., Smith, J. R. 2005. The role of auditing in investor protection.
Account. Rev. 80, 289-313. Petersen, M. A. 2009. Estimating standard errors in finance panel data sets: Comparing approaches.
Rev. Fin. St. 22, 435-480.
Piotroski, J. D., Wong, T. J., Zhang, T. 2015. Political incentives to suppress negative information: Evidence from Chinese listed firms. J. Account. Res. 53, 405-459.
Qian, M., Pan, H., Yeung, B. 2011. Expropriation of minority interest shareholders in politically
connected firms. Working Paper - NUS Business School. Rosenbaum, P. R., Rubin, D. B. 1983. The central role of the propensity score in observational studies
for causal effects. Biometrika 70, 41-55.
Rosenbaum, P. R., Rubin, D. B. 1985. Constructing a control group using multivariate matched
sampling methods that incorporate the propensity score. Am. Stat. 39, 33-38. Ryngaert, M. D., Thomas, S. 2012. Not all related-party transactions (RPTs) are the same: ex ante
versus ex post RPTs. J. Account. Res. 50, 845–882.
33
Sebastian, L. C., Iisgindarsah 2013. Taking stock of military reform in Indonesia. In: Rüland, J.,
Manea, M.G., Born, H. (eds.) The Politics of Military Reform Expoeriences from Indonesia and Nigeria. Verlag Berlin Heidelberg Springer.
Thomas, W. B., Herrmann, D. R., Inoue, T. 2004. Earnings management through affliated
transactions. J. Int. Account. Res. 3, 1-25.
Tucker, J. W. 2010. Selection bias and econometric remedies in accounting and finance research. J. Account. Lit. 29, 31-57.
Utama, C. A., Utama, S. 2013. Corporate governance, size and disclosure of related party
transactions, and firm value: Indonesia evidence. Int. J. Disc. Gov. 1-25. Utama, C. A., Utama, S. 2014. Determinants of disclosure level of related party transactions in
Indonesia. Int. J. Disc. Gov. 11, 74-98.
Wang, Q., Wong, T. J., Xia, L. 2008. State ownership, the institutional environment, and auditor choice: Evidence from China J. Account. Econ. 46, 112– 134.
Watts, R. L., Zimmerman, J. L. 1983. Agency problems, auditing, and the theory of the firm: Some
evidence. J. Law. Econ. 26, 613-633.
Whisenant, S., Sankaraguruswamy, S., Raghunandan, K. 2003. Evidence on the joint determination of audit and non-audit fees. J. Account. Res. 41, 721-744.
Wulandari, E. R., Rahman, A. R. 2005. Political patronage, cross-holdings and corporate governance
in Indonesia. In: Gul, F. A., Tsui, J. S. L. (eds.) The governance of East Asian corporations. N.Y.: Palgrave Macmillan.
Ying, Q., Wang, L. 2013. Propping by controlling shareholders, wealth transfer and firm
performance: Evidence from Chinese listed companies. China J. Account. Res. 6, 133-147.
Appendix I: Audit environment in Indonesia
The Indonesian Institute of Accountants9 has responsibility for establishing accounting
standards, auditing standards, and ethics codes for accountants. The International aid agencies
such as the World Bank and IMF and foreign investors recommend that the government both
enforce tougher disclosure rules and harmonize accounting practices in Indonesia. In
response to those requests, following the enactment of the Sarbanes-Oxley Act in July 2002,
the Indonesian government issued the Minister of Finance Regulation Number
423/KMK.06/2002 dated 30 September 2002 regarding Public Accountant Services10
.
Basioudis and Fifi (2004) claim that accounting standards and corporate governance
regulations in Indonesia have been influenced significantly by the developed western
countries and, sometimes, go even further than those of western regulations. They point to
9 Ikatan Akuntan Indonesia (IAI - the Indonesian Institute of Accountants) is the organization officially
recognized by the government. IAI was established on 23 December 1957. 10
The amendment of the previous Minister of Finance Regulation number 43/KMK.017/1997. This regulation was further revised by the issuance of the Minister of Finance Regulation number
17/PMK.01/2008, dated 05 February 2008, and was upgraded to the law by the enactment of Law number
5, year 2011, regarding Public accountants on 03 May 2011.
