Content uploaded by Kung-Cheng Ho
Author content
All content in this area was uploaded by Kung-Cheng Ho on Mar 12, 2019
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=raae20
Asia-Pacific Journal of Accounting & Economics
ISSN: 1608-1625 (Print) 2164-2257 (Online) Journal homepage: http://www.tandfonline.com/loi/raae20
Corporate social responsibility and managerial
short-termism
Yujing Gong & Kung-Cheng Ho
To cite this article: Yujing Gong & Kung-Cheng Ho (2018): Corporate social responsibility
and managerial short-termism, Asia-Pacific Journal of Accounting & Economics, DOI:
10.1080/16081625.2018.1540941
To link to this article: https://doi.org/10.1080/16081625.2018.1540941
Published online: 08 Nov 2018.
Submit your article to this journal
Article views: 12
View Crossmark data
Corporate social responsibility and managerial short-termism
Yujing Gong
a
and Kung-Cheng Ho
b
a
Wenlan School of Business, Zhongnan University of Economics and Law, Wuhan, China;
b
School of Finance,
Zhongnan University of Economics and Law, Wuhan, China
ABSTRACT
This study examines whether corporate social responsibility (CSR) miti-
gates or exacerbates managerial short-termism. Using earnings manage-
ment (EM) as a proxy for managerial short-termism, we find that socially
responsible firms engage in less accrual-based and real EM activities. This
finding supports the argument that socially responsible firms not only
focus on current profits but also maintain long-term sustainable devel-
opment. Further analyses reveal that the moderate effect of CSR on
managerial short-termism is evident only for mandatory CSR disclosure
firms. This suggests that regulatory non-financial reporting potentially
mitigates information asymmetry and effectively constrains managerial
short-termism.
ARTICLE HISTORY
Received 15 December 2017
Accepted 16 October 2018
KEYWORDS
Corporate social
responsibility; mandatory
disclosure; managerial short-
termism; earnings
management
1. Introduction
Managers should ideally secure a firm’s long-term value while achieving short-term goals.
However, both researchers and practitioners find that managers tend to make decisions yielding
short-term gains at the expense of long-term growth, thereby foregoing valuable investments and
destroying shareholders’long-term values (Narayanan 1985; Stein 1989; Graham, Harvey, and
Rajgopal 2005; Roychowdhury 2006; Davies et al. 2014). Studies mainly focus on the role of
external factors, such as enhancing monitoring and granting equity incentives to managers, in
alleviating managerial short-termism (Bushee 1998; Bhojraj et al. 2009; Brochet, Loumioti, and
Serafeim 2015). Therefore, our study investigates the role of internal factors (i.e., whether
managers are socially responsible and ethical) in influencing managerial short-termism.
Despite its perceived importance, empirically examining managerial short-termism is difficult.
Numerous studies investigate managerial myopic behaviors through intangible asset investments
(Baber, Fairfield, and Haggard 1991; Dechow and Sloan 1991; Bushee 1998; Graham, Harvey, and
Rajgopal 2005), whereas some recent studies focus on managerial actions that manipulate
accounting earnings to meet short-term expectations (Call et al. 2014; Chen et al. 2015).
Specifically, Chen et al. (2015) argue that earnings management (EM) captures the essence of
managerial myopia and is a natural setting for managerial short-termism. Therefore, we employ
EM as a proxy for managerial short-termism.
A debate regarding the impact of corporate social responsibility (CSR) performance on EM is
ongoing; however, most participants of this debate have investigated the problem in developed
countries, particularly the United States. The aim of our study is to extend this framework to
China, because the earnings quality of Chinese firms is lower than that in developed markets
(Allen, Qian, and Qian 2005). The concept of social responsibility has developed in the United
CONTACT Kung-Cheng Ho kcho731101@163.com School of Finance, Zhongnan University of Economics and Law,
China
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS
https://doi.org/10.1080/16081625.2018.1540941
© 2018 City University of Hong Kong and National Taiwan University
States for many decades; most large firms voluntarily engage in CSR activities and report their
CSR performance accordingly (Dhaliwal et al. 2011). By contrast, the Chinese government began
encouraging firms to embrace CSR only in 2001 (Moon and Shen 2010). Compared with
developed markets, the Chinese stock market is characterized by weak investor protection and
poor corporate governance. Because of these unique institutional settings, we investigate how
Chinese firms perform with EM in association with their CSR performance.
In December 2008, the Shanghai Stock Exchange (SHSE) and Shenzhen Stock Exchange (SZSE)
mandated a subset of Chinese listed firms to annually disclose their CSR reports. In addition, an
increasing number of firms voluntarily disclose their CSR reports. Unlike relevant studies based on
voluntary CSR disclosure, the aforementioned 2008 CSR regulation provides us the opportunity to
differentiate our sample into voluntary and mandatory CSR disclosure firms. Therefore, we examine
how the willingness of CSR disclosure affects the association between CSR performance and EM.
Using the unique CSR performance scores from Rankins CSR Ratings (RKS) during the period
2009–2015, we find that firms with superior social performance engage in less accrual-based EM
(AEM), evidenced by their lower accruals quality (AQ). Furthermore, socially responsible firms
engage in less real EM (REM), evidenced by overproduction and opportunistic reductions in their
discretionary expenses. The potential mechanism underlying this is that CSR can facilitate man-
agers’forward thinking and enhance firms’information transparency (Orlitzky, Schmidt, and Rynes
2003; Porter and Kramer 2006; Kim, Park, and Wier 2012; Wang, Cao, and Ye 2016). Thus, socially
responsible firms not only focus on short-term profits but also aim to maintain long-term sustain-
able development (Gelb and Strawser 2001; Chih, Shen, and Kang 2007; Choi and Pae 2011).
1
Next, we separate the Chinese sample into two subsamples –voluntary and mandatory CSR
disclosure firms –and find that only firms with mandatory CSR disclosure show negative
association between CSR performance and EM. Therefore, regulators’mandatory action for
nonfinancial reporting potentially mitigates information asymmetry and constrains managerial
short-termism activities. However, the moderating effect of CSR performance on EM does not
function if firms voluntarily disclose their CSR reports. Thus, without powerful supervision,
managers may strategically use CSR to pursue their private benefits, without considering the
long-term sustainable development of their firm.
These results are robust to three alternative measures of CSR performance, and hold after
controlling for potential determinants of EM. Moreover, we also confirm the causal effect of CSR
performance on the intensity of EM when the endogenous effect is considered. In addition,
because large firms and state-owned entrepreneurs (SOEs) are subject to mandatory disclosure
and less likely to engage in EM, we examine the extent to which our results hold outside of large
firms or SOEs. We find that the negative relationship between CSR performance and EM holds
even when large firms or SOEs are excluded from the sample (see Appendix B). Overall, our
findings support the view that socially responsible firms engage in less EM, indicating less
managerial short-termism, particularly when their CSR activities are under strict supervision.
Our paper makes several contributions to the literature. First, unlike the studies that primarily
focus on voluntary CSR disclosure samples, our paper examines the impact of CSR performance
on managerial short-termism for voluntary and mandatory CSR disclosure firms. We find that
socially responsible firms in China engage in less EM activities; however, this is only true for
mandatory CSR disclosure firms. This increases the understanding of the economic consequences
of CSR performance in emerging economies. In addition, Wang, Cao, and Ye (2016) examines a
mandatory CSR subsample in China. Using the 2008 mandatory CSR requirement as a quasi-
natural experiment, the authors find that mandatory CSR disclosure firms constrained their EM
‘after’2008 (the CSR disclosure effect). By contrast, using the unique CSR scores from RKS, we
investigate the impact of CSR performance on EM for mandatory and voluntary CSR disclosure
firms, respectively (i.e., the CSR performance effect). Wang, Cao, and Ye (2016) primarily focus
on the CSR ‘disclosure’effect and ignore the CSR ‘performance’effect. Thus, our study
2Y. GONG AND K.-C. HO
complements that of Wang, Cao, and Ye (2016) based on the investigation the moderating impact
of CSR performance on EM, which is particularly strong for mandatory CSR disclosure firms.
The remainder of this paper is organized as follows. Section 2 describes the background and
development of CSR in China and develops hypotheses. Section 3 presents our data and discusses
our research designs. Section 4 reports the empirical results. Section 5 concludes the paper.
2. Background and hypotheses development
2.1. CSR in china
The Chinese government has recognized the importance of balancing economic growth with
social responsibility development. In particular, the 2008 milk powder scandal evoked a public
outcry concerning firms’CSR. The Social Science Academy of China reported that most Chinese
investors are concerned about firms’CSR activities. In 2007, the State-owned Assets Supervision
and Administration Commission released the Guidelines on the CSR Disclosure of Central
Government Enterprises, and in January 2009, the China Banking Association issued the
Guidelines on the Corporate Social Responsibilities of Banking Institutions of China.
Furthermore, the two main stock exchanges in China have disseminated CSR guidelines to the
listed firms. In September 2006, the SZSE announced the Guide on Listed Companies’Social
Responsibilities, encouraging firms to prepare and voluntarily disclose their CSR reports. Next, in
May 2008, the SHSE issued two guidelines and strengthened the regulations for disclosing social
responsibility information. In December 2008, the SHSE and SZSE jointly issued the Notice for
Better Preparing 2008 Annual Reports. Accordingly, the SZSE mandated that firms included in the
SZSE 100 index disclose CSR reports attached to their annual reports. Three types of firms listed
on the SHSE should provide mandatory CSR reports: those in the SHSE Corporate Governance
Section Index, in the financial sector, and with listed overseas shares. In addition, the SHSE
launched a CSR index in 2009.
