Journal of Business Research 91 (2018) 94–107

Contents lists available at ScienceDirect

Journal of Business Research

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

The dark side of rent-seeking: The impact of rent-seeking on earnings T

management☆
⁎
Baohua Liua, Yan Linb, , Kam C. Chanc, Hung-Gay Fungd
a
School of Economics and Management, Southwest Jiaotong University, Chengdu, China
b
School of Accounting, Yunnan University of Finance and Economics, Yunnan, China
c
Gordon Ford College of Business, Western Kentucky University, Bowling Green, KY 42101, United States
d
College of Business Administration, University of Missouri-St. Louis, St. Louis, MO 63121, United States

A R T I C LE I N FO A B S T R A C T

JEL classification: We use Chinese firm-level data from 2009 to 2014 to examine the impact of rent-seeking on a firm's earnings
G18 management, an important indicator of quality in accounting information. We show that rent-seeking strongly
G31 promotes more earnings management, hiding firm-specific information from the market. When the Chinese
Keywords: government initiates an anti-corruption campaign that depresses rent-seeking activities, the significant relation
Firm-level rent-seeking between rent-seeking and earnings management disappears, confirming the predictive link of rent-seeking and
Earnings management earnings management. Our study clearly identifies rent-seeking as a key driver of earnings management. The
SOEs positive effect of rent-seeking on earnings management is stronger at firms that are not state owned (non-SOEs)
Politically connected firms
than at state-owned enterprises and at non-SOEs without political connections than at those with political
connections.

1. Introduction investment, employment, and capital structure.2 Few studies examine

Firms engage in rent-seeking activities to garner significant eco-
nomic benefits primarily from their governments. Rent-seeking by firms
involves the practice of manipulating public policy or certain economic
conditions to gain favors as a strategy for increasing earnings. Some of
these economic benefits include better government services, subsidies,
and lower tax rates (Cai, Fang, & Xu, 2011), increased revenue (Wang &
You, 2012), approval of initial public offerings (Liu, Tang, & Tan, 2013;
Piotroski & Zhang, 2014), a lower threat of extraction by the govern-
ment (Kusnadi, Yang, & Zhou, 2015), the ability to avoid spillover of
guilt by association (Jia & Zhang, 2016), and greater ability to have
bank credit and loan applications approved (Chen, Liu, & Su, 2013; Fan,
Rui, & Zhao, 2008; Khwaja & Mian, 2005). The literature generally
suggests that rent-seeking activities are bad for an economy because it
creates misallocation of resources, but these activities are good for an
individual firm.1 In contrast, several studies suggest that the economic
benefits from rent-seeking come with costs to a firm in the form of
possible adverse impacts on firm productivity, revenue, corporate
the impact of rent-seeking, if any, on accounting information quality.
This study uses information based on entertainment expenses and
travel costs (ETC) at Chinese firms to determine the extent of rent-
seeking activities to investigate the adverse effect of rent-seeking on
earnings management. We argue that rent-seeking prompts more
earnings management, which in turn implies a poor quality of ac-
counting information revealed to the market. Accounting information is
critical to stakeholders for monitoring and evaluation of firm perfor-
mance. Any impact of a firm's rent-seeking activities on accounting
information quality is important for stakeholders in making decisions.
For instance, a high-quality accounting report can reduce information
asymmetry for better investor protection (La Porta, Lopez-de-Silance,
Shleifer, & Vishny, 1998), lower the cost of capital for firms (Lambert,
Leuz, & Verrecchia, 2007), and reduce non-diversifiable information
risk (Easley & O'Hare, 2004).
Drawing from the theory of self-preservation (Karni & Schmeidler,
1986), we deduce that that rational executives engaging in rent-seeking
are likely to take appropriate actions in the event of bad outcomes for

☆
We acknowledge the financial support of the National Natural Science Foundation of China (Grant Nos. 71702153 and 71762030), the Humanity and Social Science Foundation for
Youth Scholar of Ministry of Education of China (Grant No. 17YJC630081).
⁎
Corresponding author.
E-mail addresses: liubh@swjtu.edu.cn (B. Liu), charrieooo@126.com (Y. Lin), Johnny.chan@wku.edu (K.C. Chan), fungh@msx.umsl.edu (H.-G. Fung).
1
See, for example, Aidt, Dutta, and Sena (2008), Dong and Torgler (2010), Murphy et al. (1993), Tang et al. (2016), and Zhang et al. (2017).
2
See, for example, Asiedu and Freeman (2009), De Rosa, Gooroochurn, and Görg (2010), Fan, Titman, and Twite (2012), Fisman and Svensson (2007), Nguyen and Van Dijk (2012),
and Smith (2016).

https://doi.org/10.1016/j.jbusres.2018.05.037
Received 25 October 2017; Received in revised form 25 May 2018; Accepted 26 May 2018
0148-2963/ © 2018 Elsevier Inc. All rights reserved.
B. Liu et al.
self-protection. Executives would try to make their firms opaque to
reduce public awareness. Prior studies have shown earnings manage-
ment make firms opaque to the market (Caprio, Faccio, & McConnell,
2013; Durnev & Fauver, 2011; Smith, 2016). Thus, as an implication of
executives' self-preservation, rent-seeking promotes more earnings
management for two reasons. First, rent-seeking firms (i.e., firms that
engage in rent-seeking activities) try to disguise rent-seeking benefits
because if they are caught engaging in rent-seeking (regardless of
whether bribes are involved), especially in China, they may suffer
serious consequence. As a result, executives of these firms prefer to
maintain a low profile to minimize public scrutiny. They use earnings
management for self-protection to minimize information disclosure in
order to circumvent public awareness. As a result, accounting in-
formation quality is worse at a rent-seeking firm than a non-rent-
seeking firm.
Second, rent-seeking firms deliberately become more opaque to
deter government officials who attempt to seek personal benefits. When
government officials have no clear idea about a firm's performance, the
firm can shield itself from solicitations by them seeking personal benefit
(Caprio et al., 2013; Smith, 2016). Thus, in both explanations, earnings
management is a self-preservation realization of executives to take ac-
tions to mitigate possible bad outcomes of rent-seeking.
We conduct our analysis using firm-level data in China on A shares
(i.e., shares owned primarily by domestic investors) for firms listed on
the exchange from 2009 to 2014. China provides an ideal environment
for our examination for two reasons. First, rent-seeking activities are
common at Chinese firms. These firms actively solicit subsidies, tax
relief, bank loans, and other benefits from local and central govern-
ments using legal and illegal means.3 Second, irregularities in ac-
counting reporting in China are not new. One prominent issue in the
Chinese context is a poor accounting information environment.4 Hence,
it is particularly important to examine potential drivers behind earnings
management in China so that stakeholders can recognize any issues
early on. Our findings are valuable for domestic and international in-
vestors alike. Other emerging markets can also learn from our evidence
in the Chinese market.
We show that rent-seeking has a strong positive effect on earnings
management, which reduces the quality of a firm's accounting in-
formation. When an exogenous shock occurs—an anti-corruption
campaign initiated in December 2012 by China's President Xi Jinping
that supposedly reduced corruption at Chinese firms5—we do not find,
as expected, that rent-seeking had significant effects on earnings man-
agement, confirming a firm's rent-seeking rationale to hide information
on earnings management. In addition, the positive effect of rent-seeking
on earnings management is more pronounced among non-state-owned
enterprises (non-SOEs) and non-politically connected non-SOEs, sug-
gesting these firms have stronger incentives for engaging in rent-
seeking activities because of their inferior competitive position relative
to state-owned enterprises (SOEs) and politically connected firms.
Our paper makes three important contributions to the literature and
policy. First, we advance the literature on rent-seeking and earnings
management. We document that rent-seeking is a key driver of earnings
management (which reduces the quality of accounting information) at
the firm level. This result is new and has not been investigated before.
3
For instance, Kingsun, a Chinese manufacturer of light-emitting diodes, bribed gov-
ernment officials to obtain RMB 42.45 billion in subsidies in 2013 (http://finance.ifeng.
com/a/20150209/13490791_0.shtml, accessed August 4, 2016; in Chinese).
4
On April 16, 2014, Forbes magazine reported widespread accounting abuse at Chinese
listed firms (http://www.forbes.com/sites/ninaxiang/2014/04/16/accounting-fraud-is-
still-widespread-among-chinese-companies/#25373590723e/, accessed September 28,
2016).
5
According to China's Statistical Yearbook, from 2009 to 2012, the People's
Procuratorate (similar to the Congress in the United States) began an investigation into
serious graft, including embezzlement, bribery, and abuse of power. The average number
of people implicated was 2622; the number increased to 2871 in 2013 and peaked at 4040
in 2014.
95
Journal of Business Research 91 (2018) 94–107
Although rent-seeking activities may bring a firm economic benefits,
they also incur indirect, less visible, and long-term costs by making the
firm opaque to stakeholders. As stakeholders use poor accounting in-
formation in monitoring and evaluating firm performance, the cost of
firm capital is expected to increase because of greater information
asymmetry between the firm and the market. The adverse impact of
rent-seeking goes beyond the internal operations of a firm and affects
the entire market when resources resulting from poor-quality in-
formation are misallocated. We document that rent-seeking entails
costs to both the firm and society.
Second, we document that an anti-corruption campaign is able to
restrain rent-seeking activities. As one of leading emerging economies,
China implemented an anti-corruption campaign in 2013. We find a
weakened effect of rent-seeking on earnings management in 2013 and
after. This result suggests that an anti-corruption campaign, as a policy,
is useful in curbing rent-seeking activities. Thus, implementing anti-
corruption is a good public policy on improving accounting information
quality in a high-information-quality economy.
Third, our findings suggest that it is time to evaluate the appro-
priateness of China's accounting policy that allows excessive tax-de-
ductible expenses for rent-seeking activities. Chinese firms can claim
60% of the actual entertainment expenses, which is much more than
allowed by US rules, which permit firms to give business gifts costing
$25 and 50% of qualified entertainment expenses (Sun, 2016).
2. Literature review and hypothesis development
2.1. Rent-seeking and corruption
Rent-seeking involves economic activities that directly or indirectly
benefit government officials. For instance, a rent-seeking firm incurs
expenses to bring different government or private partners together for
a project that helps develop a certain area or district, in which the rent-
seeking firm has consequently earned favors from government officials
for future benefits. In this case, the government officials do not receive
direct benefits from the rent-seeking firm, hence, it is not considered a
bribe paid to government officials by the rent-seeking firm. Rent-
seeking is thus more general than corruption and involves direct per-
sonal benefits to the corrupt government officials in exchange for firm
benefits.6
Two conceptual frameworks can explain the pros and cons of rent-
seeking. As a form of rent-seeking activity, corruption can quicken the
pace of bureaucratic procedures to help rent-seeking firms achieve their
goals (Lui, 1985). It is possible that corruption allows the most efficient
firm to obtain scarce government resources (Beck & Maher, 1986). As
implied by the term “greasing the wheels,” rent-seeking accelerates
economic growth and benefits firms. For instance, rent-seeking firms
can attain higher sales growth (Wang & You, 2012), obtain more pro-
curement contracts from the government (Huang & Li, 2013), receive
more government subsidies (Li & Chi, 2015), and have more corporate
financing channels (Chen et al., 2013; Fan et al., 2008).
However, rent-seeking has associated costs. Economic benefits (e.g.,
government subsidies) given to rent-seeking firms mean fewer re-
sources for other firms. Competitors of the rent-seeking firm may suffer;
rent-seeking activities may stifle healthy competition. It can be also
argued that corruption can be the real driving force for government
officials to implement and maintain inefficient policies in order to en-
tice firms into rent-seeking activities (Aidt, 2003). Rent-seeking may
“squeeze out” the productive and innovative activities of a firm and
harms a country's economic development (Mauro, 1995; Murphy,
Shleifer, & Vishny, 1993). The rent-seeking firms that obtain these re-
sources might not be the most efficient ones.
6
Corruption is defined as the misuse of power by public officials for private personal
gain (Bardhan, 1997).
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 1
Variable definitions.
Variable Definitions