34
auditor rotation as stipulated in Regulation number 423/KMK.06/2002, in which audit firms
are not allowed to audit an entity for five consecutive years, and audit partners can audit an
entity for only three consecutive years11
. Further, they argue that the requirements for audit
rotation in Indonesia are more burdensome compared to those of other jurisdictions in the
world, as even the Sarbanes-Oxley Act did not require rotation of audit firms at all, or
rotation of audit partners until after five years.
All financial statements were to be prepared in accordance with the Indonesian
accounting standards which are mostly derived from the International and US accounting
standards. Further, financial statements were to be audited according to the Indonesian
auditing standards, which are significantly similar to the generally accepted auditing
standards in the US. In providing an opinion on the fairness of financial statements, auditors
refer primarily to the Indonesian Accounting Standards, followed by relevant domestic rules
and regulations, and International Accounting Standards for matters that are not covered by
the abovementioned regulations ((Basioudis and Fifi, 2004).
Wulandari and Rahman (2005) point out that even though the Indonesian accounting
standards are substantially similar to the International Accounting Standards, some
Indonesian companies still have low quality financial reports, owing to the lack of effective
and efficient enforcement mechanisms. Further, they argue that auditors might have been
responsible for playing a role in the Asian financial crisis, as many of listed companies were
granted unqualified opinions prior to the crisis, but went bankrupt during the financial crisis.
They claim that auditors had little understanding and awareness of accounting or auditing
standards. In addition, auditors’ opinions were sometimes based on the work of other
professionals, such as appraisers, when they took for granted the value of certain assets
belonging to the auditee (Wulandari and Rahman, 2005), and auditors rely heavily on 11
The rotation of auditors was revised by the issuance of the Minister of Finance Regulation number
17/PMK.01/2008 by which rotation of audit firms is now required after six consecutive years (one year
longer than the previous regulation).
35
management representations without conducting further investigation to gather corroborative
evidence in supporting their audit opinions (Rahman, 2000 as cited in Wulandari and
Rahman, 2005).
Empirical evidence on whether audit quality is valuable in Indonesia is not available.
We used our data to provide some evidence on this important issue. Since extant research has
established that financial reporting quality, as proxied by less earnings management, is better
for firms audited by Big 4 and industry specialist audit firms, we regress abnormal accruals
on auditor identity and other control variables. Untabulated results show the coefficient on
AUDITOR, is significantly negative (coefficient -0.023, t-stat -2.41), suggesting that Big 4
auditors in Indonesia constrain earnings management. This result may provide a perspective
on connected firms’ reluctance to appoint Big 4 audit firms, i.e., to conceal their tunneling
and rent-seeking activities. We also regress audit opinion (OPINION) (a dummy variable
coded 1 if clients receive a qualified audit opinion, and 0 otherwise) on the AUDITOR
indicator variable and other relevant determinants of OPINION. Untabulated results show the
coefficient on AUDITOR is negative and significant (coefficient -0.42, z-statistic -1.85,
p<0.10). However, this effect is pronounced for the PCON=1 group only (coefficient -0.76,
z-statistic -1.75, p<0.10). This suggests that connected firms tend to get more qualified
opinions compared to their non-connected counterparts and this, again, might explain
connected firms’ preference for non-Big 4 auditors.