Moreover, some firms voluntarily disclose their CSR performance. Compared with mandatory
CSR reporting, such voluntary disclosure is less comprehensive and contains vaguer information
(Dhaliwal et al. 2011). Wang, Cao, and Ye (2016) further argue that managers voluntarily disclose
CSR reports because of their strategic considerations. In 2015, 408 mandated reports and 293
voluntary reports were submitted. The numerous observations aid in distinguishing the motives of
mandatory and voluntary CSR disclosure and examining their impacts on managerial short-
termism separately.
2.2. Hypotheses development
The debate on the impact of CSR on managerial short-termism remains ongoing. One view is
based on the long-term perspective, stating that CSR can facilitate managers’forward thinking
(Orlitzky, Schmidt, and Rynes 2003; Porter and Kramer 2006). Firms with superior CSR perfor-
mance not only aim to increase short-term profits but also focus on long-term sustainable
development (Gelb and Strawser 2001; Chih, Shen, and Kang 2007; Choi and Pae 2011).
Socially responsible firms are less likely to manipulate real activities to boost short-term account-
ing earnings; information in accounting reports provided by such firms is more reliable and
transparent (Kim, Park, and Wier 2012; Wang, Cao, and Ye 2016; Yang et al. 2018). Moreover,
socially responsible firms are found to be more ethical, trustworthy, and honest (Donaldson and
Preston 1995; Jones 1995; Phillips, Freeman, and Wicks 2003; Ho et al. 2016), because their
managers are encouraged to undertake actions that boost the firms’long-term value, thereby
resulting in less managerial short-termism.
Another view is based on the opportunism perspective, stating that managers may use CSR as a
strategic tool to disguise their opportunistic behaviors, such as obtaining higher compensation
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 3
(Bergstresser and Philippon 2004; Cheng and Warfield 2005) and avoiding scrutiny from stake-
holder activists (Cespa and Cestone 2007; Ferrero, Banerjee, and Sanchez 2016). In pursuit of
private benefits, managers may engage in CSR activities by distorting earnings information to
protect their entrenchment (Prior, Surroca, and Tribo 2008; Bozzolan et al. 2015; Gao and Zhang
2015). Accordingly, firms with higher CSR scores engage in earnings aggressiveness (Chih, Shen,
and Kang 2007; Prior, Surroca, and Tribo 2008; Liu, Shi, and Wu 2017). Given the conflicting
results of relevant research, additional studies are warranted to investigate whether the engage-
ment in CSR encourages or impedes managerial short-termism. To test these opposite views, we
propose the following competing hypotheses:
Hypothesis 1a. Based on the long-term perspective, a negative association exists between CSR
performance and EM.
Hypothesis 1b. Based on the opportunism perspective, a positive association exists between CSR
performance and EM.
The 2008 mandatory CSR disclosure regulation indicates that mandatory CSR disclosure firms
are under the scrutiny and strict supervision of the Chinese regulatory authorities. In addition,
these firms are more likely to attract investors’attention, and CSR reporting tends to increase
their exposure to public attention (Wang, Cao, and Ye 2016). Both the regulatory authorities and
outsiders serve as active monitors of firms’operations, which can prevent managers from
pursuing their self-interests by sacrificing the firm’s value. In addition, opportunistic behaviors
by managers are discouraged if an effective monitoring mechanism exists (Choi, Lee, and Park
2013). Therefore, compared with voluntary CSR firms, mandatory CSR firms under strict super-
vision potentially discourage managers from using CSR as a strategic tool, but encourage them to
make business decisions based on a long-term perspective. Accordingly, we propose our second
hypothesis.
Hypothesis 2. The negative association between CSR performance and EM is stronger for mandatory
CSR disclosure firms.
3. Data and research design
3.1. Data and sample
We obtain financial-related data from the China Stock Market Financial Database –Statement
Notes of China Stock Market and Accounting Research (CSMAR). Our CSR data are from RKS.
RKS, a third party institution, ensures the reliability of CSR disclosure and performance (Cohen
et al. 2011; Marquis and Qian 2014). The sample period begins in 2009, when the CSR rating score
is first available, and ends in 2015. Our initial sample includes all common A-shares listed on the
SHSE and SZSE over the sample period of 2009–2015. We apply two filtering criteria for our
initial sample. First, each firm is required to possess a CSR report. Second, we drop observations
lacking enough information to measure EM. Our final sample consists of 3,409 firm-year
observations.
3.2. Measuring EM
We employ AQ, based on a modification of Dechow and Dichev’s (2002) model, as the measure of
AEM. Specifically, the modified model is
4Y. GONG AND K.-C. HO
CAcci;t
ATi;t1
¼;
0þ;
1
CFOi;t1
ATi;t1
þ;
2
CFOi;t
ATi;t1
þ;
3
CFOi;tþ1
ATi;t1
þ;
4
ΔRevi;t
ATi;t1
þ;
5
PPEi;t
ATi;t1
þεi;t(1)
where CAcc is current accruals or the change in working capital; CFO is cash flows from
operations; ΔRev is the change in revenues; PPE is the property, plant, and equipment; AT is
the book value of total assets; εis the error term; and AQ
i
is the standard deviation of firm i’s
residuals. Larger the AQ, more the AEM is.
In contrast to AEM, which does not affect firms’cash flows, REM is performed by firms
altering their actual business activities to achieve their desired financial goals (Roychowdhury
2006). Firms tend to engage in both of AEM and REM (Bushee 1998; Roychowdhury 2006; Cohen
and Zarowin 2010; Chen et al. 2015). In particular, the effect of REM on firms’long-term
development is severer. Therefore, we also examine the behavior of socially responsible firms
engaging in REM.
According to relevant studies, we construct three measures of REM (Roychowdhury 2006;
Cohen and Zarowin 2010; Chen et al. 2015). The first one is the abnormal production costs, which
is estimated by the following regression:
PRODi;t
ATi;t1
¼γ0þγ1
1
ATi;t1
þγ2
SALEi;t
ATi;t1
þγ2
ΔSALEi;t
ATi;t1
þγ3
ΔSALEi;t1
ATi;t1
þεi;t(2)
where PRODi;tis the sum of the cost of goods sold and the change in inventory from year t-1 to t
for firm iand SALEi;tis the firm i’s sales revenue.
The second one is the abnormal discretionary expenses, which is constructed by the following
model:
DISXi;t
ATi;t1
¼β0þβ1
1
ATi;t1
þβ2
SALEi;t1
ATi;t1
þεi;t(3)
where DISXi;tis the sum of R&D, advertising, selling, and general and administrative expenses for
firm iin year t.
The third one is the abnormal operating cash flows, which is estimated by the following model:
CFOi;t
ATi;t1
¼δ0þδ1
1
ATi;t1
þδ2
SALEi;t
ATi;t1
þδ2
ΔSALEi;t
ATi;t1
þεi;t(4)
where CFOi;tis firm i’s operating cash flows.
According to Chen et al. (2015), we define abnormal production costs (ABPROD), abnormal
discretionary expenses (ABDISX), and abnormal operating cash flows (ABCFO) as the residuals
from the three models, respectively. A higher value of ABPROD, a lower value of ABDISX, and a
lower value of ABCFO all indicate more REM.
We construct two comprehensive indices of REM by combining the three individual measures
to further capture the overall effect of REM (Cohen and Zarowin 2010; Chen et al. 2015). The first
composite index of REM (REM1) is the sum of ABPROD and the value of ABDISX multiplied by
negative 1, whereas the second measure of REM (REM2) is constructed by multiplying ABCFO
and ABDISX by negative 1 and adding the two items together. A higher value of REM1 or REM2
indicates more REM.
3.3. Measuring CSR
We measure firms’social performance using the CSR rating score annually estimated by RKS,
which is drawn from 153 measures over the following four dimensions: macrocosm, content,
technique, and industry index. Detailed definitions of each measure can be found in Appendix A.
RKS is a widely used measure in China for determining CSR performance (Li, Zhang, and Foo
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 5
2013; Li and Foo 2015). The original CSR rating score ranges from 1 to 100. According to Pan
et al. (2015), we considered the standardized CSR, defined as the difference between the original
CSR score and average CSR scores, scaled by the standard deviation of CSR scores, as an
alternative CSR measure.
3.4. Control variables
We include several control variables, discussed as potential determinants of EM (Bergstresser and
Philippon 2004;Zang 2012; Chen et al. 2015). Firm size (SIZE) is the natural logarithm of a firm’s
total assets. Cash flow (CF) is the operating cash flow divided by the book value of assets, and cash
flow volatility (CFV) is the standard deviation of cash flow over the past 3 years. Debt ratio (Debt)
is the ratio of total debts to total assets. LOSS is a dummy –equal to one if the operating income is
negative and zero otherwise. AZ is the Altman’s Z-score, which measures the default probability
(Altman 1968). MB is the ratio of the market value to book value of total assets. ROA is return on
assets. Detailed definition of each variable is provided in Table 1.
3.5. Descriptive statistics
In Panel A of Table 2, we describe summary statistics of variables for the whole sample. The mean
(37.56) and median (34.66) CSR values indicate that firms in our sample receive relatively low
CSR rating scores. Most Chinese firms in our sample receive lower than average CSR scores. The
mean AQ, REM1, and REM2 values are 0.06, −0.02, and −0.03, respectively. On average, firms in
our sample have a firm size of 24.14, market-to-book ratio of 1.65, and debt ratio of 0.47.