DA_MJ Discretionary accruals calculated using the modified Jones model

DA_FLJ Discretionary accruals calculated using the forward-looking Jones model
DA_K Discretionary accruals calculated using the performance-adjusted approach
ABS_DA Absolute value of DA
ETC Entertainment and travel costs divided by sales revenue
RENT Excessive ETC. It is the residual of an ETC regression equation shown in Appendix A.
Ln (1 + RENT per capita) The natural logarithm of 1 plus RENT per capita
SOE SOE dummy, which equals 1 if the ultimate owner is the government or other state-owned enterprises, and 0 otherwise
PC Political connection dummy, which equals 1 if the chairperson of the board of directors or CEO is a member of the National People's Congress,
Chinese People's Political Consultative Conference (CPPCC), or has political experience, and 0 otherwise
BIG4 Auditing quality dummy, which equals 1 if the firm's reporting is audited by a Big Four auditing firm, 0 otherwise
SIZE Size of the firm, which is the natural log of total assets at the end of accounting period
LEVER Leverage of long liabilities, non-current liabilities divided by the sum of owners' equity, and non-current liabilities
MTB Market-to-Book ratio: sum of equity market value and market value of liabilities divided by book value of total assets
ROA Return on assets, which is net income divided by total assets
LOSS Loss dummy, which equals 1 if net income in the preceding year is lower than zero, and 0 otherwise
EISSUE Stock issuing dummy, which equals 1 if the capital stock share is greater in year t + 1 than in year t, and 0 otherwise
TOP1 The percent of ownership by the largest shareholder
INSIDER The percent of ownership by senior executives
BOARDSIZE Board size measured by the natural logarithm of 1 plus the total number of directors on the board
INDDIR The percent of the number of independent directors to total directors on the board
CGI Government Integrity Index, which equals the mean of the sub-item “government integrity” from 2010 to 2012 in Wang et al. (2015).
LAWJUST Justice and Efficiency Index of the legal system, which equals the mean of the sub-item “justice and efficiency of legal system” from 2010 to 2012
Wang et al. (2015)

2.2. Earnings management Richardson, & Tuna, 2003), and the performance-adjusted Jones model
(Kothari, Leone, & Wasley, 2005). These methods are different varia-
Firms frequently practice earnings management to smooth out tions on capturing accounting information quality.
earnings, which then appear to be stable over time. Earnings manage- Chinese firms widely practice earnings management, casting doubt
ment is used to hide firm-specific information, and thus it is an im- on the credibility of the accounting information, and for different mo-
portant indicator of accounting information quality. Reliable ac- tives (Chen, Chen, Lobo, & Wang, 2011; Liu & Lu, 2007; Qiao, Fung,
counting information is imperative for a firm's stakeholders to make Miao, & Fung, 2017). Earnings management helps rent-seeking firms
sound decisions. Earning management distorts the true earnings of a hide firm-specific information from the public, and thereby rent-
firm, which provides an important benchmark for measuring perfor- seeking is a key factor that drives earnings management.
mance (Dechow, Ge, & Schrand, 2010) and may create different
meanings for different people (Dechow & Schrand, 2004). In practice, a
2.3. Conceptual frameworks and hypothesis development
firm can manipulate earnings only within the limits of generally ac-
cepted accounting principles to distort its genuine performance. In the
Rent seeking can be evaluated in the theoretical economic frame-
case of manipulated earnings, the firm's accounting information is
work of self-preservation (Karni & Schmeidler, 1986). In this frame-
clearly not at all useful for predicting its future performance.
work, rational executives who engage in rent-seeking activities at firms
Firms can manage earnings by two methods (Badertscher, 2011).
are likely to take appropriate actions in the event of bad outcomes for
First, a firm can use discretionary accruals (DAs) to reach the desired
self-protection. Because of the drive for self-protection, there are two
level of earnings. This practice implies that the accounting information
main rationales that link rent-seeking to earnings management. First,
quality is low if the DA level is high. Second, a firm can make changes
we contend that when rent-seeking firms engage in corruption, bribery,
in real business activities to manipulate earnings, which is called “real
or relationship reciprocity to gain favors from government officials,
earnings management” (Roychowdhury, 2006). The real earnings
they are likely to take actions to protect themselves. It is because rent-
management approach entails measurement errors because real busi-
seeking activities make firm management nervous and damage staff
ness activities are more challenging to arrange (Kim & Sohn, 2013).
morale if employees engaged in them are known to the public or are
Customers and/or suppliers may not fully cooperate with the firms to
caught in the act of wrongdoing. If the rent-seeking firms are rational,
allow flexible booking costs or revenues.
they will take a conservative approach to hide the benefits of their rent-
DAs that are value relevant can affect decision making (Healy &
seeking activities and prefer that the firm maintain a low profile to
Palepu, 2001). That is, firms commonly use DAs to manipulate earnings
minimize public scrutiny. Accordingly, executives will take steps to
and thus adversely affect decisions that are based on accounting in-
create firm opaqueness in order to shield themselves from public
formation. For instance, acquiring firms in mergers and acquisitions
scrutiny. Earnings management is closely linked to firm opaqueness
tend to use DAs to window dress their firms before a transaction (Chen,
because more earnings management, which helps hide firm-specific
Thomas, & Zhang, 2016; Kassamany, Ibrahim, & Archbold, 2017). Si-
information from the public, implies greater opaqueness (Caprio et al.,
milarly, firms making an initial public offering (IPO) use DAs to boost
2013; Durnev & Fauver, 2011; Hutton, Marcus, & Tehranian, 2009;
their chances of obtaining regulatory approval, raise the offering price,
Smith, 2016). Thus, it is not surprising that rent-seeking firms engage in
and lower the IPO discount (Aharony, Lin, & Loeb, 1993; Teoh, Welch,
earnings management for self-preservation. We call this reason the
& Wong, 1998). We examine whether rent-seeking affects DAs used to
benefits-hiding hypothesis.
capture accounting information quality.
The second rationale grows out of the political extraction risk re-
DAs are a common method used by firms to manipulate earnings.
lated to self-preservation. After garnering economic benefits from the
Several DA methods have been proposed in the literature, including the
government, rent-seeking firms may attract attention from corrupt
Jones model (Jones, 1991), the modified Jones model (Dechow, Sloan,
government officials. The corrupt government officials have incentives
& Sweeney, 1995), the forward-looking Jones model (Dechow,
for exploiting the situation for their private benefit. That is, the officials

96
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 2
Descriptive statistics.
Variable N Mean Std. Min p25 Median p75 Max

ABS_DA_MJ 8610 0.085 0.085 0 0.027 0.059 0.113 0.432

ABS_DA_FLJ 7598 0.081 0.078 0 0.026 0.057 0.107 0.383
ABS_DA_K 8610 0.082 0.081 0 0.026 0.058 0.108 0.409
ETC 8705 0.006 0.007 0 0.002 0.004 0.007 0.046
RENT 8614 0 0.006 −0.010 −0.003 −0.001 0.001 0.028
Ln(1 + RENT_per_capita) 8614 −0.001 0.928 −2.401 −0.584 0.038 0.612 2.279
BIG4 8614 0.092 0.289 0 0 0 0 1
SIZE 8614 21.82 1.166 19.31 20.98 21.69 22.52 25.22
LEVER 8614 0.146 0.176 0 0.011 0.066 0.236 0.749
MTB 8614 1.988 1.642 0.195 0.901 1.547 2.490 9.328
ROA 8614 0.0530 0.071 −0.155 0.016 0.042 0.079 0.381
LOSS 8614 0.0790 0.269 0 0 0 0 1
EISSUE 8614 0.369 0.483 0 0 0 1 1
TOP1 8614 0.361 0.152 0.090 0.239 0.340 0.471 0.754
INSIDER 8614 0.053 0.127 0 0 0 0.017 0.610
BOARDSIZE 8614 1.854 0.227 1.099 1.792 1.946 1.946 2.398
INDDIR 8614 0.803 0.431 0.430 0.500 0.667 0.857 2.667
SOE 8614 0.434 0.496 0 0 0 1 1
PC 8614 0.403 0.491 0 0 0 1 1
ANTICOR 8614 0.424 0.494 0 0 0 1 1
CGI 8614 3.336 0.170 2.955 3.195 3.395 3.425 3.650
LAWJUST 8614 2.963 0.137 2.590 2.880 2.980 3.050 3.255

This table presents descriptive statistics of the main variables. Definitions of all variables are in Table 1.