36
Appendix II: Development of RPT regulations in Indonesia
Bapepam-LK enacted Regulation Number VIII.G.7, which requires extensive disclosure on
RPTs in the notes to financial statements. Regulation Number VIII.G.7 under the Decree of
the Head of the Indonesian Capital Market Supervisory Agency number KEP-06/PM/2000
dated 13 March 2000 requires issuers and listed firms to disclose the breakdown and the total
amount of RPTs under each category of assets, liabilities, sales, and purchases (expenses),
along with their percentage to total assets, total liabilities, total sales, and total purchases
(expenses); nature, types and elements of RPTs; pricing policies and terms of RPTs and a
statement that RPTs are similar to those undertaken with third parties; reasons and basis for
the recognition of doubtful accounts for RPT receivables. If the amount of a RPT for each
category exceeds IDR 1,000,000,000 (US$ 101,96812
) or if non-core business RPTs are
conducted, separate disclosures and explanations are needed. From 25 June 2012, a new RPT
regulation was enacted, i.e., Regulation Number VIII.G.7, under a Decree of the Head of the
Indonesian Capital Market Supervisory Agency number KEP-347/BL/2012, in which the 12
(1US$=IDR 9,807 in 2013 (DataStream)
37
disclosure requirements are quite similar, but RPTs are now classified based on three
categories of parties conducting the RPT, namely, family members, related parties, and
government affiliated.
In addition, Regulation Number IX.E.1 under the Decree of the Head of the
Indonesian Capital Market Supervisory Agency, number KEP-412/BL/2009, dated 25
November 2009, requires issuers and listed firms to announce to the public any RPT, and
report the evidence of that announcement to the Indonesian Capital Market Supervisory
Agency no later than two working days after the transactions are undertaken, except for
RPTs having a value less than 0.5% of the firm’s paid capital and less than IDR
5,000,000,000 (US$509,840), in which case no public announcement is required, but such an
RPT shall be reported to the Indonesian Capital Market Supervisory Agency. In addition, if
RPTs are identified as conflict of interest transactions13
, they need to get approval from the
independent shareholders, except for RPTs having a low value as specified above.
Recent RPT cases in the United Kingdom involved an Indonesian listed firm. Asia
Resource Mineral Plc (formerly Bumi Plc) suffered fines in the amount of £4,651,200 from
the Financial Conduct Authority because its subsidiary, namely, PT Berau Coal Energy Tbk
(listed in Indonesian Stock Exchange) conducted an RPT which breached UKLA Listing
Rules (Authority, 2015). Those two companies are controlled by Bakrie Group, and are
classified as politically connected firms in our study.
13
According to Bapepam-LK rule IX.E.I, conflict of interest takes place in the transaction when the
economic interest of the firm differs from the personal interest of directors, commissioners, or controlling
shareholders, such that the transaction might ruin the firm.
38
Appendix – III: Variable definitions
Variables Definitions
AUDITOR Dummy variable, 1 for the firm audited by Big 4 auditors, 0 for otherwise.
PCON Dummy variable, 1 for politically connected firms, 0 for otherwise.
GCON Dummy variable, 1 for government connected firms, 0 for otherwise. MCON Dummy variable, 1 for military connected firms, 0 for otherwise.
SCON Dummy variable, 1 for Suharto connected firms, 0 for otherwise.
RPT_RATIO Gross RPT divided by total assets of the firm.
OPRPT_RATIO Gross operating RPTs divided by total assets of the firm. LOAN_RATIO Gross RPTLOAN divided by total assets of the firm.
OWNCON Total percentage of shares owned by the five largest shareholders
FOWN Total percentage of shares owned by foreign institutional investors. SIZE Natural logarithm of total assets.
LEV Total long term debt divided by total assets.
GROWTH Market value of equity divided by book value of equity. INV The ratio of inventory to total assets
SEGMENT Natural logarithm of number of business segments
FINANCE A dummy variable that takes the value of one if the sum of new long-term debt
plus new equity exceeds 20% of total assets ROA Return on assets (earnings before extraordinary items plus discontinued operations
for the preceding year divided by total assets for the same year).
AC Dummy variable, 1 for the firm having audit committee, 0 for otherwise. IC Dummy variable, 1 for the firm having independent commissioners, 0 for
otherwise.
ACC Total accruals derived from earnings before extraordinary items and discontinued operations minus operating cash flows.
|DAC| Absolute discretionary accruals calculated with the Modified Jones model (1995).
To estimate DAC we use the cross-sectional modified Jones model, controlling for
firm performance (Kotari et al., 2005). We estimate the following model for all firms in the same industry (using economic sector code):
39
)2.....()()/(]/)[()/1(/
3
1211101
t
tt tttttt
ROA TAPPE TADEBTORS SALES TA TAACC
The coefficient estimates from Equation (2) are used to predict non-discretionary
component of total accruals (NDAC) for our sample firms. Thus, discretionary accruals is the residual from equation (2), i.e. DAC=ACC-NDAC.