Panel B of Table 2 presents descriptive statistics of variables for mandatory and voluntary CSR
disclosure subsamples. Compared with voluntary CSR firms, on average, mandatory CSR firms
have superior CSR performance and engage in less AEM and REM. Moreover, mandatory CSR
firms are characterized by a significantly large firm size, high debt ratio, low MB ratio, and great
return on assets. Nonrandom selection of mandatory CSR firms may cause bias in our results.
Literature suggests that larger firms usually construct a more transparent information
Table 1. Variable definitions.
Variables Definition
AQ Accrual quality, measured as the uncertainty in the accrual-to-cash flows. A high AQ value implies low
accounting quality. (Dechow and Dichev 2002)
AZ Altman’s Z-score, defined as (3.3*operating income + sales + 1.4*retained earnings + 1.2*(current assets –
current liability))/total assets
Beta CAPM beta, estimated from past 60 months
CF Cash flow from operations/total assets
CFV Standard deviation of cash flow over the past three years
CSR The score of CSR rating ranging from 1 (the lowest) to 100 (the highest)
CSRs Standardized CSR score, defined as the difference between CSR and the average value of CSR scaled by the
standard deviation of CSR. (Pan et al. 2015)
DEBT Total debt divide by book value of asset
LOSS Dummy variable, where equals to 1 if operating income is negative, 0 otherwise
Mandatory Dummy variable, where equals to1 if the firms are mandated to disclose their CSR reports, 0 otherwise.
MB (Total assets –book equity + market equity)/total assets
Ranking Rating level of CSR, ranging from AA+ (19, the highest) to CCC (1, the lowest).
RankL Ranking level of CSR. In specific, firm with the highest CSR rating score is assigned a Ranking value of 1, firm
with the second highest CSR rating score is assigned a Ranking value of 2, and so on and so forth, firm with
the lowest CSR rating score is assigned a N (N is the total number of firms in any given year of our sample).
REM1 Real earnings management index 1, defined as ABPROD plus (−1) * ABDISX (Chen et al. 2015)
REM2 Real earnings management index 2, defined as (−1) * ABCFO plus (−1) * ABDISX (Chen et al. 2015)
RET Return on assets
SIZE The natural logarithm of the total assets
6Y. GONG AND K.-C. HO
environment than do smaller firms; however, the impact of mandatory CSR firms on EM should
be underestimated, rather than overestimated (Wang, Cao, and Ye 2016).
Table 3 presents the Pearson’s correlation matrix for the variables discussed in previous
sections. The CSR performance is negatively correlated with AQ and two REM indices. In
addition, the weak correlations between our control variables suggest that they may capture
non-overlapping information regarding EM.
2
4. Empirical results
4.1. Test of the main results
In this section, we examine the effect of CSR performance on EM. The baseline regression is as
follows:
EMi;t¼α0þα1CSRi;t1þα2SIZEi;t1þα3CFi;t1þα4CFVi;t1þα5DEBTi;t1þ
α6LOSSi;t1þα7AZi;t1þα8MBi;t1þα9RETi;t1þIND þYEAR þεi;t;(5)
Table 2. Summary statistics.
Panel A: Summary statistics
MEAN STD P1 Q1 MEDIAN Q3 P99
AQ 0.06 0.05 0.01 0.03 0.05 0.07 0.23
REM1 −0.02 0.16 −0.47 −0.05 0.01 0.05 0.24
REM2 −0.03 0.19 −0.50 −0.09 0.00 0.05 0.33
CSR 37.56 12.94 17.89 28.58 34.66 42.82 78.14
CSRs 0.00 1.00 −1.43 −0.66 −0.26 0.33 3.26
Mandatory 0.66 0.47 0.00 0.00 1.00 1.00 1.00
SIZE 24.14 1.23 21.52 23.26 24.11 24.98 27.01
CF 0.09 0.19 −0.08 0.03 0.07 0.12 0.39
CFV 0.05 0.18 0.00 0.01 0.03 0.05 0.28
DEBT 0.47 0.22 0.02 0.31 0.48 0.64 0.94
LOSS 0.12 0.33 0.00 0.00 0.00 0.00 1.00
AZ 2.99 4.51 −0.13 0.94 2.25 3.95 13.61
MB 1.65 1.73 0.08 0.63 1.18 2.09 8.00
RET 0.09 0.15 −0.33 0.04 0.09 0.15 0.39
Panel B: Descriptive statistics of subsample
Mandatory CSR Voluntary CSR Difference
MEAN MEDIAN STD MEAN MEDIAN STD MEAN t-statistics
AQ 0.05 0.04 0.04 0.07 0.05 0.07 −0.01 −5.22 ***
REM1 −0.02 0.01 0.14 −0.01 0.01 0.20 −0.01 −1.57
REM2 −0.04 −0.01 0.16 −0.02 0.00 0.24 −0.03 −4.15 ***
CSR 38.46 35.04 14.02 35.74 34.25 9.92 2.72 5.99 ***
CSRs 0.12 −0.19 1.09 −0.23 −0.39 0.74 0.34 9.83 ***
SIZE 24.38 24.38 1.21 23.67 23.62 1.11 0.71 17.05 ***
CF 0.09 0.07 0.11 0.09 0.07 0.27 0.00 0.01
CFV 0.05 0.03 0.11 0.05 0.03 0.28 −0.01 −0.99
DEBT 0.50 0.51 0.22 0.40 0.39 0.22 0.10 12.96 ***
LOSS 0.11 0.00 0.31 0.15 0.00 0.36 −0.05 −4.00 ***
AZ 2.95 2.15 3.70 3.04 2.37 5.65 −0.09 −0.54
MB 1.47 1.01 1.54 2.16 1.60 2.10 −0.69 −11.04 ***
RET 0.10 0.10 0.12 0.07 0.08 0.21 0.03 5.76 ***
Panel A presents summary statistics for the whole sample. Panel B reports descriptive statistics for mandatory CSR disclosure
subsample (Mandatory CSR) and voluntary CSR (Voluntary CSR) disclosure subsample, respectively. AQ is the accrual quality.
REM1 is the first index of real earnings management and REM2 is the second index of real earnings management. CSR is the
original CSR rating score and CSRs is the standardized CSR rating. Mandatory is a dummy variable of mandatory CSR
disclosure. SIZE is the natural logarithms of total assets. CF is cash flow from operations to total assets and CFV is the
standard deviation of cash flow. DEBT is firm’s leverage ratio. LOSS is a dummy variable, equaling to 1 if operating income is
negative, 0 otherwise. AZ is the Altman’s Z-score. MB is the market to book ratio and ROA is return on assets.
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 7
where EM is earnings management, for which we use AQ, REM1, and REM2 as proxies, and CSR
is the measure of CSR performance. Control variables are outlined in Section 3.4. The model is
estimated using pooled ordinary least square with year- and industry-fixed effects.
Table 4 shows the estimated results for baseline regressions. AQ is negatively and significantly
associated with CSR performance in Model 1 and 2, suggesting that socially responsible firms
engage in less AEM activities.
3
On including control variables in Model 2, the results are more in
agreement with relevant studies (Bergstresser and Philippon 2004; Zang 2012; Chen et al. 2015).
AQ is positively related to cash flow volatility, loss dummy variable, and Altman’s Z-score,
suggesting that less stable firms manage their earnings more. AQ is positively related to mar-
ket-to-book ratio, but negatively to debt ratio.
Moreover, the coefficients of REM1 and REM2 are all significant and negative in Model 3 and
5, indicating that firms with a higher level of CSR performance engage in less REM. The results of
control variables in Model 4 and 6 almost corroborate those of relevant literature Zang (2012) and
Chen et al. (2015). Firms with insufficient cash flow and lower MB ratio are more likely to engage
in acceleration of sales, overproduction, and opportunistic reduction of R&D expenditure.
The aforementioned analyses focus on the role of CSR performance in managing earnings
activities. However, studies suggest that distinguish between CSR performance and CSR disclosure
is essential (Richardson, Welker, and Hutchinson 1999; Wang, Cao, and Ye 2016). Therefore, in
this section, we further investigate how the willingness of CSR disclosure affects the relationship
between CSR performance and EM. As previously described, our sample enables differentiation of
the sample into two subsamples, voluntary and mandatory CSR disclosure.
Table 5 reports the regression results of EM on CSR performance across different subsamples.
The coefficients of CSR rating scores are negative and significant in all models for mandatory CSR
firms. By contrast, no coefficient of CSR performance is significant for voluntary CSR firms. Thus,
these results suggest that mandatory CSR reporting under supervision ensures quality; this could
mitigate information asymmetry and constrain EM activities3F.
4
4.2. Endogeneity
Studies document that the extent of EM and CSR performance might be simultaneously
determined by other external factors (Larcker and Rusticus 2010; Choi, Lee, and Park 2013;
Wang, Cao, and Ye 2016). Both CSR and EM are chosen by managers (Wang, Cao, and Ye
2016). To protect their entrenchment, managers distorting earnings information are also more
likely to engage in CSR activities (Prior, Surroca, and Tribo 2008). Thus, a major concern is
that our results could be driven by the endogeneity of CSR engagement. To address this
problem, we first adopt the 2SLS approach to control for the endogenous relationship between
CSR and EM.