may engage in political extraction from the rent-seeking firms. To
minimize this political extraction risk, rent-seeking firms have in-
centives for hiding their investments and cash holdings (Caprio et al.,
2013; Smith, 2016) and decrease corporate governance (Durnev &
Fauver, 2011).7 The political extraction risk framework predicts that
rent-seeking firms will engage in earnings management to decrease
corporate governance and increase firm opaqueness. We call this the
political extraction risk hypothesis.
We explore rent-seeking behavior in terms of entertainment and
travel costs (ETC) (Cai et al., 2011). Chinese firms typically use ETC as a
general category of expenses that involves networking with government
officials to seek economic benefits. Fake or inflated expenses are also
submitted for reimbursement for illegitimate uses. ETC serves the
functions of both “protection money” and “grease money” for the two
rationales behind rent-seeking discussed earlier. As seen in Cai et al.
(2011) using survey data, the World Bank and the Enterprise Survey
Organization of the National Bureau of Statistics of China confirm that
ETC encompasses the extent of rent-seeking in China.
ETC at a firm also includes normal business-related expenses,
however. To account for normal business-related ETC expenses, we
follow Cai et al. (2011) in using a multiple regression model that as-
sociates ETC with normal business activities and firm attributes. Then,
we recover the residuals from the regression equation and use them to
represent rent-seeking expenses (RENT). These residuals are excessive
ETC, and they do not include the impact of the firm's normal business
on total ETC expenses. The procedure allows us to purge the business-
related component of ETC to derive a proxy for firm rent-seeking.
Hiding economic benefits and avoiding political extraction motivate
firms to engage in earnings management, so we use this proxy for rent-
seeking in proposing the following hypothesis.
H1. A firm's rent-seeking activity as indicated by RENT is positively
associated with earnings management.
Both the benefits-hiding and political extraction risk hypotheses are
the expected result of executives' self-preservation. An implication is
7
Although some studies argue that corruption could lower the political extraction risk
because it plays a role as “protection money” (Cai et al., 2011; Li & Huang, 2010), this
strategy entails risks for the firm because of frequent turnover among government offi-
cials, and it is debatable whether corruption enhances firm performance over time
(Svensson, 2003).
that when a government implements some public policies, such as anti-
corruption campaign, to crack down corruption (which is a part of rent-
seeking activities), executives' need for self-preservation to mitigate risk
of being caught is likely reduced due to a lower level of rent-seeking
activities. Then, our second testable hypothesis is:
H2. The effect of rent-seeking on earnings management weakens in an
anti-corruption campaign.
Two moderating variables are important in the relation between
rent-seeking and earnings management. First, state ownership is an
important component of rent-seeking in an emerging economy because
in this environment the government plays a greater role. Many Chinese
firms are SOEs, which clearly have close ties with the government. SOEs
have low incentives for deliberately hiding their economic benefits
because they are part of the government and do not have a strong sense
of self-preservation. Thus, they do not have a great incentive for en-
gaging in rent-seeking because they can easily obtain resources from
the government through normal channels. The government prioritizes
SOEs over non-SOEs in terms of subsidies and other financial support as
they help the government realize economic plans to achieve its goals. In
contrast, non-SOEs, which have private ownership, have a strong in-
centive for hiding economic benefits. Non-SOEs need to engage in rent-
seeking to secure help from government officials, such as obtaining
bank loans and subsidies (Ayyagari, Demirgüç-Kunt, & Maksimovic,
2010). Executives at non-SOEs can receive better compensation if they
can garner government resources because their compensation is more
flexible than that of counterparts at SOEs (Xu, Li, Yuan, & Chan, 2014).
As a result, non-SOEs have more incentives than SOEs for engaging in
rent-seeking. Thus, the state ownership variable moderates the relation
between rent-seeking and earnings management.
The second moderating variable relates to the risk of political ex-
traction. When a firm is politically connected (PC), the risk of political
extraction is low. PC firms have access to government officials and
government resources. Corrupt government officials show restraint in
extracting benefits from PC firms because of a concern over political
repercussions due to PC firms' relation with other government officials,
implying that PC firms have less need than non-PC firms to engage in
rent-seeking.
Overall, PC firms can leverage their political network to obtain
government resources. In addition, when a firm is politically connected,
it has less need to manage earnings because it has a lower risk of

97
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 3
Pearson correlation matrices.
(1) (2) (3) (4) (5) (6) (7) (8) (9)

ABS_DA_MJ (1) 1.00

ABS_DA_FLJ (2) 0.92*** 1.00

(0.00)
ABS_DA_K (3) 0.94*** 0.93*** 1.00
(0.00) (0.00)
RENT (4) 0.02** 0.03** 0.03** 1.00
(0.02) (0.02) (0.01)
BIG4 (5) −0.03** −0.03** −0.02* 0.00 1.00
(0.02) (0.02) (0.07) (0.94)
SIZE (6) −0.04*** −0.03** −0.04*** 0.00 0.19*** 1.00
(0.00) (0.02) (0.00) (0.73) (0.00)
LEVER (7) 0.06*** 0.06*** 0.05*** −0.00 0.06*** 0.50*** 1.00
(0.00) (0.00) (0.00) (0.81) (0.00) (0.00)
MTB (8) 0.10*** 0.07*** 0.08*** 0.01 −0.03** −0.54*** −0.39*** 1.00
(0.00) (0.00) (0.00) (0.43) (0.02) (0.00) (0.00)
ROA (9) 0.15*** 0.10*** 0.12*** 0.01 0.01 0.01 −0.23*** 0.32*** 1.00
(0.00) (0.00) (0.00) (0.64) (0.33) (0.54) (0.00) (0.00)
LOSS (10) 0.09*** 0.11*** 0.10*** 0.03** −0.01 −0.08*** 0.10*** 0.04*** −0.16***
(0.00) (0.00) (0.00) (0.01) (0.43) (0.00) (0.00) (0.00) (0.00)
EISSUE (11) 0.04*** 0.03** 0.02** −0.02** 0.00 −0.05*** −0.09*** 0.14*** 0.20***
(0.00) (0.03) (0.03) (0.02) (0.80) (0.00) (0.00) (0.00) (0.00)
TOP1 (12) −0.02* −0.03** −0.02* 0.01 0.09*** 0.27*** 0.08*** −0.10*** 0.11***
(0.10) (0.03) (0.08) (0.54) (0.00) (0.00) (0.00) (0.00) (0.00)
INSIDER (13) −0.00 −0.01 −0.00 0.00 −0.03*** −0.23*** −0.22*** 0.17*** 0.13***
(0.92) (0.45) (0.77) (0.83) (0.01) (0.00) (0.00) (0.00) (0.00)
BOARDSIZE (14) −0.03*** −0.02** −0.03** −0.00 0.02* 0.22*** 0.13*** −0.13*** −0.00
(0.01) (0.03) (0.01) (0.94) (0.10) (0.00) (0.00) (0.00) (0.82)
INDDIR (15) 0.03** 0.02** 0.02* 0.00 0.07*** 0.02 −0.02** 0.04*** −0.01
(0.02) (0.04) (0.07) (1.00) (0.00) (0.10) (0.05) (0.00) (0.50)
SOEs (16) −0.07*** −0.05*** −0.06*** 0.00 0.05*** 0.33*** 0.28*** −0.23*** −0.13***
(0.00) (0.00) (0.00) (0.98) (0.00) (0.00) (0.00) (0.00) (0.00)
PC (17) −0.03*** −0.03*** −0.02** 0.00 −0.00 0.08*** 0.00 −0.06*** 0.05***
(0.01) (0.00) (0.03) (0.71) (0.90) (0.00) (0.89) (0.00) (0.00)
ANTICOR (18) 0.00 0.00 0.01 0.00 0.24*** 0.10*** −0.00 0.02** −0.05***
(0.98) (0.97) (0.58) (0.86) (0.00) (0.00) (0.64) (0.04) (0.00)

(10) (11) (12) (13) (14) (15) (16) (17) (18)

LOSS (10) 1.00

EISSUE (11) −0.10*** 1.00

(0.00)
TOP1 (12) −0.09*** 0.00 1.00
(0.00) (0.95)
INSIDER (13) −0.08*** 0.20*** −0.04*** 1.00
(0.00) (0.00) (0.00)
BOARDSIZE (14) 0.00 −0.03*** −0.02* −0.15*** 1.00
(0.96) (0.00) (0.05) (0.00)
INDDIR (15) 0.02 0.05*** 0.00 0.06*** −0.46*** 1.00
(0.12) (0.00) (0.77) (0.00) (0.00)
SOE (16) 0.07*** −0.20*** 0.14*** −0.35*** 0.25*** −0.11*** 1.00
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
PC (17) −0.06*** 0.02** 0.02** −0.02** 0.02** 0.02 −0.06*** 1.00
(0.00) (0.02) (0.03) (0.02) (0.03) (0.11) (0.00)
ANTICOR (18) −0.00 0.11*** −0.02 0.06*** −0.12*** 0.28*** −0.13*** 0.01 1.00
(0.76) (0.00) (0.12) (0.00) (0.00) (0.00) (0.00) (0.19)

This table reports the Pearson correlation matrices, respectively. P-values are in parentheses. Definitions of all variables are in Table 1. *, **, and *** indicate
statistical significance at the 10%, 5%, and 1% levels, respectively.

political extraction than non-PC firms. In light of these two arguments,
we propose the following hypothesis:
H3. The positive effect of rent-seeking on earnings management is
stronger for non-SOEs than SOEs and for non-SOEs without political
connections than for those with political connections.
If we find support for H3, then the underlying conceptual frame-
works on the impact of rent-seeking on earnings management are valid.
This is because (1) non-SOEs have a stronger tendency than SOEs to
hide rent-seeking benefits using earnings management, and (2) PC non-
SOEs incur less political extraction risk than do non-SOEs without PC.
3. Research design
3.1. Data source and sample selection
We hand-collect the ETC items from annual financial reports to
examine the impact of rent-seeking on a firm's earnings management.
Specifically, we use the sum of the accounting items “entertainment
fees” and “travel fees” to proxy for ETC. Although mandatory disclosure
of “entertainment fees” and “travel fees” began in 2010, the 2010 an-
nual reports disclosed information for 2009. Hence, our sample is from
2009 to 2014. Based on the regression results in Appendix A, rent-

98
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 4
Univariate analysis.
Panel A High RENT vs. low RENT group

Low RENT (1) High RENT (2) (1)–(2)

Obs. Mean Obs. Mean Difference

ABS_DA_MJ 2584 0.090 2583 0.089 0.002

ABS_DA_FLJ 2242 0.084 2260 0.084 0.000
ABS_DA_K 2584 0.086 2583 0.085 0.001

Panel B SOEs vs. non-SOEs (full sample)

Non-SOEs (1) SOEs (2) (1)–(2)

Obs. Mean Obs. Mean Difference

ABS_DA_MJ 4874 0.090 3736 0.079 0.011***

ABS_DA_FLJ 4066 0.084 3532 0.076 0.008***
ABS_DA_K 4874 0.086 3736 0.077 0.010***

Panel C PC vs. non-PC (non-SOEs sample)

Non-PC & non-SOEs (1) PC & non-SOEs (2) (1)–(2)

Obs. Mean Obs. Mean Difference

ABS_DA_MJ 2787 0.093 2087 0.086 0.007***

ABS_DA_FLJ 2333 0.087 1733 0.080 0.007***
ABS_DA_K 2787 0.089 2087 0.083 0.006**

This table presents comparison of key variables between different subsamples. Panel A reports the differences of key variables between high RENT and low RENT
groups. Panel B reports the differences of key variables between SOEs and non-SOEs, Panel C reports the variations of key variables between politically connected
companies and companies with no political ties. We define a firm as politically connected if the chairperson of the board or CEO has political experience or a political
identity. Definitions of other variables are in Table 1. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.