ΔSALES Change in sales from year t-1 to year t.
∆DEBTORS Change in accounts receivable from year t-1 to year t. PPE Gross property, plant, and equipment.
ROA Return on assets (earnings before extraordinary items plus discontinued operations
for the preceding year divided by total assets for the same year).
TABLE 1: Sample selection procedure and descriptive statistics
PANEL A: Sample selection procedure
Sample Selection Process Observations Number of non-financial firm-year observations from 2004 to 2013 3,149
Less: Firm year observations pertaining to 2004-2006 sample period dropped because of
too few available annual reports
( 772)
Less: Firm year observations with unavailable audit reports during 2007-2013 period ( 481)
Less: Firm year observations with zero RPT values ( 113)
Less: Number of firm-year observations with negative book value (distress firms) ( 85)
Number of firm-year observations with complete non zero RPT data 1,698
Less: missing data of other control variables ( 269)
Number of firm-year observations for the baseline regression 1,429
PANEL B: Descriptive statistics
Variables Mean SD 25% Median 75%
AUDITOR 0.43 0.50 0.00 0.00 1.00
PCON 0.39 0.49 0.00 0.00 1.00
GCON 0.24 0.43 0.00 0.00 0.00
MCON 0.12 0.32 0.00 0.00 0.00
SCON 0.03 0.19 0.00 0.00 0.00
RPT_RATIO 0.50 0.88 0.04 0.21 0.66
OPRPT_RATIO 0.28 0.57 0.00 0.03 0.27
LOAN_RATIO 0.11 0.29 0.00 0.01 0.08
OWNCON 0.71 0.19 0.59 0.74 0.85
FOWN 0.28 0.30 0.00 0.15 0.52
SIZE 19.00 1.67 17.89 18.99 20.15
LEV 0.13 0.15 0.00 0.08 0.21
GROWTH 3.17 5.52 0.94 1.66 3.28 INV 0.16 0.16 0.02 0.12 0.24
SEGMENT 0.90 0.53 0.69 1.1 1.39
40
FINANCE 0.14 0.34 0.00 0.00 0.00
ROA 0.07 0.12 0.01 0.05 0.11
AC 0.81 0.39 1.00 1.00 1.00
IC 0.92 0.27 1.00 1.00 1.00
|DAC| 0.10 0.10 0.03 0.07 0.13
PANEL C: Industry distributions
Sector Code Economic Sector Description Observations % of
observations 1000 Materials 340 23.79
2000 Consumer Discretionary 287 20.08
3000 Consumer Staples 222 15.54
3500 Health Care 42 2.94
4000 Energy 41 2.87
5000 Real Estate Management & Development 168 11.76
6000 Industrials 248 17.35
8000 Information Technology 40 2.80
8600 Telecommunication Service 41 2.87
1,429 100.00
Panel D: Univariate test Variables PCON=1 [n=562] PCON=0 [n=867] t-test of difference in mean
AUDITOR 0.53 0.37 5.53***
RPT_RATIO 0.47 0.51 -1.02
OPRPT_RATIO 0.24 0.30 -1.72*
LOAN_RATIO 0.11 0.12 -0.42
OWNCON 0.51 0.50 0.68
FOWN 0.23 0.31 -4.55***
SIZE 20.03 18.32 21.95***
LEV 0.16 0.11 5.23***
GROWTH 3.73 2.81 3.08***
INV 0.14 0.18 -4.32***
SEGMENT 3.06 2.55 7.29***