Table 3. Correlation coefficients.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
(1) CSR 1.00
(2) CSRs 0.93 1.00
(3) Mandatory 0.10 0.16 1.00
(4) SIZE 0.35 0.26 0.28 1.00
(5) CF 0.00 0.00 0.00 0.07 1.00
(6) CFV −0.02 −0.02 −0.02 0.01 0.16 1.00
(7) DEBT 0.11 0.17 0.21 0.13 −0.18 −0.07 1.00
(8) LOSS −0.02 −0.06 −0.07 −0.08 −0.17 0.00 0.11 1.00
(9) AZ −0.09 −0.04 −0.01 −0.05 0.66 0.03 0.00 −0.06 1.00
(10) MB −0.17 −0.18 −0.18 0.04 0.16 0.05 −0.44 −0.07 0.07 1.00
(11) RET 0.06 0.10 0.10 0.13 0.24 0.02 −0.04 −0.42 0.10 0.15 1.00
The table presents the Pearson’s correlation matrix of all the variables used in the sample. Bold font denotes statistical
significance at the 10% level.
8Y. GONG AND K.-C. HO
Table 4. Regressions of EM proxies on CSR: main results.
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
constant −0.0134 −0.0243 0.2282*** 0.1916*** 0.4589*** 0.4191***
(−0.47) (−0.82) (3.06) (2.51) (6.10) (5.44)
CSR −0.0002*** −0.0008*** −0.0009***
(−2.82) (−3.71) (−3.83)
CSRs −0.0029*** −0.0099*** −0.0105***
(−2.78) (−3.62) (−3.83)
SIZE 0.0033*** 0.0033*** −0.0068** −0.0068*** −0.0142*** −0.0142***
(2.92) (2.92) (−2.33) (−2.34) (−4.82) (−4.81)
CF 0.0108 0.0108 −0.3107*** −0.3108*** −0.7847*** −0.7848***
(0.84) (0.84) (−8.59) (−8.60) (−21.52) (−21.52)
CFV 0.0329*** 0.0329*** −0.0001 0.0001 −0.0050 −0.0049
(7.00) (7.01) (−0.01) (0.00) (−0.36) (−0.35)
DEBT −0.0225*** −0.0225*** −0.0041 −0.0041 −0.0402*** −0.0401***
(−3.92) (−3.91) (−0.27) (−0.26) (−2.59) (−2.59)
LOSS 0.0123*** 0.0123*** 0.0228*** 0.0229*** 0.0055 0.0056
(3.67) (3.68) (2.52) (2.52) (0.61) (0.62)
AZ 0.0022*** 0.0022*** −0.0021*** −0.0021*** 0.0007 0.0007
(7.82) (7.82) (−2.70) (−2.69) (0.92) (0.92)
MB 0.0013 0.0014 −0.0172*** −0.0171*** −0.0064*** −0.0064***
(1.53) (1.55) (−7.92) (−7.89) (−2.94) (−2.92)
RET −0.0020 −0.0020 −0.1381*** −0.1382*** −0.0812*** −0.0812***
(−0.22) (−0.22) (−5.57) (−5.57) (−3.24) (−3.24)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.15 0.15 0.27 0.27 0.37 0.37
Observations 2,187 2,187 2,715 2,715 2,715 2,715
This table reports the estimated results of the OLS regression. AQ is the accrual quality. REM1 is the first index of real earnings
management and REM2 is the second index of real earnings management. CSR is the rating score of corporate social
responsibility, ranging from 1 to 100. CSRs is the standardized CSR. SIZE is the natural logarithms of total assets. CF is cash
flow from operations to total assets and CFV is the standard deviation of cash flow. DEBT is firm’s leverage ratio. LOSS is a
dummy variable, equaling to 1 if operating income is negative, 0 otherwise. AZ is the Altman’s Z-score. MB is the market to
book ratio and ROA is return on assets. T-statistics is reported in parentheses. ***, **, * denote statistical significance at the
1%, 5%, and 10% levels, based on industry fixed effects (Industry FE) and year fixed effects (Year FE), respectively.
Table 5. Regressions of EM proxies on CSR for mandatory CSR firms and voluntary CSR firms.
Panel A: Mandatory CSR disclosure firms
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
constant −0.0685*** −0.0824*** −0.0359 −0.0826 0.1255 0.0708
(−2.37) (−2.78) (−0.39) (−0.88) (1.39) (0.76)
CSR −0.0003*** −0.0011*** −0.0012***
(−4.01) (−4.43) (−4.91)
CSRs −0.0037*** −0.0127*** −0.0143***
(−3.92) (−4.28) (−4.91)
SIZE 0.0053*** 0.0053*** 0.0045 0.0044 −0.0004 −0.0003
(4.71) (4.69) (1.28) (1.26) (−0.11) (−0.09)
CF 0.0197 0.0197 −0.2777*** −0.2780*** −0.7360*** −0.7360***
(1.59) (1.59) (−6.81) (−6.81) (−18.30) (−18.31)
CFV 0.0460*** 0.0460*** 0.0002 0.0005 0.0222 0.0224
(5.48) (5.49) (0.01) (0.02) (0.81) (0.82)
DEBT −0.0100* −0.0099* 0.0199 0.0200 0.0052 0.0055
(−1.81) (−1.80) (1.12) (1.12) (0.30) (0.31)
LOSS 0.0121*** 0.0121*** 0.0229** 0.0229** 0.0151 0.0152
(3.61) (3.62) (2.10) (2.10) (1.40) (1.41)
AZ 0.0011*** 0.0011*** −0.0017* −0.0017* 0.0010 0.0011
(4.10) (4.10) (−1.88) (−1.87) (1.17) (1.18)
MB 0.0017** 0.0017** −0.0162*** −0.0161*** −0.0060*** −0.0059***
(2.07) (2.11) (−6.41) (−6.36) (−2.39) (−2.36)
(Continued)
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 9
First stage:
CSRi;t¼#0þ#1ΔCSRi;t1þ#2ΔCSRi;t2þ#3Industry CSRi;tþXvi;tCONTROLi;tþεi;t(6)
Second stage:
EMi;t¼ω0þω1d
CSRi;tþXwi:tCONTROLi;tþμi;t(7)
For the first-stage regression, the industry-median CSR (Industry CSRi;t), the differential of one-
period-lagged CSR (ΔCSRi;t1), and the differential of two-period-lagged CSR (ΔCSRi;t2) are used
as the instrumental variables (Cui, Jo, and Na 2018). Subsequently, we regress EM on the
estimated value of CSR ( d
CSRi;t) from the first-stage regression in the second-stage regression.
Table 5. (Continued).
Panel A: Mandatory CSR disclosure firms
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
RET 0.0124 0.0122 −0.2651*** −0.2656*** −0.1500*** −0.1502***
(1.16) (1.14) (−7.73) (−7.74) (−4.44) (−4.44)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.19 0.19 0.35 0.35 0.45 0.45
Observations 1,663 1,663 1,885 1,885 1,885 1,885
Panel B: Voluntary CSR disclosure firms
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
constant −0.0112 0.0001 0.4284*** 0.4092*** 0.8226*** 0.7950***
(−0.13) (0.00) (2.66) (2.52) (4.96) (4.75)
CSR 0.0003 −0.0004 −0.0006
(0.89) (−0.72) (−1.05)
CSRs 0.0030 −0.0052 −0.0074
(0.81) (−0.77) (−1.06)
SIZE 0.0033 0.0033 −0.0167*** −0.0167*** −0.0302*** −0.0302***
(0.91) (0.91) (−2.56) (−2.56) (−4.49) (−4.49)
CF −0.0020 −0.0018 −0.3743*** −0.3741*** −1.0142*** −1.0141***
(−0.05) (−0.04) (−4.24) (−4.24) (−11.14) (−11.14)
CFV 0.0308*** 0.0308*** −0.0058 −0.0058 −0.0208 −0.0208
(4.23) (4.22) (−0.35) (−0.35) (−1.21) (−1.21)
DEBT −0.0572*** −0.0572*** −0.0142 −0.0143 −0.0858*** −0.0858***
(−3.37) (−3.37) (−0.46) (−0.46) (−2.70) (−2.70)
LOSS 0.0112 0.0112 0.0175 0.0176 −0.0187 −0.0186
(1.33) (1.33) (1.08) (1.08) (−1.12) (−1.11)
AZ 0.0050*** 0.0050*** −0.0031** −0.0031** −0.0007 −0.0007
(6.85) (6.85) (−1.96) (−1.96) (−0.40) (−0.40)
MB −0.0046 −0.0047 −0.0166*** −0.0166*** −0.0051 −0.0051
(−1.54) (−1.54) (−3.77) (−3.77) (−1.11) (−1.12)
RET −0.0179 −0.0179 −0.0264 −0.0263 0.0081 0.0082
(−0.94) (−0.94) (−0.67) (−0.66) (0.20) (0.20)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.15 0.15 0.14 0.14 0.27 0.27
Observations 524 524 830 830 830 830
This table reports the estimated results of the OLS regression for mandatory CSR firms and voluntary CSR firms. Panel A shows the
mandatory CSR disclosure firms, and Panel B shows the voluntary CSR disclosure firm. AQ is the accrual quality. REM1 is the first
index of real earnings management and REM2 is the second index of real earnings management. CSR is the rating score of
corporate social responsibility, ranging from 1 to 100. CSRs is the standardized CSR. SIZE is the natural logarithms of total assets.