seeking activity (RENT) is the residual from the following ETC regres-
sion equation:
ETCi, t = β0 + β1 CGIi, t + β2 BIG 4i, t + β3 SIZEi, t + β4 LEVERi, t + β5 MTBi, t
+ β6 ROAi, t + β7 LOSSi, t + β8 EISSUEi, t + β9 TOP1i, t
+ β10 INSIDERi, t + β11 BOARDSIZEi, t + β12 INDDIRi, t + β13 SOEi, t
+ β14 PCi, t + β15 AN TICORi, t + β16 LAWJUSTi, t + θj + γt + μi, t
(1)
where CGI is a government integrity index, BIG4 is audit quality, SIZE is
firm size, LEVER is financial leverage, MTB is the market-to-book ratio,
ROA is the return on assets, LOSS is a dummy to indicate negative
earnings, EISSUE is a dummy for stock issues, TOP1 is the percentage of
ownership of the largest shareholder, INSIDER is the percentage of
ownership of senior executives, BOARDSIZE is the number of directors
on the board, INDDIR is the percentage of independent directors on the
board, SOE is the state-ownership, PC is a dummy variable for political
connections, ANTICOR is a dummy variable for the period of the anti-
corruption campaign, and LAWJUST is a justice and efficiency index of
the legal system. θj and γt represent industry and year effects. Detailed
definitions are in Table 1.
We include all A-share firms listed on the Shanghai Stock Exchange
illegal activities.
In addition, we include (1) information on ultimate control at firms
obtained from the China Center for Economic Research and Financial
Database; (2) the ranking of Chinese certified public accounting firms
from the Chinese Institute of Certified Public Accountants8; (3) data
from the Government Integrity Index and Justice and Efficiency Index
of the legal system for each province in Wang, Yu, and Fan (2015); and
(4) political connections and financial data from the China Stock
Market and Accounting Research database. We winsorize the con-
tinuous variables at the 1% and 99% levels.
3.2. Measurement of key variables
We use three DAs to measure earnings management: the modified
Jones model (Dechow et al., 1995), the forward-looking Jones model
(Dechow et al., 2003), and the performance-adjust Jones model
(Kothari et al., 2005).
Jones (1991) provides a first model for capturing earnings man-
agement. He suggests first estimating Eq. (2) separately in an industry-
year subsample. Then, Jones (1991) uses Eq. (3) to compute DA (called
DA_J), in which the estimated coefficients are obtained from Eq. (2).
TACt 1 ⎞ ΔSalest ⎞ PPEt ⎞
= α0 ⎛ ⎜ ⎟ + α1 ⎛⎜ + α2 ⎛
⎟ ⎜ + εt
⎟

and the Shenzhen Stock Exchange. After deleting firms in the financial TAt − 1 ⎝ TAt − 1 ⎠ ⎝ TAt − 1 ⎠ ⎝ TAt − 1 ⎠ (2)
industry and those with missing values, we have 8614 firm-years from
2171 distinct firms. When using DA_FAJ measure, we have smaller TACt 1 ⎞ ΔSalest ⎞ PPEt ⎞
DAJ = 0 ⎛
−α ⎜ ⎟ 1 ⎛
−α ⎜ 2 ⎛
−α ⎟ ⎜ ⎟

sample size due to missing values. Using the universe of public firms, TAt − 1 ⎝ TAt − 1 ⎠ ⎝ TAt − 1 ⎠ ⎝ TAt − 1 ⎠ (3)
our approach offers an alternative to a survey data approach to gauge
where TAC is total accruals, which is calculated using Eq. (3):
the impact of rent-seeking on earnings management. Our firm-level
approach circumvents the common weaknesses of using a survey,
8
especially when survey respondents are reluctant to report negative or Available at http://www.cicpa.org.cn.

99
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 5 After estimating k1 by using the industry-level and year-level sample
Empirical results of rent-seeking and earnings management. separately, the normal portion of accounts receivable is k1 ∗ ΔSales,
(1) (2) (3) while the abnormal portion of accounts receivable is ΔAR − k1 ∗ ΔSales.
In addition, the current accruals are correlated with accruals from
ABS_DA_MJ ABS_DA_FLJ ABS_DA_K earlier periods and the firm's growth capacity (as measured by sales
⁎⁎⁎ ⁎⁎⁎ growth, GR_Sale). Dechow et al. (2003) use Eqs. (8) and (9) to estimate
RENT 1.530 1.418 1.680⁎⁎⁎
(3.71) (3.35) (4.28) the normal as well as abnormal portions of accruals. The abnormal
BIG4 −0.002 0.000 −0.001 portion (DA_FLJ) is the earnings management measure of a forward-
(−0.29) (0.01) (−0.14) looking Jones model:
SIZE 0.023⁎⁎⁎ 0.021⁎⁎⁎ 0.020⁎⁎⁎
(4.14) (3.77) (3.76) TACt (1 + κ )̂ ΔSalest − ΔARt ⎞ PPEt ⎞ TACt − 1 ⎞
LEVER 0.043⁎⁎⁎ 0.019 0.026⁎ = β0 + β1 ⎛ ⎜ + β2 ⎛ ⎟ + β3 ⎛
⎜ ⎟ ⎜ ⎟

TAt − 1 ⎝ TAt − 1 ⎠ ⎝ TAt − 1 ⎠ ⎝ TAt − 2 ⎠

(2.75) (1.24) (1.76)
MTB 0.003⁎ 0.002 0.002 + β4 (GRSalet ) + εt (8)
(1.91) (1.44) (1.45)
ROA 0.306⁎⁎⁎ 0.250⁎⁎⁎ 0.284⁎⁎⁎ TACt (1 + κ )̂ ΔSalest − ΔARt ⎞  ⎛ PPEt ⎞
(10.72) (8.61) (10.44) DAFLJ = − β0 − β1 ⎛ ⎜ − β2 ⎟ ⎜ ⎟

LOSS 0.004 0.007 0.004 TAt − 1 ⎝ TAt − 1 ⎠ ⎝ TAt − 1 ⎠

(0.88) (1.58) (0.93) TAC
EISSUE 0.005⁎⁎ 0.003 0.003 − β3 ⎛
⎜
t−1 ⎞
⎟ − β4 (GRSalet )
(2.01) (1.49) (1.43) ⎝ TAt − 2 ⎠ (9)
TOP1 −0.082⁎⁎⁎ −0.091⁎⁎⁎ −0.074⁎⁎
Dechow et al. (1995) find that, at firms with extreme performance,
(−2.69) (−3.04) (−2.55)
INSIDER −0.017 0.012 −0.016 DAs are influenced by performance. Based on this, Kothari et al. (2005)
(−0.83) (0.44) (−0.82) add performance factors to the currently used accruals-estimating
BOARDSIZE 0.023⁎⁎ 0.026⁎⁎⁎ 0.025⁎⁎⁎ model to derive a “performance-adjusted Jones model.” For the per-
(2.49) (2.78) (2.76)
formance-adjusted Jones model, we follow Kothari et al. (2005) in es-
INDDIR 0.005⁎ 0.005 0.004
(1.71) (1.64) (1.44)
timating Eq. (10) separately in an industry-year subsample and then
SOE −0.013⁎⁎ −0.011⁎⁎ −0.010⁎ bring the parameters estimated into Eq. (11) to calculate DAs (DA_K).
(−2.25) (−1.98) (−1.73)
−0.004 −0.004 −0.001 TACt 1 ⎞ ΔSalest − ΔARt ⎞ PPEt ⎞
PC = γ0 + γ1 ⎛ ⎜ ⎟ + γ2 ⎛ ⎜ + γ3 ⎛ ⎟ ⎜ ⎟

(−0.68) (−0.76) (−0.22) TAt − 1 ⎝ TA t − 1 ⎠ ⎝ TA t − 1 ⎠ TA

⎝ t−1 ⎠
ANTICOR −0.017⁎⁎⁎ −0.016⁎⁎⁎ −0.015⁎⁎⁎
(−3.13) (−2.79) (−2.76) + γ4 ROAt − 1 + εt (10)
Intercept −0.455⁎⁎⁎ −0.441⁎⁎⁎ −0.402⁎⁎⁎
(−3.60) (−3.43) (−3.34) TACt 1 ⎞ ΔSalest − ΔARt ⎞ PPEt ⎞
Firm Yes Yes Yes DAKt = − γ0 ̂ − γ1 ̂ ⎛
⎜ − γ2 ̂ ⎛
⎟ ⎜ − γ3 ̂ ⎛⎟ ⎜ ⎟

TAt − 1 ⎝ TAt − 1 ⎠ ⎝ TAt − 1 ⎠ ⎝ TAt − 1 ⎠

Industry Yes Yes Yes
Year Yes Yes Yes − γ4 ̂ROAt − 1 (11)
N 8610 7598 8610
Adj. R2 0.207 0.168 0.189 Thus, we have DA_MJ (from Eq. (6)), DA_FLJ (from Eq. (9)), and
DA_K (from Eq. (11)) to capture three different aspects of earnings
This table presents the results of the impact rent-seeking has on earnings management for robustness. To circumvent the positive and negative
management. All variables are defined in Table 1. t statistics in parentheses are
values of different measures, we take absolute values of DA_MJ, DA_FLJ,
adjusted for heteroskedasticity. *, **, and *** indicate statistical significance at
and DA_K.9 Large absolute values of these measures suggest poor
the 10%, 5%, and 1% levels, respectively.
earnings management.