FINANCE 0.14 0.13 0.75
ROA 0.08 0.06 1.88*
AC 0.89 0.76 6.00***
IC 0.91 0.92 -0.57
|DAC| 0.09 0.10 -0.99
Note: Variable definitions are in Appendix III. ***, ** are significant at p<0.01 and <0.05 respectively (two-tailed test)
41
TABLE 2: Correlation analysis
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17)
AUDITOR (1) 1.00
PCON (2) 0.15 1.00
RPT_RATIO (3) -0.11 -0.03 1.00
LOAN_RATIO (4) 0.14 -0.05 0.33 1.00
OPRPT_RATIO (5) -0.17 -0.01 0.59 -0.11 1.00
OWNCON(6) -0.04 0.11 0.49 0.08 0.12 1.00
FOWN (7) 0.18 -0.15 -0.04 0.17 -0.02 -0.06 1.00
SIZE (8) 0.16 -0.12 0.04 0.14 0.00 -0.03 0.30 1.00
LEV (9) 0.40 0.50 -0.02 -0.01 -0.07 0.11 -0.21 -0.09 1.00
GROWTH (10) 0.00 0.13 -0.05 -0.16 0.15 -0.08 -0.14 -0.01 0.27 1.00
INV(11) 0.10 0.08 -0.02 -0.01 0.05 0.07 0.02 0.02 0.01 0.00 1.00
SEGMENT(12) 0.00 -0.12 -0.09 0.11 -0.05 -0.10 0.06 -0.02 -0.09 -0.30 -0.06 1.00
FINANCE(13) 0.05 0.14 -0.04 0.00 -0.05 -0.02 -0.17 -0.12 0.20 0.02 -0.06 -0.03 1.00
ROA (14) -0.02 0.02 0.06 -0.02 0.08 0.04 0.00 -0.01 0.04 0.17 0.08 -0.13 -0.01 1.00
AC (15) 0.22 0.04 -0.05 0.08 -0.16 0.13 0.09 0.01 0.17 -0.28 0.20 0.05 -0.02 0.24 1.00
IC (16) 0.13 0.16 -0.06 -0.06 -0.12 0.08 -0.09 -0.06 0.23 0.06 -0.01 -0.07 0.03 0.04 0.08 1.00
|DAC| (17) 0.05 -0.01 0.06 0.03 -0.01 0.06 -0.05 -0.05 0.11 0.07 -0.05 -0.04 0.02 0.05 -0.01 0.20 1.00
Note: Italicized and bold-faced correlations are significant at p<0.01. Variable definitions are in Appendix III. The correlation is based on a full sample of 1,429 firm-year observations.
42
TABLE 3: Political connections, RPTs, and auditor choice
)1.......(..............................||13121110987
6543210
YearIndustryDACICACROAFINANCESEGMENTINV
GROWTHLEVSIZEFOWNOWNCONPCONAUDITOR
PANEL A: Political connections, RPTs, and auditor choice Only PCON=1 Only PCON=1
RPT_RATIO RPT_RATIO RPT_RATIO
Expected
Sign
(1) (2) (3) (4) (5)
Baseline Baseline PCON GCON=1 MCON=1
Variables Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
PCON - -0.363** - 0.007 - -
[-2.22] [0.04]
GCON - - -0.877*** - - -
[-4.36]
MCON - - -0.193 - - -
[-0.88]
SCON + - 2.358*** - - -
[4.05]
RPT_RATIO - - - -0.271** -2.417*** -1.16
[-2.21] [-4.96] [-0.78]
PCON*RPT_RATIO - - - -1.163*** - -
[-4.12]
OWNCON + 2.150*** 2.229*** 2.620*** 4.340*** 5.609***
[6.85] [7.03] [8.12] [5.30] [3.63]
FOWN + 1.866*** 1.857*** 1.936*** 4.229*** 5.163***
[7.59] [7.36] [7.32] [6.01] [5.11]
SIZE + 0.805*** 0.840*** 0.906*** 0.731*** 1.427***
[13.02] [12.81] [12.94] [3.37] [3.76] LEV - -1.283** -1.429** -1.606** -3.125 -6.753***
[-2.03] [-2.22] [-2.36] [-1.63] [-2.63]