CF is cash flow from operations to total assets and CFV is the standard deviation of cash flow. DEBT is firm’s leverage ratio. LOSS
is a dummy variable, equaling to 1 if operating income is negative, 0 otherwise. AZ is the Altman’s Z-score. MB is the market to
book ratio and ROA is return on assets. T-statistics is reported in parentheses. ***, **, * denote statistical significance at the 1%,
5%, and 10% levels, based on industry fixed effects (Industry FE) and year fixed effects (Year FE), respectively.
10 Y. GONG AND K.-C. HO
The results of the first-stage regression are shown in Model 1 and 2 in Panel A of Table 6.We
find that CSR performance is significantly and negatively correlated with the differentials of one-
and two-period-lagged CSR, but significantly and positively correlated with the industry-median
CSR. Model 3–8 show the estimated results of the second-stage regression. The significant and
negative coefficients of CSR performance confirm the negative impact of CSR on EM, indicating
that socially responsible firms engage in less AEM and REM. As shown in Panel B, we also find
that the negative association between CSR performance and EM exists for mandatory CSR
disclosure firms.
Moreover, the generalized method of moments (GMM) approach is used to model the
endogeneity of CSR (Riccardo 2002; Tran and Tsionas 2013; Seo and Shin 2016). After controlling
for endogeneity, the results in Panel A of Table 7 suggest that the long-term perspective hypoth-
esis remains a valid dominant factor for determining CSR performance. CSR performance
remains negatively associated with EM. In Panel B, the negative association between CSR
performance and EM also exists for mandatory CSR disclosure firms.
In addition, only 3,409 firm-year observations exist in our sample and firms without CSR
reports are excluded, which may induce a selection bias and endogeneity. As suggested by
Table 6. Results of EM proxies on CSR: 2SLS regressions.
Panel A: Earnings management and CSR using 2SLS regressions
First stage: Second stage:
Dependent
variable CSR CSRs AQ REM1 REM2
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
constant −75.3529*** −6.1749*** −0.0462 −0.0580* 0.2724*** 0.2272** 0.5232*** 0.4803***
(−5.21) (−8.98) (−1.41) (−1.72) (2.65) (2.16) (5.07) (4.53)
ΔCSR
t-1
−0.6126***
(−12.26)
ΔCSR
t-2
−0.6024***
(−13.31)
Industry_CSR 0.8927***
(2.86)
ΔCSRs
t-1
−0.7063***
(−13.93)
ΔCSRs
t-2
−0.5930***
(−12.70)
Industry_CSRs 0.7837**
(2.23)
c
CSR −0.0003*** −0.0010*** −0.0010***
(−2.80) (−3.36) (−3.13)
c
CSRs −0.0032*** −0.0122*** −0.0115***
(−2.76) (−3.32) (−3.11)
SIZE 3.0040*** 0.2481*** 0.0048*** 0.0048*** −0.0081** −0.0081** −0.0168*** −0.0168***
(9.90) (9.85) (3.70) (3.69) (−2.04) (−2.04) (−4.18) (−4.17)
CF 5.2971 0.4425 −0.0049 −0.0048 −0.1390*** −0.1388*** −0.6950*** −0.6947***
(1.47) (1.48) (−0.35) (−0.35) (−3.03) (−3.03) (−15.08) (−15.07)
CFV −0.7170 −0.0628 0.0454*** 0.0454*** −0.0067 −0.0067 −0.0021 −0.0022
(−0.37) (−0.40) (6.28) (6.28) (−0.27) (−0.28) (−0.09) (−0.09)
DEBT 2.7070* 0.2138 −0.0153*** −0.0153*** −0.0073 −0.0072 −0.0262 −0.0262
(1.70) (1.62) (−2.43) (−2.42) (−0.36) (−0.36) (−1.29) (−1.29)
LOSS −0.0466 −0.0095 0.0083** 0.0083** 0.0365*** 0.0366*** 0.0170 0.0171
(−0.05) (−0.13) (2.30) (2.31) (3.15) (3.16) (1.46) (1.47)
AZ 0.0224 0.0017 0.0022*** 0.0022*** −0.0049*** −0.0049*** −0.0020* −0.0020*
(0.26) (0.23) (6.64) (6.64) (−4.43) (−4.42) (−1.81) (−1.80)
MB −1.2593*** −0.1098*** 0.0020* 0.0020* −0.0130*** −0.0129*** 0.0007 0.0007
(−4.71) (−4.96) (1.67) (1.69) (−3.80) (−3.78) (0.20) (0.22)
RET 2.1876 0.1795 0.0007 0.0007 −0.1615*** −0.1618*** −0.1178*** −0.1180***
(Continued)
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 11
Table 6. (Continued).
Panel A: Earnings management and CSR using 2SLS regressions
First stage: Second stage:
Dependent
variable CSR CSRs AQ REM1 REM2
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
(0.83) (0.82) (0.07) (0.06) (−4.82) (−4.83) (−3.50) (−3.50)
Year FE YES YES YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES YES YES
Adj. R
2
0.39 0.34 0.17 0.17 0.30 0.30 0.43 0.43
Observations 1,462 1,462 1,294 1,294 1,462 1,462 1,462 1,462
Panel B: Earnings management and mandatory CSR disclosure using 2SLS regressions
First stage: Second stage:
Dependent
variable CSR CSRs AQ REM1 REM2
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
constant −78.9575*** −6.8223*** −0.0509 −0.0621* 0.1265 0.0828 0.3496*** 0.3028**
(−4.39) (−7.53) (−1.47) (−1.75) (0.97) (0.62) (2.71) (2.29)
ΔCSR
t-1
−0.6800***
(−11.45)
ΔCSR
t-2
−0.6485***
(−11.65)
Industry_CSR 0.7592**
(2.07)
ΔCSRs
t-1
−0.7704***
(−12.74)
ΔCSRs
t-2
−0.6059***
(−10.62)
Industry_CSRs 0.7701*
(1.82)
CSR −0.0003*** −0.0010*** −0.0011***
(−2.87) (−3.01) (−3.16)
CSRs −0.0031*** −0.0121*** −0.0127***
(−2.80) (−2.92) (−3.11)
SIZE 3.3335*** 0.2751*** 0.0046*** 0.0046*** −0.0016 −0.0017 −0.0097** −0.0097**
(8.38) (8.33) (3.48) (3.46) (−0.31) (−0.33) (−1.96) (−1.96)
CF 2.3923 0.1970 −0.0822*** −0.0822*** −0.1937*** −0.1936*** −0.8015*** −0.8013***
(0.43) (0.43) (−4.66) (−4.66) (−2.86) (−2.86) (−12.04) (−12.04)
CFV 3.9996 0.3464 0.2437*** 0.2437*** 0.0951 0.0948 0.2798*** 0.2796***
(0.50) (0.52) (9.56) (9.55) (0.97) (0.97) (2.91) (2.90)
DEBT 2.1419 0.1565 −0.0128** −0.0128** −0.0057 −0.0056 −0.0090 −0.0089
(1.08) (0.95) (−2.02) (−2.01) (−0.23) (−0.23) (−0.38) (−0.37)
LOSS 0.8252 0.0625 0.0051 0.0051 0.0282* 0.0283* 0.0149 0.0149
(0.70) (0.63) (1.35) (1.35) (1.95) (1.95) (1.04) (1.05)
AZ 0.0121 0.0011 0.0002 0.0002 −0.0057*** −0.0056*** −0.0018 −0.0018
(0.10) (0.11) (0.53) (0.53) (−3.86) (−3.85) (−1.25) (−1.23)
MB −2.0425*** −0.1766*** 0.0016 0.0016 −0.0090** −0.0088** 0.0045 0.0046
(−5.67) (−5.92) (1.34) (1.38) (−2.02) (−1.97) (1.02) (1.06)
RET 8.8631** 0.7406** 0.0455*** 0.0454*** −0.2392*** −0.2399*** −0.1503*** −0.1508***
(2.27) (2.28) (3.63) (3.62) (−5.01) (−5.03) (−3.20) (−3.21)
Year FE YES YES YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES YES YES
Adj. R
2
0.42 0.37 0.25 0.25 0.33 0.33 0.46 0.46
Observations 1,054 1,054 1,002 1,002 1,054 1,054 1,054 1,054
Panel A & B report the estimated results of two-stage least-squares regression. AQ is the accrual quality. REM1 is the first index
of real earnings management and REM2 is the second index of real earnings management. CSR is the rating score of
corporate social responsibility. CSRs is the standardized CSR. The differential of one-period-lagged CSR (ΔCSRi;t1), the
differential of two-period-lagged CSR (ΔCSRi;t2), the industry-median CSR (Industry CSRi;t), and the estimated value of CSR
(d
CSRi;t). SIZE is the natural logarithm of total assets. CF is cash flow from operations to total assets and CFV is the standard
deviation of cash flow. DEBT is firm’s leverage ratio. LOSS is a dummy variable, equaling to 1 if operating income is negative,
0 otherwise. AZ is the Altman’s Z-score. MB is the market to book ratio and ROA is return on assets. T-statistics is reported in
parentheses. ***, **, * denote statistical significance at the 1%, 5%, and 10% levels, based on industry fixed effects (Industry
FE) and year fixed effects (Year FE), respectively.
12 Y. GONG AND K.-C. HO
Table 7. Results of EM proxies on CSR: GMM regressions.