TAC = (∆CA − ∆CASH ) − (∆CL − ∆STD − ∆ITP ) − DEP (4) 3.3. Empirical model
In Eq. (4), Δ indicates the change in amount, CA is current assets,
We use a regression model in Eq. (12) to test H1 to H3 as follows:
CASH is cash and cash equivalents, CL is current liabilities, STD is long-
term liabilities that mature within one year, ITP is the income tax that EM = β0 + β1 RENT + β2 BIG 4 + β3 SIZE + β4 LEVER + β5 MTB + β6 ROA
should be paid, DEP is the depreciation of tangible assets, TA is total
+ β7 LOSS + β8 EISSUE + β9 TOP1 + β10 INSIDER
assets, ΔSales measures a change in sales, PPE is the gross value of fixed
assets, and αi is the estimated coefficient of αi in Eq. (2). + β11 BOARDSIZE + β12 INDDIR + β13 SOE + β14 PC
Jones's model assumes that total sales are not easily manipulated by + β15 ANTICOR + αi + θi + γi + μi, t (12)
firms. To improve on the Jones model, Dechow et al. (1995) use Eqs. (5)
and (6) to modify Jones's model to calculate DAs and accrual profit where EM is earnings management, measured by various absolute DA
(DA_MJ). values; RENT is rent-seeking, measured by the residual of the ETC re-
gression equation in Appendix A. RENT is used to test H1 in Eq. (12).
1 ⎞ ΔSalest − ΔARt ⎞ PPEt ⎞
0 ⎛
NDAMJt = α ⎜ 1 ⎛
+α
⎟ ⎜ 2 ⎛
+α ⎟ ⎜ ⎟ The control variables are defined in Eq. (1). We also include industry-
⎝ TA ⎠ ⎝ TA ⎠ TA
⎝ t−1 ⎠ (5)
t − 1 t − 1
fixed effects (αi), year-fixed effects (γi), and firm-fixed effects (θi). These
TACt control variables incorporate the period of an anti-corruption cam-
DAMJt = − NDAMJt paign, industry-level factors, and firm managerial factors. We do not
TAt − 1 (6)
control for region-fixed effects because firms do not change the city in
i is the estimated coefficient of αi in Eq. (2). ΔAR is the
In Eq. (5), α which they are chartered, and therefore firm-fixed effects incorporate
changes in accounts receivable, and other variables have the same de- region-fixed effects. We run Eq. (12) using ordinary least squares. If H1
finitions as in Eqs. (2) to (4). is valid, β1 will be positive and significant.
Dechow et al. (2003) use Eq. (7) to estimate the normal level of
accounts receivable to measure its impact on earnings management:
9
The findings using Jones's (1991) model are qualitatively similar. To conserve space,
∆AR = k 0 + k1 ∗∆Sales + ε (7) we do not report the results.

100
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 6
Rent-seeking and earnings management: evidence from the shock of the anti-corruption campaign.
(1) (2) (3) (4) (5) (6)

ANTICOR = 1 ANTICOR = 0 ANTICOR = 1 ANTICOR = 0 ANTICOR = 1 ANTICOR = 0

ABS_DA_MJ ABS_DA_MJ ABS_DA_FLJ ABS_DA_FLJ ABS_DA_K ABS_DA_K

RENT 0.729 2.195⁎⁎⁎ 0.759 1.974⁎⁎⁎ 1.164 2.378⁎⁎⁎

(0.76) (3.81) (0.73) (3.19) (1.29) (4.41)
BIG4 0.018 0.028 0.012 0.025 0.014 0.039⁎
(0.78) (1.42) (0.50) (1.18) (0.61) (1.85)
SIZE 0.096⁎⁎⁎ 0.019⁎⁎ 0.101⁎⁎⁎ 0.014 0.089⁎⁎⁎ 0.018⁎⁎
(6.91) (2.06) (6.62) (1.52) (6.69) (2.05)
LEVER 0.089⁎⁎ 0.060⁎⁎⁎ 0.023 0.041⁎ 0.056 0.041⁎
(2.24) (2.70) (0.55) (1.89) (1.37) (1.90)
MTB 0.002 0.002 0.004 0.000 0.001 0.001
(0.48) (0.82) (1.05) (0.18) (0.16) (0.57)
ROA 0.320⁎⁎⁎ 0.316⁎⁎⁎ 0.172⁎⁎⁎ 0.263⁎⁎⁎ 0.293⁎⁎⁎ 0.304⁎⁎⁎
(4.07) (7.80) (2.65) (6.23) (4.16) (8.07)
LOSS −0.008 0.002 −0.010 0.010 −0.008 0.004
(−0.90) (0.25) (−1.06) (1.47) (−0.91) (0.60)
EISSUE −0.004 0.003 −0.003 0.001 −0.004 0.002
(−0.77) (1.04) (−0.62) (0.35) (−0.85) (0.76)
TOP1 −0.014 −0.041 0.024 −0.073 0.014 −0.061
(−0.20) (−0.78) (0.34) (−1.46) (0.21) (−1.27)
INSIDER −0.045 −0.016 −0.023 0.070 −0.025 −0.019
(−0.94) (−0.48) (−0.44) (1.16) (−0.53) (−0.62)
BOARDSIZE 0.046⁎⁎ 0.004 0.049⁎⁎⁎ 0.010 0.057⁎⁎⁎ 0.003
(2.50) (0.24) (2.78) (0.64) (3.21) (0.20)
INDDIR 0.004 −0.002 0.005 −0.003 0.005 −0.004
(1.04) (−0.41) (1.20) (−0.46) (1.19) (−0.70)
SOE −0.013⁎ −0.012 −0.010 −0.010 −0.011 −0.006
(−1.65) (−0.86) (−1.27) (−0.69) (−1.42) (−0.44)
PC −0.009 0.006 −0.004 0.006 −0.006 0.009
(−0.58) (0.70) (−0.29) (0.72) (−0.40) (1.15)
Intercept −2.095⁎⁎⁎ −0.343⁎ −2.216⁎⁎⁎ −0.257 −1.962⁎⁎⁎ −0.318⁎
(−6.83) (−1.72) (−6.49) (−1.28) (−6.59) (−1.69)
Firm Yes Yes Yes Yes Yes Yes
Industry Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes
N 3648 4962 3417 4181 3648 4962
Adj. R2 0.293 0.206 0.242 0.164 0.274 0.202

This table reports the regression results of Eq. (4) in the pre-campaign period (ANTICOR = 0, year ∈ (2009, 2012)) and post-campaign period (ANTICOR = 1,
year ∈ (2013, 2014)). T-statistics in parentheses are adjusted for heteroskedasticity. Definitions of all variables are in Table 1. *, **, and *** indicate statistical
significance at the 10%, 5%, and 1% levels, respectively.

To test H2, we test whether China's recent strict anti-corruption
campaign influences the relation between RENT and EM measures, by
splitting the sample into pre- and post–anti-corruption campaign sub-
samples. In addition, the recent anti-corruption campaign is an exo-
genous shock to the system, which alleviates the endogeneity concern
between rent-seeking and earnings management.
To examine H3, we split the firms in our sample according to
whether they have state ownership (SOEs vs. non-SOEs) and political
connections (PC vs. non-PC). If H3 is valid, then β1 for the RENT vari-
able will be positive and significant for the non-SOE and non-PC sub-
samples.
4. Results and discussion
4.1. Descriptive statistics
Table 2 lists the descriptive statistics for the major variables. Over
the period 2009 to 2014, the mean of absolute values are 0.085 for
DA_MJ, 0.081 for DA_FLJ, and 0.082 for DA_K; and the standard de-
viations are 0.085, 0.078, and 0.081, respectively. These values in-
dicate that although DAs vary across firms, they have similar profiles.
The mean of RENT is 0.000 and the standard deviation is 0.006,
implying variations in the variable among companies. The mean of ETC
comprises 0.6% of sales revenue in our sample, which differs from that
of Cai et al. (2011).10
The mean of SOE is 0.434 and the mean of PC is 0.403, implying
that SOEs comprise 43.4% of the companies on the Chinese stock
market. Firms whose chairperson of the board or CEO has a political
background comprise 40.3% of firm-years. These values show that SOEs
and PC firms are common in China.
Table 3 presents the Pearson and Spearman correlation coefficients
of the major variables. As expected, the different earnings management
measures are highly correlated. However, the correlations among
earnings management measures with other variables are relatively low.
The results do not present a serious multicollinearity problem in
10
The mean operating revenue in the sample is RMB 518 million, so the mean of ETC
is RMB 3.108 million. In Cai et al. (2011), the mean of ETC/Sales is 1.5%, while our mean
of ETC/Sales is 0.6%, with the difference likely caused by firm size and time. Their
samples come from three surveys conducted by the World Bank and National Bureau of
Statistics of China. Although Cai et al. (2011) do not directly report firm size, survey data
typically include much smaller firms. For instance, the mean of firm age in Cai et al.
(2011) is 2.284 (logarithm value) whereas for the same variable in our study the mean is
2.601 (logarithm value). In addition, the mean of logarithm CEO pay is 3.63 in Cai et al.
(2011), while in our study, the mean of the same variable is 13.16, indicating that CEO
compensation in our sample is much higher than theirs. Our firm size is bigger than theirs
because of the positive correlation between firm size, firm age, and CEO pay.

101
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 7
Rent-seeking and earnings management: analysis of the nature of owners.
(1) (2) (3) (4) (5) (6)

SOE = 1 SOE = 0 SOE = 1 SOE = 0 SOE = 1 SOE = 0

ABS_DA_MJ ABS_DA_MJ ABS_DA_FLJ ABS_DA_FLJ ABS_DA_K ABS_DA_K

RENT 0.288 1.948⁎⁎⁎ 0.249 1.707⁎⁎⁎ 0.263 2.141⁎⁎⁎

(0.38) (3.79) (0.34) (2.94) (0.37) (4.30)
BIG4 0.000 −0.007 0.004 −0.007 0.004 −0.007
(0.06) (−0.85) (0.60) (−0.74) (0.66) (−0.86)
SIZE 0.021⁎⁎ 0.022⁎⁎⁎ 0.012 0.023⁎⁎ 0.018⁎⁎ 0.017⁎⁎
(2.42) (2.58) (1.43) (2.42) (2.22) (2.04)
LEVER 0.053⁎⁎ 0.042⁎ 0.035⁎ −0.002 0.022 0.035
(2.40) (1.71) (1.69) (−0.08) (1.07) (1.49)
MTB −0.001 0.003 −0.003 0.003 −0.003 0.003
(−0.56) (1.44) (−1.03) (1.39) (−1.12) (1.27)
ROA 0.307⁎⁎⁎ 0.324⁎⁎⁎ 0.259⁎⁎⁎ 0.248⁎⁎⁎ 0.274⁎⁎⁎ 0.310⁎⁎⁎
(6.39) (8.25) (5.66) (5.77) (5.99) (8.26)
LOSS 0.011⁎ −0.006 0.014⁎⁎ −0.002 0.012⁎⁎ −0.005
(1.85) (−0.80) (2.32) (−0.23) (2.02) (−0.72)
EISSUE 0.003 0.005 0.000 0.005 0.001 0.003
(0.75) (1.61) (0.09) (1.39) (0.27) (1.12)
TOP1 −0.047 −0.135⁎⁎⁎ −0.065 −0.138⁎⁎⁎ −0.056 −0.111⁎⁎
(−0.93) (−2.90) (−1.35) (−2.93) (−1.18) (−2.43)
INSIDER −0.115 −0.021 −0.029 0.002 −0.064 −0.020
(−1.04) (−1.00) (−0.24) (0.07) (−0.53) (−0.99)
BOARDSIZE 0.003 0.032⁎⁎ 0.018 0.031⁎⁎ 0.007 0.035⁎⁎⁎
(0.21) (2.33) (1.35) (2.07) (0.50) (2.71)
INDDIR −0.002 0.006 0.003 0.002 −0.002 0.005
(−0.40) (1.63) (0.70) (0.55) (−0.44) (1.41)
PC −0.004 −0.005 −0.009 0.006 −0.002 0.001
(−0.62) (−0.53) (−1.41) (0.60) (−0.34) (0.15)
ANTICOR −0.018⁎⁎ −0.016⁎ −0.015⁎⁎ −0.014 −0.019⁎⁎⁎ −0.010
(−2.42) (−1.76) (−1.97) (−1.41) (−2.67) (−1.16)
Intercept −0.376⁎ −0.434⁎⁎ −0.233 −0.477⁎⁎ −0.307⁎ −0.384⁎⁎
(−1.94) (−2.21) (−1.21) (−2.17) (−1.70) (−1.98)
Firm Yes Yes Yes Yes Yes Yes
Industry Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes
N 3736 4874 3532 4066 3736 4874
Adj. R2 0.165 0.227 0.145 0.178 0.148 0.211