GROWTH + 0.047** 0.052*** 0.047** 0.049* 0.116
[2.47] [2.74] [2.46] [1.81] [0.93]
INV + 1.225*** 1.441*** 1.062** 3.881** -0.632
[2.60] [3.01] [2.09] [2.24] [-0.33]
SEGMENT + 0.049 0.072 0.015 -0.022 0.005
[0.36] [0.51] [0.11] [-0.05] [0.03]
FINANCE ? -0.341 -0.312 -0.387 0.066 0.289
[-1.49] [-1.33] [-1.63] [0.13] [0.37]
ROA + 1.984** 1.803** 1.512* 11.362*** -1.338 [2.44] [2.15] [1.76] [3.36] [-0.40]
AC + 0.726*** 0.847*** 0.633*** -0.577 4.304***
[3.69] [4.18] [3.28] [-0.78] [4.38]
IC + 0.066 0.262 0.184 -0.101 -0.524
[0.26] [1.11] [0.72] [-0.18] [-0.45]
|DAC| ? 0.123 0.090 -0.035 -1.352 7.542*
[0.15] [0.11] [-0.04] [-0.48] [1.79]
Constant -16.710*** -17.614*** -18.435*** -14.519*** -31.578***
[-12.41] [-12.62] [-12.77] [-3.36] [-4.10]
Industry YES YES YES YES YES
Year YES YES YES YES YES
Pseudo R2 0.29 0.31 0.32 0.46 0.48
Observations 1,429 1,429 1,429 342 155
***, **, and * represent statistical significance at the 1%, 5%, and 10% levels respectively (two-tailed test). Variable
definitions are in Appendix III.
43
PANEL B: Categories of RPTs, political connections, and auditor choice
Operating RPTs RPT loan and guarantees Operating RPTs and RPT loans (1) (2) (3) (4) (5) (6) (7) (8) (9)
Predicted
sign
Full sample Only
PCON=1 sample
Only
PCON=1 sample
Full sample Only
PCON=1 sample
Only
PCON=1 sample
Full sample Only PCON=1
sample Only PCON=1
sample
GCON=1 MCON=1 GCON=1 MCON=1 GCON=1 MCON=1
Variables Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
PCON - -0.22 - - -0.338* - - -0.125 - -
[-1.19] [-1.92] [-0.66]
OPRPT_RATIO + 0.47*** -0.185 0.348 - - - 0.430*** -0.333 0.380
[3.40] [-0.75] [0.55] [3.08] [-1.26] [0.59]
LOAN_RATIO - - - -1.180** -3.160*** 0.882 -1.058** -3.719*** 0.880
[-2.42] [-3.49] [0.84] [-2.19] [-4.07] [0.83]
PCON*OPRPT_RATIO + -0.517 - - - - - -0.56 - -
[-1.16] [-1.29]
PCON*LOAN_RATIO - - - - -1.373* - - -1.550** - -
[-1.73]
[-1.96]
Other control variables YES YES YES YES YES YES YES YES YES
Industry YES YES YES YES YES YES YES YES YES
Year YES YES YES YES YES YES YES YES YES
Pseudo R2 0.29 0.37 0.47 0.31 0.42 0.47 0.31 0.42 0.47
Observations 1,429 342 155 1,429 342 155 1,429 342 155
44
TABLE 4: Endogeneity tests
Heckman (1979) self-selection tests
..(5)……………………………… +YearFE + FEIndustry ROASEGMENTGROWTH +LEV +SIZE +FOWN +OWNCON +HQ%PCON_IND =PCON
98
7654321
)1.......(..............................|| 141312111098
76543210
YearIndustryIMRDACICACROAFINANCESEGMENT
INVGROWTHLEVSIZEFOWNOWNCONPCONAUDITOR
(1) (2) (3) (4) (5)
1st stage
probit
model
DV=PCON
Expected
sign for the
2nd stage