Panel A: Earnings management and CSR using GMM regressions
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
constant 0.0048 0.0081 0.1936*** 0.1966*** 0.3259*** 0.2981***
(0.14) (0.23) (2.61) (2.68) (4.23) (3.85)
CSR −0.0003** −0.0009*** −0.0008***
(−2.29) (−3.90) (−3.96)
CSRs −0.0026** −0.0110*** −0.0121***
(−2.04) (−4.08) (−4.85)
SIZE 0.0027* 0.0022 −0.0054* −0.0071*** −0.0091*** −0.0095***
(1.91) (1.61) (−1.76) (−2.41) (−2.82) (−3.00)
CF 0.0068 0.0048 −0.4005*** −0.3997*** −0.8472*** −0.8400***
(0.36) (0.26) (−4.21) (−4.15) (−5.96) (−5.95)
CFV 0.0308*** 0.0309*** 0.0070 0.0053 0.0018 0.0009
(4.10) (4.10) (0.68) (0.54) (0.13) (0.07)
DEBT −0.0269*** −0.0284*** −0.0161 −0.0165 −0.0422*** −0.0420***
(−3.05) (−3.26) (−0.87) (−0.91) (−2.34) (−2.35)
LOSS 0.0106*** 0.0106*** 0.0407*** 0.0403*** 0.0081 0.0101
(2.89) (2.91) (3.42) (3.38) (0.55) (0.69)
AZ 0.0024*** 0.0026*** −0.0004 −0.0013 −0.0003 −0.0004
(2.72) (2.91) (−0.16) (−0.56) (−0.09) (−0.12)
MB 0.0010 0.0009 −0.0199*** −0.0194*** −0.0089** −0.0088**
(0.86) (0.82) (−4.68) (−4.57) (−2.02) (−2.01)
RET 0.0316* 0.0319* 0.0215 0.0228 0.0215 0.0225
(1.73) (1.73) (1.11) (1.14) (1.44) (1.46)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.15 0.15 0.26 0.26 0.36 0.37
Observations 2,187 2,187 2,175 2,175 2,175 2,175
Panel B: Earnings management and mandatory CSR disclosure using GMM regressions
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
constant −0.0385 −0.0296 0.0363 0.0373 −0.0132 −0.0420
(−1.15) (−0.89) (0.43) (0.44) (−0.17) (−0.54)
CSR −0.0003*** −0.0012*** −0.0012***
(−2.85) (−4.59) (−5.13)
CSRs −0.0029** −0.0144*** −0.0166***
(−2.26) (−4.56) (−5.85)
SIZE 0.0043*** 0.0035*** 0.0015 −0.0006 0.0049 0.0039
(3.12) (2.66) (0.41) (−0.18) (1.52) (1.24)
CF 0.0257 0.0215 −0.4303*** −0.4201*** −0.8997*** −0.8741***
(1.53) (1.31) (−3.76) (−3.64) (−5.27) (−5.21)
CFV 0.0198 0.0195 0.0140 0.0122 0.0220 0.0226
(1.40) (1.36) (0.48) (0.42) (0.60) (0.61)
DEBT −0.0174** −0.0195** −0.0018 −0.0041 0.0002 −0.0008
(−2.02) (−2.29) (−0.08) (−0.18) (0.01) (−0.03)
LOSS 0.0052 0.0052 0.0511*** 0.0509*** 0.0233 0.0248
(1.44) (1.43) (3.17) (3.16) (1.23) (1.31)
AZ 0.0014*** 0.0016*** 0.0006 −0.0009 0.0043 0.0031
(2.40) (2.66) (0.24) (−0.34) (0.98) (0.76)
MB 0.0012 0.0012 −0.0215*** −0.0208*** −0.0088* −0.0093*
(1.15) (1.20) (−3.93) (−3.86) (−1.69) (−1.78)
RET 0.1098*** 0.1130*** 0.0617 0.0699 0.0117 0.0215
(3.72) (3.74) (1.11) (1.25) (0.23) (0.44)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.21 0.21 0.33 0.33 0.43 0.44
Observations 1,663 1,663 1,885 1,885 1,885 1,885
Panel A & B report the estimated results of GMM regression. AQ is the accrual quality. REM1 is the first index of real earnings
management and REM2 is the second index of real earnings management. CSR is the rating score of corporate social
responsibility. CSRs is the standardized CSR. SIZE is the natural logarithm of total assets. CF is cash flow from operations to
total assets and CFV is the standard deviation of cash flow. DEBT is firm’s leverage ratio. LOSS is a dummy variable, equaling
to 1 if operating income is negative, 0 otherwise. AZ is the Altman’s Z-score. MB is the market to book ratio and ROA is return
on assets. T-statistics is reported in parentheses. ***, **, * denote statistical significance at the 1%, 5%, and 10% levels, based
on industry fixed effects (Industry FE) and year fixed effects (Year FE), respectively.
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 13
Armstrong, Balakrishnan, and Cohen (2012), we mitigate these potential effects by including all
A-share listed firms during 1999–2015 and employing a difference-in-difference (DID) specifica-
tion to further isolate the impact of CSR disclosure on EM, thus explaining the causal effect of
CSR in our study. The first difference compares EM before and after CSR disclosure is imple-
mented, whereas the second compares the change in EM between CSR and non-CSR firms during
the same period. This approach illustrates the regression framework when many CSR disclosures
are staggered over time (Bertrand and Mullainathan 2003). Because different firms began to
disclose their CSR reports in different years, we use the following model to perform a DID
regression (Bertrand and Mullainathan 2003; Armstrong, Balakrishnan, and Cohen 2012; Gong
and Ho 2018):
EMit ¼β0þβ1CSR dummyit þXβi;tCONTROLi;tþεi;t(8)
where CSR_dummyit is an indicator variable, taking a value of 1 if firm istarts to disclose CSR
reports during year t, 0 otherwise. We use CSR (CSR
DID
) and mandatory CSR disclosure
(CSR
mandtory
) as proxies for CSR dummy. In this setting, the control group includes all firms
not starting to report their CSR activities at year t, even though they have already began reporting
or will report after year t.
Panel A (B) of Table 8 presents the estimated results of the DID regressions. We find that the
estimated coefficients of CSR
DID
(CSR
mandtory
) are negative and significant, indicating that firms
with CSR (mandatory CSR) disclosures are less likely to engage in EM activities. Thus, our
previous analyses in this paper are robust.
4.3. Alternative measures of CSR
We examine the robustness of our results across alternative measures of CSR performance. In
addition to the CSR rating score, RKS has provided two alternative measures of CSR performance
since 2010. The first one is the rating level of CSR (Rating_L), ranging from AA+ (19, the highest)
to CCC (1, the lowest). The second measure is the ranking level of CSR (Ranking). Each firm is
assigned a ranking level –from 1 (highest) to N (lowest), according to its annual CSR rating score
in any given year. In particular, the firm with the highest CSR rating score is assigned a Ranking
value of 1, the firm with the second highest CSR rating score is assigned a value of 2 and so on,
until finally, the firm with the lowest CSR rating score is assigned a value of N (N = the total
number of firms in any given year of our sample). The results of Panel A from Table 9 show that
Rating_L is negatively and significantly associated with EM, whereas Ranking is positively and
significantly associated with EM, suggesting that our results are robust to alternative measures of
CSR performance. The results reported in Panel B and C from Table 9 are consistent with
Hypothesis 2 that mandatory CSR reporting could mitigate information asymmetry and constrain
EM activities.
5. Conclusion
This study examined the effect of CSR performance on managerial short-termism. Using EM as a
proxy for managerial short-termism and the unique CSR rating information from RKS, we
identified that firms with superior CSR performance engage in less AEM and REM. Specifically,
this negative association is only evident for mandatory CSR firms.
Socially responsible firms not only focus on current profits but also make efforts to foster long-
term relationships with general stakeholders. Moreover, our findings emphasize the role of
regulators in improving the quality of firms’financial reporting by mandating firms to disclose
nonfinancial information. Thus, stakeholders can evaluate a firm’sfinancial reports more precisely
14 Y. GONG AND K.-C. HO
Table 8. Results of EM proxies on CSR: DID regressions.