This table presents the impact of the nature of the rent-seeking firm's ownership on the relation between rent-seeking and accounting information quality. We split
our sample into SOEs and non-SOEs based on the nature of the firm's ultimate controller. Columns 1 to 6 show the results of OLS regressions. t statistics in parentheses
are adjusted for heteroskedasticity. Definitions of all variables are presented in Table 1. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels,
respectively.

subsequent regression analyses. 4.2. Impact of rent-seeking on earnings management

Table 4 presents several univariate analyses across subsamples. We
separate the sample into 10 groups according to the values of RENT
from high to low, then we designate the top three groups as the “high
RENT group,” and the bottom three groups as the “low RENT group,”
and compare EM (using all three DAs) between the two groups in Panel
A. In terms of DA_MJ, DA_FLJ, and DA_K, the high RENT group does not
have significantly different mean values from the low RENT group.
Although it does not confirm H1, we need to be cautious due to the
univariate nature of the analysis. Many factors contribute to earnings
management.
We split the full sample into SOEs and non-SOEs and present the
variations of EM in Panel B. Panel B shows that three measurements of
EM and RENT are significantly higher for non-SOEs than for SOEs. We
attribute these findings to the privileges held by SOEs in accessing
scarce resources, such as financing, but non-SOEs need to smooth their
performance through manipulating DAs to obtain financing.
Panel C of Table 4 compares EM between PCs and non-PCs. Because
SOEs have government connections, we exclude SOEs and separate non-
SOE samples into those with and without political connections. Panel C
of Table 4 shows that all three EM measures are significantly higher for
firms without political connections (non-PC firms) than for PC firms,
indicating that PC firms have less pronounced EM than do non-PC firms.
The results in Panels B and C of Table 4 support H3.
4.2.1. Base results
We present the results of testing H1 in columns 1 to 3 of Table 5. In
all columns, the coefficients of RENT are positive and significant at the
1% level, suggesting that rent-seeking and earnings management are
positively correlated and implying that RENT motivates more earnings
management. The results are also economically significant. In column
1, the coefficient of RENT is 1.530, suggesting that a one-standard-de-
viation increase in RENT is associated with a 10.8% increase in the
standard deviation of ABS_DA_MJ (absolute value of discretionary ac-
cruals by a modified Jones model). Columns 2 to 3 have qualitatively
similar results and economic significance. Other control variables, if
they are significant, have the expected signs. For instance, when a firm
is large (SIZE is large), the coefficients of SIZE are all positive at the 1%
level of significance, suggesting that a larger firm, on average, drums up
its DAs, as it has more flexibility in managing its accounting informa-
tion. ROA has a positive effect on earnings management across all four
models, implying that more profitable firms are likely to manage their
earnings. A profitable firm tends to give investors an impression of
earnings stability to avoid speculation. Similarly, a larger board size has
a positive and significant effect on earnings management, portraying
the stability and sustainability of firm performance.

102
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 8
Rent-seeking and earnings management: analysis of political connection for non-state-owned firms.
(1) (2) (3) (4) (5) (6)

PC = 1 PC = 0 PC = 1 PC = 0 PC = 1 PC = 0

ABS_DA_MJ ABS_DA_MJ ABS_DA_FLJ ABS_DA_FLJ ABS_DA_K ABS_DA_K

RENT 0.936 2.465⁎⁎⁎ 0.258 2.499⁎⁎⁎ 1.344 2.495⁎⁎⁎

(1.01) (3.78) (0.26) (3.24) (1.45) (4.04)
BIG4 −0.012 −0.004 −0.009 −0.007 −0.012 −0.004
(−0.95) (−0.41) (−0.66) (−0.56) (−0.97) (−0.42)
SIZE 0.027⁎ 0.024⁎⁎ 0.017 0.030⁎⁎ 0.024⁎ 0.016
(1.94) (2.03) (1.15) (2.28) (1.75) (1.39)
LEVER 0.069⁎ 0.022 0.030 −0.018 0.055 0.015
(1.93) (0.62) (0.81) (−0.49) (1.54) (0.45)
MTB 0.002 0.005⁎ 0.005 0.005 0.002 0.004⁎
(0.61) (1.87) (1.06) (1.53) (0.52) (1.73)
ROA 0.282⁎⁎⁎ 0.323⁎⁎⁎ 0.217⁎⁎⁎ 0.235⁎⁎⁎ 0.274⁎⁎⁎ 0.317⁎⁎⁎
(4.16) (6.05) (3.23) (3.79) (4.31) (6.20)
LOSS −0.011 0.000 −0.003 0.003 −0.009 0.000
(−0.83) (0.01) (−0.23) (0.27) (−0.66) (0.01)
EISSUE 0.006 0.006 0.011⁎⁎ 0.001 0.005 0.003
(1.18) (1.24) (2.08) (0.27) (1.11) (0.67)
TOP1 −0.130⁎ −0.105 −0.170⁎⁎ −0.107 −0.109 −0.072
(−1.86) (−1.54) (−2.54) (−1.49) (−1.60) (−1.06)
INSIDER 0.007 −0.038 0.051 −0.031 0.008 −0.035
(0.19) (−1.41) (0.86) (−0.91) (0.21) (−1.38)
BOARDSIZE 0.023 0.035⁎ 0.023 0.037⁎ 0.028 0.034⁎
(1.17) (1.80) (1.10) (1.65) (1.44) (1.85)
INDDIR 0.006 0.005 0.004 0.001 0.004 0.006
(1.18) (0.96) (0.72) (0.19) (0.70) (1.06)
ANTICOR −0.027⁎ −0.010 −0.022 −0.011 −0.019 −0.003
(−1.92) (−0.78) (−1.47) (−0.76) (−1.38) (−0.27)
Intercept −0.504⁎ −0.445⁎ −0.327 −0.621⁎ −0.475 −0.337
(−1.66) (−1.66) (−1.00) (−1.95) (−1.58) (−1.30)
Firm Yes Yes Yes Yes Yes Yes
Industry Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes
N 2087 2787 1733 2333 2087 2787
Adj. R2 0.220 0.234 0.158 0.191 0.198 0.218

This table presents the impact of the political connections of the rent-seeking firm on the relation between rent-seeking and earnings management. We split our non-
SOE sample into PCs and non-PCs. We define the firm as politically connected if its chairperson of board or CEO has political experience or a political identity.
Because SOEs by definition have connections with the government, we exclude SOEs when considering political connections. Columns 1 to 6 show the results of OLS
regressions. t statistics in parentheses are adjusted for heteroskedasticity. Definitions of all variables are presented in Table 1. *, **, and *** indicate significance at
the 10%, 5%, and 1% levels, respectively.

4.2.2. H2 and external shock test corruption campaign.

We document a positive link between RENT and earnings manage-
ment in Table 5. If H2 is valid, an anti-corruption campaign would
weaken the impact of RENT on EM. On December 4, 2012, two weeks
after Xi Jinping was elected president of China, the eighteenth National
People's Congress (the legislature) announced a policy to improve
working efficiency among government officials as well as to eradicate
corruption. The policy, considered the beginning of the recent anti-
corruption campaign, aims to curb the extravagant lifestyle of officials
and encourage austerity in all branches of government (Lin, Morck,
Yeung, & Zhao, 2016; Tang, Lin, Peng, Du, & Chan, 2016; Zhang, Liu,
Xie, & Ye, 2017). Based on statistics (not reported here), during the
anti-corruption period, the average ETC falls to 0.005, a 17% decline
over the mean of 0.006 ETC for the full sample (shown in Table 2).
The anti-corruption, if effective, also serves as an external shock
approach that reduces corruption inevitably weakens the effect of RENT
on earnings management. So, using this external shock approach to test
the effect of RENT on earnings management is appealing.
The external shock of this anti-corruption campaign increases the
cost of government officials who take bribes, as well as the cost of the
connection between those corrupt officials and rent-seeking firms.
Under such circumstances, the benefits from rent-seeking activities
decrease and the need for executive to self-preserve is less. The in-
centives for engaging in rent-seeking also decrease, weakening the ef-
fect of RENT on earnings management in 2013–2014 during the anti-
Following Lin et al. (2016) and Zhang (2018), we divide our sample
period into the pre-campaign period (ANTICOR = 0, 2009 to 2012) and
the anti-corruption campaign period (ANTICOR = 1, 2013 and 2014).
The results are in Table 6. Columns 1, 3, and 5 show that, after the anti-
corruption campaign began, RENT is not statistically significantly as-
sociated with all three earnings management measures. In contrast, the
results in columns 2, 4, and 6 show that, before the anti-corruption
campaign, the higher the RENT, the higher the earnings management,
and the coefficients are significant at the 1% level. In general, Table 6
offers evidence that after the anti-corruption campaign, firms engage in
less rent-seeking, so the positive correlation between rent-seeking and
earnings management is weaker during the post-campaign period. The
findings in Table 6 support H2. They also suggest that the support for
H1 is not due to endogeneity. The external shock approach alleviates
these concerns.
4.3. The moderating role of state ownership and political connections
We split the sample into SOE and non-SOE subsamples and run Eq.
(12) to show the moderating effect of state ownership on the relation
between RENT and earnings management measures. The results, in
Table 7, show that at SOEs, RENT does not influence earnings man-
agement, while at non-SOEs, an increase in RENT leads to a higher level
of earnings management. These results support H3. In addition,

103
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Table 9 RENT and EM is robust with respect to alternative measures of earnings