variables
PCON=1 &
GCON=1
PCON=1 &
MCON=1
Variables Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
PCON - - -0.029 -0.199 - -
[-0.15] [-1.07]
RPT_RATIO - - -0.310** - - -
[-2.38]
OPRPT_RATIO - + - 0.390*** -0.338 0.493
[2.71] [-1.29] [0.81]
LOAN_RATIO - - - -1.120*** -3.266*** 1.337
[-2.75] [-3.80] [1.29]
PCON*RPT_RATIO - - -1.146*** - - -
[-3.97]
PCON*OPRPT_RATIO - + - -0.535 - -
[-1.20]
PCON*LOAN_RATIO - - - -1.510* - -
[-1.89]
HQ 0.219** NA - - - -
[2.16]
%PCON_IND 1.311* NA - - - -
[1.76]
OWNCON -0.092 + 2.715*** 2.436*** 3.910*** 4.808***
[-0.43] [8.24] [7.21] [4.35] [2.98]
FOWN -0.328** + 2.137*** 2.145*** 3.944*** 4.905***
[-2.37] [5.04] [5.37] [5.69] [3.95]
SIZE 0.478*** + 0.866** 0.576 0.664*** 0.792
[14.49] [2.04] [1.41] [2.59] [1.34]
LEV -0.310 -1.654** -0.954 -1.647 -7.587**
[-1.00] [-2.25] [-1.51] [-0.86] [-2.48]
GROWTH 0.032*** + 0.042 0.031 0.037 0.066
[3.00] [1.37] [1.04] [1.11] [0.46]
INV - + 1.076** 1.075** 3.384* -2.623
[2.09] [2.03] [1.87] [-1.41]
SEGMENT 0.172*** + -0.069 -0.181 -0.269 -0.115
[2.22] [-0.34] [-0.89] [-0.72] [-0.21]
FINANCE - ? -0.382 -0.304 0.277 0.137
[-1.60] [-1.38] [0.57] [0.17]
ROA -0.515 + 1.691 1.683* 11.204*** -0.344
[-1.35] [1.62] [1.92] [3.44] [-0.09]
AC - + 0.665*** 0.694*** -0.139 4.633***
[3.39] [3.33] [-0.20] [4.54]
IC - + 0.188 0.023 -0.728 -1.407
[0.73] [0.08] [-1.24] [-1.22]
45
|DAC| - ? -0.043 0.037 -0.723 8.041*
[-0.05] [0.05] [-0.26] [1.68]
IMR - ? -0.344 -1.514 0.886 -4.662
[-0.16] [-0.72] [1.02] [-1.48]
Constant -9.742*** -17.425* -17.63*** -13.587** -15.837
[-11.90] [-1.75] [-8.75] [-2.34] [-1.15]
Industry YES YES YES YES YES
Year YES YES YES YES YES
Pseudo R2 0.26 0.31 0.30 0.44 0.44
Observations 1,429 1,429 1,429 342 155
46
Table 5: Propensity-matched technique
PANEL A: Propensity-matched variables
NN method Kernel method
Variable Treated Control t-stat p-value Treated Control t-stat p-value
OWNCON 0.51 0.52 -0.75 0.451 0.51 0.52 -0.61 0.54
FOWN 0.23 0.29 -3.15 0.002 0.23 0.28 -2.49 0.013
SIZE 20.04 19.97 0.83 0.407 19.91 19.90 0.15 0.883
LEV 0.16 0.17 -1.43 0.152 0.15 0.16 -1.17 0.243
INV 0.14 0.14 -0.16 0.874 0.15 0.14 0.07 0.941
SEGMENT 0.99 0.93 1.91 0.057 0.98 0.93 1.66 0.097
FINANCE 0.14 0.13 0.83 0.408 0.14 0.13 0.46 0.646
ROA 0.08 0.09 -1.62 0.115 0.08 0.09 -1.62 0.115
AC 0.89 0.89 0.08 0.939 0.89 0.89 -0.05 0.963
|DAC| 0.09 0.10 -0.74 0.458 0.09 0.10 -0.34 0.737
47
PANEL B: Regression results
(1) (2) (3) (4) (5) (6) (7) (8)
NN NN NN NN Kernel Kernel Kernel Kernel
Variables Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