Panel A: Earnings management and CSR using DID regressions
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3
constant 0.0846*** 0.1207*** 0.0834***
(7.14) (5.47) (3.22)
CSRDID −0.0086*** −0.0231*** −0.0345***
(−4.88) (−6.26) (−7.94)
SIZE −0.0016*** −0.0078*** −0.0044***
(−3.42) (−8.40) (−4.03)
CF 0.0426*** −0.0651*** −0.5921***
(18.58) (−15.42) (−119.35)
CFV 0.0804*** 0.0160*** −0.0146***
(42.41) (4.31) (−3.35)
DEBT −0.0141*** 0.0800*** 0.0248***
(−4.87) (14.84) (3.91)
LOSS 0.0141*** 0.0741*** 0.0155***
(9.61) (26.14) (4.63)
AZ 0.0001*** −0.0022*** −0.0018***
(5.77) (−97.22) (−67.02)
MB −0.0027*** −0.0001* 0.0004***
(−14.59) (−1.70) (4.30)
RET −0.0022*** −0.0006 0.0010
(−2.85) (−0.93) (1.28)
Year FE YES YES YES
Industry FE YES YES YES
Adj. R
2
0.23 0.32 0.52
Observations 14,521 25,065 25,065
Panel B: Earnings management and mandatory CSR disclosure using DID regressions
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3
constant 0.0811*** 0.1126*** 0.0697***
(6.82) (5.06) (2.67)
CSRmandatory −0.0104*** −0.0264*** −0.0407***
(−5.65) (−6.61) (−8.66)
SIZE −0.0015*** −0.0074*** −0.0038***
(−3.09) (−7.95) (−3.47)
CF 0.0426*** −0.0649*** −0.5918***
(18.63) (−15.39) (−119.33)
CFV 0.0803*** 0.0159*** −0.0147***
(42.38) (4.30) (−3.38)
DEBT −0.0135*** 0.0812*** 0.0269***
(−4.68) (15.04) (4.23)
LOSS 0.0140*** 0.0740*** 0.0152***
(9.52) (26.08) (4.55)
AZ 0.0001*** −0.0022*** −0.0018***
(5.78) (−97.22) (−67.03)
MB −0.0027*** −0.0001* 0.0004***
(−14.59) (−1.69) (4.31)
RET −0.0022*** −0.0006 0.0010
(−2.82) (−0.92) (1.30)
Year FE YES YES YES
Industry FE YES YES YES
Adj. R
2
0.23 0.32 0.52
Observations 14,521 25,065 25,065
Panel A & B report the estimated results of DID regression. AQ is the accrual quality. REM1 is the first index of real earnings
management and REM2 is the second index of real earnings management. CSR (CSR
DID
) and mandatory CSR disclosure
(CSR
mandtory
) as proxy for CSR dummy. SIZE is the natural logarithm of total assets. CF is cash flow from operations to total
assets and CFV is the standard deviation of cash flow. DEBT is firm’s leverage ratio. LOSS is a dummy variable, equaling to 1 if
operating income is negative, 0 otherwise. AZ is the Altman’s Z-score. MB is the market to book ratio and ROA is return on
assets. T-statistics is reported in parentheses. ***, **, * denote statistical significance at the 1%, 5%, and 10% levels, based on
industry fixed effects (Industry FE) and year fixed effects (Year FE), respectively.
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 15
Table 9. Regressions of EM proxies on CSR: alternative measures of CSR performance.
Panel A: Effect of alternative measures of CSR on earnings management
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
constant −0.0151 −0.0081 0.2436*** 0.2878*** 0.4874*** 0.5324***
(−0.47) (−0.26) (2.92) (3.55) (5.75) (6.45)
Ranking 1.E-05** 0.0001*** 0.0001***
(2.04) (3.66) (3.57)
Rating_L −0.0008*** −0.0033*** −0.0032***
(−2.38) (−3.62) (−3.42)
SIZE 0.0030*** 0.0032*** −0.0097*** −0.0096*** −0.0171*** −0.0170***
(2.45) (2.58) (−3.12) (−3.05) (−5.39) (−5.34)
CF 0.0048 0.0051 −0.2751*** −0.2752*** −0.7663*** −0.7665***
(0.35) (0.37) (−7.18) (−7.18) (−19.64) (−19.64)
CFV 0.0349*** 0.0349*** −0.0002 −0.0001 −0.0084 −0.0083
(6.70) (6.70) (−0.01) (−0.01) (−0.54) (−0.54)
DEBT −0.0252*** −0.0250*** −0.0115 −0.0113 −0.0479*** −0.0477***
(−4.09) (−4.07) (−0.70) (−0.69) (−2.88) (−2.87)
LOSS 0.0118*** 0.0118*** 0.0254*** 0.0256*** 0.0083 0.0085
(3.29) (3.31) (2.62) (2.65) (0.84) (0.87)
AZ 0.0023*** 0.0023*** −0.0020*** −0.0021*** 0.0007 0.0007
(7.89) (7.89) (−2.43) (−2.46) (0.83) (0.79)
MB 0.0013 0.0013 −0.0162*** −0.0163*** −0.0055*** −0.0057***
(1.46) (1.38) (−7.13) (−7.18) (−2.40) (−2.44)
RET −0.0017 −0.0015 −0.1348*** −0.1337*** −0.0784*** −0.0776***
(−0.18) (−0.16) (−5.17) (−5.13) (−2.96) (−2.92)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.15 0.15 0.27 0.27 0.37 0.37
Observations 1,942 1,942 2,413 2,413 2,413 2,413
Panel B: Alternative measures of CSR for mandatory CSR disclosure firms
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
constant −0.0782*** −0.0668** −0.0263 0.0230 0.1235 0.1765*
(−2.44) (−2.13) (−0.25) (0.23) (1.20) (1.75)
Ranking 2.E-05*** 0.0001*** 0.0001***
(3.35) (4.05) (4.26)
Rating_L −0.0011*** −0.0042*** −0.0043***
(−3.59) (−4.13) (−4.24)
SIZE 0.0050*** 0.0052*** 0.0014 0.0019 −0.0029 −0.0025
(4.16) (4.30) (0.37) (0.48) (−0.76) (−0.66)
CF −0.0065 −0.0059 −0.2475*** −0.2453*** −0.7274*** −0.7252***
(−0.47) (−0.43) (−5.37) (−5.32) (−15.83) (−15.78)
CFV 0.0887*** 0.0886*** 0.0135 0.0133 0.0466 0.0464
(6.50) (6.50) (0.30) (0.29) (1.02) (1.02)
DEBT −0.0115** −0.0115** 0.0099 0.0104 −0.0009 −0.0004
(−1.96) (−1.95) (0.52) (0.54) (−0.05) (−0.02)
LOSS 0.0109*** 0.0110*** 0.0249** 0.0255** 0.0189 0.0195*
(3.07) (3.10) (2.11) (2.16) (1.60) (1.65)
AZ 0.0013*** 0.0013*** −0.0015 −0.0016* 0.0010 0.0010
(4.37) (4.34) (−1.58) (−1.66) (1.08) (1.00)
MB 0.0018** 0.0017** −0.0142*** −0.0145*** −0.0042 −0.0045*
(2.14) (2.04) (−5.36) (−5.45) (−1.59) (−1.69)
RET 0.0198* 0.0200* −0.2648*** −0.2637*** −0.1456*** −0.1447***
(1.73) (1.74) (−7.09) (−7.06) (−3.91) (−3.89)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.20 0.20 0.35 0.35 0.45 0.45
Observations 1,446 1,446 1,624 1,624 1,624 1,624
Panel C: Alternative measures of CSR for voluntary CSR disclosure firms
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
constant −0.0022 −0.0143 0.4175*** 0.4620*** 0.8081*** 0.8585***
(−0.02) (−0.16) (2.45) (2.75) (4.58) (4.95)
(Continued)
16 Y. GONG AND K.-C. HO
with this additional nonfinancial information and can supervise and detect managers’short-
termism behaviors effectively. This is particularly true for emerging markets.
Further analyses regarding which aspects of CSR performance have greater impacts on EM are
warranted. Additional studies on other proxies for managerial short-termism is necessary.
Notes
1. We empirically examine this argument. Specifically, we firstly run a regression of concurrent financial
performance (returns on asset (ROA) in year t) on CSR performance (CSR rating score in year t) and find a
significant and positive association. Then, we run a regression of future financial performance (ROA in year
t + 1, t + 2, t + 3, t + 4, respectively) on CSR performance (CSR rating score in year t) and also find a
significantly positive association.
2. We use Variance Inflation Factor (VIF) to assess multicollinearity. We find that all VIFs are lower than 10,
which indicates that multicollinearity does not seem to be a problem for our test model.
3. The sample size for AQ is relatively small because there are more missing variables in Equation (1). Thus, all
models of AQ have a smaller number of observations than those of REM1 and REM2.
4. Because large firms and SOEs are subject to mandatory disclosure and less likely to engage in earnings
management, we examine the extent to which our results hold outside of large firms or SOEs. First, we
Table 9. (Continued).
Panel C: Alternative measures of CSR for voluntary CSR disclosure firms
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Ranking 0.0000 0.0001 0.0001
(−0.72) (1.44) (1.59)
Rating_L 0.0011 −0.0025 −0.0034
(0.89) (−1.16) (−1.52)
SIZE 0.0036 0.0036 −0.0182*** −0.0184*** −0.0312*** −0.0313***
(0.96) (0.95) (−2.70) (−2.74) (−4.50) (−4.51)
CF 0.0010 0.0008 −0.3614*** −0.3654*** −1.0047*** −1.0084***
(0.02) (0.02) (−4.00) (−4.04) (−10.75) (−10.80)
CFV 0.0308*** 0.0309*** −0.0066 −0.0066 −0.0215 −0.0217
(4.13) (4.14) (−0.39) (−0.39) (−1.23) (−1.23)
DEBT −0.0582*** −0.0581*** −0.0161 −0.0166 −0.0881*** −0.0885***
(−3.31) (−3.30) (−0.51) (−0.52) (−2.70) (−2.71)
LOSS 0.0119 0.0117 0.0192 0.0193 −0.0179 −0.0179
(1.35) (1.34) (1.15) (1.15) (−1.03) (−1.03)
AZ 0.0051*** 0.0050*** −0.0032** −0.0032** −0.0008 −0.0007
(6.68) (6.66) (−1.97) (−1.97) (−0.45) (−0.44)
MB −0.0049 −0.0049 −0.0163*** −0.0162*** −0.0046 −0.0044
(−1.56) (−1.57) (−3.61) (−3.57) (−0.98) (−0.93)
RET −0.0179 −0.0180 −0.0243 −0.0237 0.0084 0.0093
(−0.92) (−0.93) (−0.60) (−0.59) (0.20) (0.22)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.15 0.15 0.14 0.14 0.27 0.27
Observations 496 496 789 789 789 789
This table presents the estimated results of the relationship between earnings management and alternative measures of CSR
performance. Rating_L is the rating level of CSR, ranging from AA+ (19, the highest) to CCC (1, the lowest). Ranking is the
ranking level of CSR. Each year, firm is assigned a ranking level, ranging from 1 (the highest) to N (the lowest), according to
its CSR rating score. In specific, firm with the highest CSR rating score is assigned a Ranking value of 1, firm with the second
highest CSR rating score is assigned a Ranking value of 2, and so on and so forth, firm with the lowest CSR rating score is
assigned a N (N is the total number of firms in any given year of our sample). AQ is the accrual quality. REM1 is the first index
of real earnings management and REM2 is the second index of real earnings management. SIZE is the natural logarithm of
total assets. CF is cash flow from operations to total assets and CFV is the standard deviation of cash flow. DEBT is firm’s
leverage ratio. LOSS is a dummy variable, equaling to 1 if operating income is negative, 0 otherwise. AZ is the Altman’s
Z-score. MB is the market to book ratio and ROA is return on assets. T-statistics is reported in parentheses. ***, **, * denote
statistical significance at the 1%, 5%, and 10% levels, based on industry fixed effects (Industry FE) and year fixed effect (Year
FE), respectively.