Robustness tests: alternative measures of RENT and discretionary accruals and management and RENT measures.
using reduced sample. Second, in Table 5, the sample size is smaller (N = 7598) for the
(1) (2) (3) regression equation of ABS_DA_FLJ. For robustness, we examine the
regression equations for ABS_DA_MJ and ABS_DA_K with the same re-
ABS_DA_MJ ABS_DA_FLJ ABS_DA_K duced sample. The results in Panel B of Table 9 show that the coeffi-
cients of RENT continue to be positive and significant at the 1% level.
Panel A: Alternative measure of rent-seeking
LN (RENT per capita) 0.002⁎⁎ −0.002 0.002⁎⁎ Our hypothesis development is based on the conceptual notion that
(2.09) (−0.74) (2.12) firms hide their benefits from rent-seeking activities. The arguments
Control variables Yes Yes Yes concerning both hiding benefits and political extraction risk imply that
Intercept 0.183⁎⁎⁎ −0.426⁎⁎⁎ 0.118⁎⁎⁎
firms try to manipulate their earnings downward. That is, earnings
(6.03) (−3.31) (4.08)
Industry Yes Yes Yes
management deterioration is more pronounced in negative DAs than in
Year Yes Yes Yes positive DAs. To validate this conjecture, we divide earnings manage-
N 8610 8610 7598 ment into positive and negative DAs and examine H1. The findings are
Adj. R2 0.074 0.204 0.057 in Appendix B. As seen in columns 2, 4, and 6, the coefficients of RENT
Panel B: Using N = 7598 for negative DAs are negative and significant at the 1% or 5% levels.
RENT 1.725⁎⁎⁎ 1.418⁎⁎⁎ 1.766⁎⁎⁎ The same coefficients for positive DAs are not significant in columns 1
(3.76) (3.35) (4.04)
and 3 and positive and significant at the 10% level in column 5. That is,
Control variables Yes Yes Yes
Intercept −0.427⁎⁎⁎ −0.441⁎⁎⁎ −0.393⁎⁎⁎ when a firm engages in a high level of rent-seeking, its negative DA is
(−3.07) (−3.43) (−2.95) even smaller (it manages earnings downward). The findings in
Firm Yes Yes Yes Appendix B are consistent with our conjecture and support our con-
Industry Yes Yes Yes ceptual underpinnings.
Year Yes Yes Yes
Adj. R2 0.188 0.168 0.172

Panel A of this table presents the results of the impact rent-seeking has on
earnings management using natural logarithm of 1 plus RENT per capita to
measure rent-seeking activity. Panel B uses N = 7598 as the sample size, along
the sample size in ABS_DA_FLJ. All variables are defined in Table 1. t statistics in
parentheses are adjusted for heteroskedasticity. *, **, and *** indicate statis-
tical significance at the 10%, 5%, and 1% levels, respectively.
ANTICOR is more negative and significant at SOEs than at non-SOEs.
This result implies that SOEs are more strongly driven to reduce earn-
ings management during the imposition of anti-corruption measures,
which is to be expected as these efforts primarily target government
officials who work at SOEs.
To examine H3 further, we delete the SOEs and partition the non-
SOE sample into PC firms and non-PC firms. We define a firm as poli-
tically connected if the chairperson of the board or CEO has a political
background, namely, political experience or a political identity. Table 8
presents the regression results of Eq. (12) using the subsamples above.
The results in columns 1, 3, and 5 show that in PC firms, RENT does not
influence EM. Columns 2, 4, and 6 show that at non-PC firms, an in-
crease in RENT significantly leads to increasing EM. These findings
support H3. In addition, ANTICOR has a negative and significant effect
on the earnings management at PC firms in Model 1 while the negative
effect is insignificant for other models in the period of the anti-cor-
ruption campaign. The negative effect is reasonable because the anti-
corruption efforts that threaten corrupt government officials are in
force during this period.
4.4. Robustness tests
We conduct two robustness tests. First, in previous results, we
normalize RENT by total revenue as the proxy for rent-seeking. For
robustness, we use the logarithm value of 1 plus RENT per capita (Cai
et al., 2011; Wei, Dong, & Jin, 2015) in Eq. (12) as the explanatory
variable. The results are in Panel A of Table 9. For the sake of brevity,
we do not present the coefficients of the other control variables. The
coefficients of Ln (RENT per capita) are positive and significant at the
5% level in columns 1 and 3. Overall, the positive correlation between
5. Conclusions
We use information on entertainment and travel costs (ETC) re-
ported in Chinese firms' financial statements to develop a proxy for
rent-seeking. Because ETC contains normal business-related expenses,
we first develop a predictive model of ETC and then purge the business-
related component of ETC to develop a proxy for rent-seeking, which is
used to investigate the effect of rent-seeking on earnings management.
We document that rent-seeking leads to higher earnings management
(in terms of discretionary accruals), implying that rent-seeking moti-
vates earnings management. Our findings are robust to different mea-
sures of rent-seeking and earnings management and different estima-
tion methods. After the strict anti-corruption campaign that
commenced at the end of 2012, the rent-seeking behavior of firms de-
clined, leading to the insignificant effect of rent-seeking on earnings
management. This result further confirms the predicted relation be-
tween rent-seeking and earnings management.
Our analyses also show that firms that are neither state owned nor
politically connected experience a stronger expected effect of rent-
seeking on earnings management than SOEs because non-SOEs, which
lack resources and depend on government help, are more willing to
engage in rent-seeking. Our results also call for reexamination of the
accounting policy that allows entertainment and travel expenses in
China to be tax deductible. They are excessive when compared to those
in the US because actual entertainment expenses in China are for tax
deductible while certain qualified entertainment expenses in the US
firms are allowed for tax purpose.
Our study offers robust evidence that rent-seeking in China imposes
earnings management costs. We suggest continued future research on
rent-seeking in other emerging markets to identify its positive and ne-
gative impacts on the economy and operating results of rent-seeking
firms.
Declarations of interest
None.

104
B. Liu et al. Journal of Business Research 91 (2018) 94–107

Appendix A. Rent-seeking level after accounting for normal business and agency problem-related ETC

We run a regression on a set of firm and provincial characteristics to account for normal business-related ETC. The definitions of all variables are
in Table 1. The residual of the regression equation captures the ETC expense beyond normal business needs and executives' agency problems. The
residuals are the firm's rent-seeking expense (RENT). *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.

Dependent variable = ETC

CGI −0.003⁎⁎⁎
(−3.86)
LAWJUST 0.003⁎⁎⁎
(2.74)
SIZE −0.001⁎⁎⁎
(−12.74)
LEVER 0.001⁎⁎⁎
(2.73)
MTB 0.001⁎⁎⁎
(21.92)
ROA −0.014⁎⁎⁎
(−13.05)
TOP1 −0.005⁎⁎⁎
(−10.08)
INSIDER −0.001⁎⁎
(−2.12)
BOARDSIZE 0.001⁎⁎⁎
(3.05)
INDDIR 0.000
(0.78)
BIG4 −0.000
(−0.93)
SOE −0.001⁎⁎⁎
(−4.48)
PC −0.001⁎⁎⁎
(−3.97)
ANTICOR 0.001⁎⁎⁎
(4.13)
Intercept 0.029⁎⁎⁎
(12.73)
N 8614
Adj. R2 0.258

Appendix B. The impact of rent-seeking on earnings quality using positive and negative abnormal accruals as earnings management

This table presents the results of the impact that rent-seeking has on positive and negative abnormal accruals. All variables are defined in Table 1.
t statistics in parentheses are adjusted for heteroskedasticity. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, re-
spectively.

(1) (2) (3) (4) (5) (6)

DA_MJ DA_FLJ DA_K

>0 <0 >0 <0 >0 <0

⁎⁎⁎ ⁎⁎ ⁎
RENT 0.714 −1.981 1.022 −1.517 1.094 −1.923⁎⁎⁎
(1.05) (−2.69) (1.23) (−2.29) (1.68) (−3.11)
BIG4 0.009 0.014 0.001 0.005 0.009 0.006
(0.97) (1.45) (0.12) (0.61) (0.90) (0.78)
SIZE 0.010 −0.047⁎⁎⁎ 0.009 −0.045⁎⁎⁎ 0.005 −0.040⁎⁎⁎
(1.17) (−4.28) (0.95) (−4.55) (0.60) (−4.24)
LEVER 0.125⁎⁎⁎ 0.038 0.097⁎⁎⁎ 0.055⁎⁎ 0.091⁎⁎⁎ 0.037
(5.04) (1.30) (3.51) (1.98) (3.55) (1.50)
MTB 0.001 −0.003 0.002 −0.000 0.001 −0.001
(0.53) (−1.17) (0.75) (−0.05) (0.27) (−0.48)
ROA 0.503⁎⁎⁎ −0.055 0.467⁎⁎⁎ 0.003 0.505⁎⁎⁎ −0.065

105
B. Liu et al. Journal of Business Research 91 (2018) 94–107

(9.92) (−1.11) (9.78) (0.06) (9.96) (−1.50)

LOSS 0.004 −0.012⁎ 0.013 −0.014⁎⁎ 0.010 −0.004
(0.49) (−1.71) (1.55) (−2.13) (1.21) (−0.64)
EISSUE 0.001 −0.005 −0.002 −0.009⁎⁎ −0.000 −0.007⁎⁎
(0.29) (−1.30) (−0.49) (−2.21) (−0.03) (−2.02)
TOP1 −0.134⁎⁎ 0.014 −0.105⁎ 0.057 −0.101⁎ 0.054
(−2.45) (0.27) (−1.75) (1.22) (−1.83) (1.24)
INSIDER −0.005 0.031 0.033 −0.009 0.002 0.017
(−0.14) (0.90) (0.81) (−0.22) (0.06) (0.62)
BOARDSIZE 0.015 −0.019 0.016 −0.010 0.036⁎⁎ −0.017
(1.01) (−1.14) (0.95) (−0.65) (2.19) (−1.21)
INDDIR −0.002 −0.003 −0.000 −0.002 −0.001 −0.003
(−0.34) (−0.53) (−0.03) (−0.38) (−0.16) (−0.56)
SOE −0.011 0.014 −0.003 0.007 −0.005 0.011
(−1.08) (1.31) (−0.25) (0.63) (−0.46) (1.14)
PC 0.001 0.004 −0.008 0.004 −0.001 0.004
(0.07) (0.40) (−0.80) (0.46) (−0.06) (0.49)
ANTICOR 0.004 0.036⁎⁎⁎ 0.012 0.033⁎⁎⁎ 0.009 0.033⁎⁎⁎
(0.46) (3.71) (1.29) (3.72) (0.93) (3.84)
Intercept −0.131 1.028⁎⁎⁎ −0.155 0.934⁎⁎⁎ −0.123 0.777⁎⁎⁎
(−0.69) (4.09) (−0.72) (4.21) (−0.65) (3.63)
Firm Yes Yes Yes Yes Yes Yes
Industry Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes
N 4604 4006 3733 3865 4201 4409
Adj. R2 0.244 0.164 0.203 0.179 0.214 0.182