Coefficient
[z-stat]
PCON - -0.613** - -0.401* -0.569*** -0.765*** -0.360* -0.679***
[-3.54] [-1.91] [-2.79] [-4.74] [-1.77] [-3.56]
GCON - - -0.724*** - - - -1.203*** - -
[-3.62] [-6.22]
MCON - - -0.967*** - - - -0.647*** - -
[-4.20] [-2.83]
SCON + - 2.696*** - - - 2.247*** - -
[3.85] [3.62]
RPT_RATIO - - - -0.664*** - - - -0.554** -
[-2.81] [-2.48]
OPRPT_RATIO + - - - -0.237 - - - -0.225
[-1.00] [-1.06]
LOAN_RATIO - - - - -1.667*** - - - -1.276**
[-2.92] [-2.38]
PCON*RPT_RATIO - - - -0.723** - - - -1.153*** -
[-2.11] [-3.36]
PCON*OPRPT_RATIO + - - - 0.016 - - - -0.012
[0.06] [-0.04]
PCON*LOAN_RATIO - - - - -2.047* - - - -1.590*
[-1.80] [-1.76]
OWNCON + 2.980*** 3.332*** 3.753*** 3.429*** 2.192*** 2.331*** 2.872*** 2.508***
[6.96] [7.51] [8.20] [7.60] [5.56] [5.73] [6.83] [6.06]
FOWN + 4.040*** 4.356*** 4.230*** 4.181*** 2.141*** 2.124*** 2.221*** 2.171***
[12.35] [12.69] [12.50] [12.38] [6.69] [6.44] [6.56] [6.58]
SIZE + 0.855*** 0.840*** 0.973*** 0.882*** 0.748*** 0.783*** 0.906*** 0.793***
[9.97] [9.50] [10.10] [9.58] [8.90] [8.81] [9.47] [8.84]
LEV - -1.317** -1.106 -1.355* -0.900 -0.576 -0.638 -0.824 -0.497
[-1.97] [-1.61] [-1.95] [-1.30] [-0.89] [-0.96] [-1.22] [-0.74]
GROWTH + -0.029* -0.032* -0.043** -0.040** -0.001 0.003 -0.012 -0.006
[-1.72] [-1.91] [-2.43] [-2.30] [-0.04] [0.19] [-0.66] [-0.35]
48
INV + 0.326 0.889 0.771 0.910 1.044* 1.452** 0.994 1.170*
[0.56] [1.44] [1.25] [1.44] [1.69] [2.29] [1.51] [1.78]
SEGMENT + -0.386** -0.172 -0.365** -0.413** 0.145 0.206 0.158 0.072
[-2.33] [-0.99] [-2.13] [-2.42] [0.89] [1.22] [0.92] [0.43]
FINANCE ? -0.261 -0.193 -0.372 -0.335 -0.326 -0.295 -0.444* -0.321
[-1.07] [-0.77] [-1.43] [-1.29] [-1.34] [-1.17] [-1.72] [-1.26]
ROA + 5.829*** 5.699*** 5.266*** 5.446*** 4.808*** 4.667*** 4.602*** 4.312***
[6.34] [6.24] [5.46] [5.67] [5.22] [4.93] [4.81] [4.55]
AC + 1.062*** 1.161*** 0.877*** 0.970*** 0.927*** 1.129*** 0.765*** 0.875***
[3.87] [4.06] [3.06] [3.44] [3.27] [3.80] [2.61] [3.01]
IC + 0.357 0.745** 0.514 0.379 -0.142 0.067 0.016 -0.146
[1.09] [2.12] [1.45] [1.09] [-0.44] [0.20] [0.05] [-0.43]
|DAC| ? 0.717 0.049 0.711 0.689 -1.201 -1.391 -1.407 -1.367
[0.78] [0.05] [0.75] [0.71] [-1.26] [-1.39] [-1.42] [-1.39]
Constant -17.826*** -18.295*** -20.159*** -18.379*** -15.018*** -16.078*** -17.872*** -15.473***
[-9.26] [-9.15] [-9.54] [-8.99] [-8.36] [-8.44] [-8.87] [-8.11]
Industry YES YES YES YES YES YES YES YES
Year YES YES YES YES YES YES YES YES
Pseudo R2 0.43 0.45 0.45 0.44 0.29 0.33 0.34 0.31
Observations 1,150 1,150 1,144 1,144 1,398 1,398 1,391 1,391