References
Aharony, J., Lin, C., & Loeb, M. P. (1993). Initial public offerings, accounting choices, and
earnings management. Contemporary Accounting Research, 10, 61–82.
Aidt, T., Dutta, J., & Sena, V. (2008). Governance regimes, corruption and growth: Theory
and evidence. Journal of Comparative Economics, 36(2), 195–220.
Aidt, T. S. (2003). Economic analysis of corruption: A survey. Economic Journal, 113(491),
632–652.
Asiedu, E., & Freeman, J. (2009). The effect of corruption on investment growth:
Evidence from firms in Latin America, Sub-Saharan Africa, and transition countries.
Review of Development Economics, 13(2), 200–214.
Ayyagari, M., Demirgüç-Kunt, A., & Maksimovic, V. (2010). Formal versus informal fi-
nance: Evidence from China. Review of Financial Studies, 23(8), 3048–3097.
Badertscher, B. A. (2011). Overvaluation and the choice of alternative earnings man-
agement mechanisms. Accounting Review, 86, 1491–1518.
Bardhan, P. (1997). Corruption and development: A review of issues. Journal of Economic
Literature, 35, 1320–1346.
Beck, P. J., & Maher, M. W. (1986). A comparison of bribery and bidding in thin markets.
Economics Letters, 20(1), 1–5.
Cai, H., Fang, H., & Xu, L. C. (2011). Journal of Law and Economics, 54(1), 55–78.
Caprio, L., Faccio, M., & McConnell, J. J. (2013). Sheltering corporate assets from poli-
tical extraction. Journal of Law, Economics, and Organization, 29(2), 332–354.
Chen, H., Chen, J. Z., Lobo, G. J., & Wang, Y. (2011). Effects of audit quality on earnings
management and cost of equity capital: Evidence from China. Contemporary
Accounting Research, 28(3), 892–925.
Chen, S., Thomas, J., & Zhang, F. (2016). Spring-loading future performance when no one
is looking? Earnings and cash flow management around acquisitions. Review of
Accounting Studies, 21, 1081–1115.
Chen, Y., Liu, M., & Su, J. (2013). Greasing the wheels of bank lending: Evidence from
private firms in China. Journal of Banking & Finance, 37(7), 2533–2545.
De Rosa, D., Gooroochurn, N., & Görg, H. (2010). Corruption and productivity: Firm-level
evidence from the BEEPS survey. World Bank policy research working paper series.
Dechow, P., Ge, W., & Schrand, C. (2010). Understanding earnings quality: A review of
the proxies, their determinants and their consequences. Journal of Accounting and
Economics, 50, 344–401.
Dechow, P. M., Richardson, S. A., & Tuna, I. (2003). Why are earnings kinky? An ex-
amination of the earnings management explanation. Review of Accounting Studies,
8(2), 355–384.
Dechow, P. M., & Schrand, C. (2004). Earnings quality. Charlottesville, Virginia: CFA
Institute Research Foundation.
Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1995). Detecting earnings management.
Accounting Review, 70(2), 193–225.
Dong, B., & Torgler, B. (2010). The consequences of corruption: Evidences from China.
SSRN working paper.
Durnev, A., & Fauver, L. (2011). Stealing from thieves: Expropriation risk, firm govern-
ance, and performance. Paper presented at the 2nd annual conference on empirical legal
studies.
Easley, D., & O'Hare, M. (2004). Information and the cost of capital. Journal of Finance,
59(4), 1553–1583.
Fan, J. P., Rui, O. M., & Zhao, M. (2008). Public governance and corporate finance:
Evidence from corruption cases. Journal of Comparative Economics, 36(3), 343–364.
Fan, J. P., Titman, S., & Twite, G. (2012). An international comparison of capital structure
and debt maturity choices. Journal of Financial and Quantitative Analysis, 47(1),
23–56.
Fisman, R., & Svensson, J. (2007). Are corruption and taxation really harmful to growth?
Firm level evidence. Journal of Development Economics, 83, 63–75.
Healy, P. M., & Palepu, K. G. (2001). Information asymmetry, corporate disclosure, and
the capital markets: A review of the empirical disclosure literature. Journal of
Accounting and Economics, 31, 405–440.
Huang, J., & Li, K. (2013). Eat-drink, corruption and firm's purchase order. Economic
Research Journal, 48(6), 71–84 (in Chinese).
Hutton, A. P., Marcus, A. J., & Tehranian, H. (2009). Opaque financial reports, R2, and
crash risk. Journal of Financial Economics, 94(1), 67–86.
Jia, M., & Zhang, Z. (2016). How long does the influence of organizational deviance have
on innocent firms? Journal of Business Research, 69(8), 2649–2663.
Jones, J. J. (1991). Earnings management during import relief investigations. Journal of
Accounting Research, 29(2), 193–228.
Karni, K., & Schmeidler, D. (1986). Self-preservation as a foundation of rational behavior
under risk. Journal of Economic Behavior and Organization, 7(1), 71–81.
Kassamany, T., Ibrahim, S., & Archbold, S. (2017). Accrual and real-based earnings
management by UK acquirers. Journal of Accounting and Organizational Change, 13,
492–519.
Khwaja, A., & Mian, A. (2005). Do lenders favor politically connected firms? Rent pro-
vision in an emerging financial market. Quarterly Journal of Economics, 120(4),
1371–1411.
Kim, J. B., & Sohn, B. C. (2013). Real earnings management and cost of capital. Journal of
Accounting and Public Policy, 32, 518–543.
Kothari, S. P., Leone, A. J., & Wasley, C. E. (2005). Performance matched discretionary
accrual measures. Journal of Accounting and Economics, 39(1), 163–197.
Kusnadi, Y., Yang, Z., & Zhou, Y. (2015). Institutional development, state ownership, and
corporate cash holdings: Evidence from China. Journal of Business Research, 68(2),
351–359.
La Porta, R., Lopez-de-Silance, R., Shleifer, A., & Vishny, R. W. (1998). Law and finance.
Journal of Political Economy, 106(6), 1113–1150.
Lambert, R., Leuz, C., & Verrecchia, R. (2007). Accounting information, disclosure, and
the cost of capital. Journal of Accounting Research, 45, 385–420.
Li, J., & Huang, Y. (2010). Corruption and firm growth in transitional economics. China
Economic Quarterly, 9(4), 1467–1487 (in Chinese).
Li, W., & Chi, Q. (2015). Research on the mechanism of the effect of executive perquisite
consumption on firm performance: Perspective on ownership type. China Industrial
Economics, 4, 122–134 (in Chinese).
Lin, C., Morck, R., Yeung, B., & Zhao, X. (2016). Anti-corruption reforms and shareholder
valuations: Event study evidence from China. National Bureau of Economic Research
working paper.
Liu, Q., & Lu, Z. (2007). Corporate governance and earnings management in the Chinese

106
B. Liu et al.
listed companies: A tunneling perspective. Journal of Corporate Finance, 13, 881–906.
Liu, Q., Tang, J., & Tan, G. G. (2013). Does political capital create value in the IPO
market? Evidence from China. Journal of Corporate Finance, 23, 395–413.
Lui, F. T. (1985). An equilibrium queuing model of bribery. Journal of Political Economy,
93(4), 760–781.
Mauro, P. (1995). Corruption and growth. Quarterly Journal of Economics, 110(3),
681–712.
Murphy, K. M., Shleifer, A., & Vishny, R. W. (1993). Why is rent-seeking so costly to
growth? American Economic Review, 83(2), 409–414.
Nguyen, T. T., & Van Dijk, M. A. (2012). Corruption, growth, and governance: Private vs.
state-owned firms in Vietnam. Journal of Banking & Finance, 36(11), 2935–2948.
Piotroski, J. D., & Zhang, T. (2014). Politicians and the IPO decision: The impact of im-
pending political promotions on IPO activity in China. Journal of Financial Economics,
111, 111–136.
Qiao, P., Fung, A., Miao, J., & Fung, H. G. (2017). Powerful CEOs and firm performance:
Integrating agency and stewardship theory. China & World Economy, 25(6), 100–119.
Roychowdhury, S. (2006). Earnings management through real activities manipulation.
Journal of Accounting and Economics, 42, 335–370.
Smith, J. D. (2016). US political corruption and firm financial policies. Journal of Financial
Economics, 121(2), 350–367.
Sun, F. (2016). How to manage client entertainment in China. Business Horizons, 59,
401–410.
Svensson, J. (2003). Who must pay bribes and how much? Evidence from a cross section
of firms. Quarterly Journal of Economics, 118(1), 207–230.
Tang, X., Lin, Y., Peng, Q., Du, J., & Chan, K. C. (2016). Politically connected directors
and firm value: Evidence from forced resignations in China. North American Journal of
Economics and Finance, 37, 148–167.
Teoh, S. H., Welch, I., & Wong, T. J. (1998). Earnings management and the long-run
market performance of initial public offerings. Journal of Finance, 53, 1935–1974.
Wang, X., Yu, J., & Fan, G. (2015). Business environment index for China's provinces 2013
report. Beijing: City Press.
Wang, Y., & You, J. (2012). Corruption and firm growth: Evidence from China. China
Economic Review, 23(2), 415–433.
107
Journal of Business Research 91 (2018) 94–107
Wei, X., Dong, Z., & Jin, Z. (2015). Corruption and corporate vitality: An empirical re-
search on the effect rent-seeking and rent-extracting on operating rate. Journal of
World Economy, 1, 105–125 (in Chinese).
Xu, N., Li, X., Yuan, Q., & Chan, K. C. (2014). Excess perks and stock price crash risk:
Evidence from China. Journal of Corporate Finance, 25, 419–434.
Zhang, J. (2018). Public governance and corporate fraud: Evidence from the recent anti-
corruption campaign in China. Journal of Business Ethics, 148(2), 375–396.
Zhang, M., Liu, Y., Xie, L., & Ye, T. (2017). Does the cutoff of “red capital” raise a red flag?
Political connections and stock price crash risk. North American Journal of Economics
and Finance, 39, 89–109.
Baohua Liu is an assistant professor of accounting at Southwest Jiaotong University,
Chengdu, China. He received his Ph.D. in accounting at Southwestern University of
Finance and Economics. His research interests include executive compensation, corporate
governance, and financial reporting.
Yan Lin is an assistant professor of accounting at Yunnan University of Finance and
Economics, Kunming, China. She received his Ph.D. in accounting at Southwestern
University of Finance and Economics. Her research interests include directors, corporate
governance, and financial reporting.
Kam C. Chan is the Leon Page Endowed Chair of Finance and a University Distinguished
Professor at Western Kentucky University, Bowling Green, USA. He has published > 190
articles. He received his Ph.D. from the University of Alabama and holds a CFA charter.
His research interests include corporate governance, analyst behavior, and political
connections.
Hung-Gay Fung is the Dr. Y.S. Tsiang Professor in Chinese Studies and Curators'
Professor of Finance at the University of Missouri at St. Louis, St. Louis, USA. He received
his Ph.D. from Georgia State University and has published > 200 articles. His research
interests include corporate finance and Chinese financial markets.