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 17
examine a sample that excludes the largest 20% of firms from our sample to re-estimate our main regression
specified in Equation (5). We find similar results that the moderating effect of CSR on earnings management
is evident only for mandatory CSR disclosure firms. We also find that the strong negative association
between CSR performance and earnings management for mandatory CSR firms holds even when the SOEs
are excluded from our sample. The results are shown in Appendix B.
Acknowledgments
We grateful to the referees for their helpful comments and suggestions. This work was supported by National
Natural Science Foundation of China [grant number 71771217; 71801226]; Zhongnan University of Economics and
Law - the corporate finance team [grant number the first class discipline construction fund]; the Institute of
Guangdong and Taiwan of Shantou University for their open funding financial support.
Disclosure statement
The authors declare no conflict of interest.
Funding
This work was supported by the National Natural Science Foundation of China [71771217, 71801226];Zhongnan
University of Economics and Law, The first class discipline construction fund - the corporate finance team;the
Institute of Guangdong and Taiwan of Shantou University for their open funding financial support.;
ORCID
Yujing Gong http://orcid.org/0000-0001-8706-4004
Kung-Cheng Ho http://orcid.org/0000-0002-3475-2089
References
Allen, A., J. Qian, and M. Qian. 2005.“Law, Finance, and Economic Growth in China.”Journal of Financial
Economics 77: 57–116. doi:10.1016/j.jfineco.2004.06.010.
Altman, E. 1968.“Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy.”The
Journal of Finance 23 (4): 189–209. doi:10.2307/2978933.
Armstrong, C. S., K. Balakrishnan, and D. Cohen. 2012.“Corporate Governance and the Information Environment:
Evidence from State Antitakeover Laws.”Journal of Accounting and Economics 53 (1–2): 185–204. doi:10.1016/j.
jacceco.2011.06.005.
Baber, W. R., P. M. Fairfield, and J. A. Haggard. 1991.“The Effect of Concern about Reported Income on
Discretionary Spending Decisions: The Case of Research and Development.”The Accounting Review 66 (4):
818–829.
Bergstresser, D., and T. Philippon. 2004.“CEO Incentives and Earnings Management.”Journal of Financial
Economics 80 (3): 511–529. doi:10.1016/j.jfineco.2004.10.011.
Bertrand, M., and S. Mullainathan. 2003.“Enjoying the Quiet Life? Corporate Governance and Managerial
Preferences.”Journal of Political Economy 111: 1043–1075. doi:10.1086/376950.
Bhojraj, S., P. Hribar, M. Picconi, and J. McInnis. 2009.“Making Sense of Cents: An Examination of Firms that
Marginally Miss or Beat Analyst Forecasts.”The Journal of Finance 64 (5): 2361–2388. doi:10.1111/j.1540-
6261.2009.01503.x.
Bozzolan, S., M. Fabrizi, C. A. Mallin, and G. Michelon. 2015.“Corporate Social Responsibility and Earnings
Quality: International Evidence.”The International Journal of Accounting 50: 361–396. doi:10.1016/j.
intacc.2015.10.003.
Brochet, F., M. Loumioti, and G. Serafeim. 2015.“Speaking of the Short-Term: Disclosure Horizon and Managerial
Myopia.”Review of Accounting Studies 20 (3): 1122–1163. doi:10.1007/s11142-015-9329-8.
Bushee, B. 1998.“The Influence of Institutional Investors on Myopic R&D Investment Behavior.”The Accounting
Review 73 (3): 305–333.
Call, A. C., S. Chen, B. Miao, and Y. H. Tong. 2014.“Short-Term Earnings Guidance and Accrual-Based Earnings
Management.”Review of Accounting Studies 19 (2): 955–987. doi:10.1007/s11142-013-9270-7.
18 Y. GONG AND K.-C. HO
Cespa, G., and G. Cestone. 2007.“Corporate Social Responsibility and Managerial Entrenchment.”Journal of
Economics and Management Strategy 16: 741–771. doi:10.1111/j.1530-9134.2007.00156.x.
Chen, Y., S. G. Rhee, M. Veeraraghavan, and L. Zolotoy. 2015.“Stock Liquidity and Managerial Short-Termism.”
Journal of Banking & Finance 60 (11): 44–59. doi:10.1016/j.jbankfin.2015.07.007.
Cheng, Q., and T. D. Warfield. 2005.“Equity Incentives and Earnings Management.”The Accounting Review 80 (2):
441–476. doi:10.2308/accr.2005.80.2.441.
Chih, H. L., C. H. Shen, and F. C. Kang. 2007.“Corporate Social Responsibility, Investor Protection, and Earnings
Management: Some International Evidence.”Journal of Business Ethics 79: 179–198. doi:10.1007/s10551-007-
9383-7.
Choi, B. B., D. Lee, and Y. Park. 2013.“Corporate Social Responsibility, Corporate Governance and Earnings
Quality: Evidence from Korea.”Corporate Governance: An International Review 21 (5): 447–467. doi:10.1111/
corg.12033.
Choi, T. H., and J. Pae. 2011.“Business Ethics and Financial Reporting Quality: Evidence from Korea.”Journal of
Business Ethics 103: 403–427. doi:10.1007/s10551-011-0871-4.
Cohen, A., and P. Zarowin. 2010.“Accrual-Based and Real Earnings Management Activities around Seasoned
Equity Offerings.”Journal of Accounting and Economics 50 (1): 2–19. doi:10.1016/j.jacceco.2010.01.002.
Cohen, J., L. Holder-Webb, L. Nath, and D. Wood. 2011.“Retail Investors’Perceptions of the Decision-Usefulness
of Economic Performance, Governance, and Corporate Social Responsibility Disclosures.”Behavioral Research
in Accounting 23 (1): 109–129. doi:10.2308/bria.2011.23.1.109.
Cui, J., H. Jo, and H. Na. 2018.“Does Corporate Social Responsibility Affect Information Asymmetry?”Journal of
Business Ethics 148 (3): 549–572. doi:10.1007/s10551-015-3003-8.
Davies, R., A. G. Haldane, M. Nielsen, and S. Pezzini. 2014.“Measuring the Costs of Short-Termism.”Journal of
Financial Stability 12: 16–25. doi:10.1016/j.jfs.2013.07.002.
Dechow, P., and R. Sloan. 1991.“Executive Incentives and the Horizon Problem: An Empirical Investigation.”
Journal of Accounting and Economics 14 (1): 51–89. doi:10.1016/0167-7187(91)90058-S.
Dechow, P. M., and I. D. Dichev. 2002.“The Quality of Accruals and Earnings: The Role of Accrual Estimation
Errors.”The Accounting Review 77: 35–59. doi:10.2308/accr.2002.77.s-1.35.
Dhaliwal, D., L. Oliver, A. Tsang, and Y. Yang. 2011.“Voluntary Nonfinancial Disclosure and the Cost of Equity
Capital: The Initiation of Corporate Social Responsibility Reporting.”The Accounting Review 86 (1): 59–100.
doi:10.2308/accr.00000005.
Donaldson, T., and L. E. Preston. 1995.“The Stakeholder Theory of the Corporation: Concepts, Evidence, and
Implications.”The Academy of Management Review 20 (1): 65–91. doi:10.2307/258887.
Ferrero, J. M., S. Banerjee, and I. M. G. Sanchez. 2016.“Corporate Social Responsibility as a Strategic Shield against
Costs of Earnings Management Practices.”Journal of Business Ethics 133 (2): 305–324. doi:10.1007/s10551-014-
2399-x.
Gao, L., and J. H. Zhang. 2015.“Firms’Earnings Smoothing, Corporate Social Resistibility and Valuation.”Journal
of Corporate Finance 32: 108–127. doi:10.1016/j.jcorpfin.2015.03.004.
Gelb, D. S., and J. A. Strawser. 2001.“Corporate Social Responsibility and Financial Disclosure: An Alternative
Explanation for Increased Disclosure.”Journal of Business Ethics 33 (1): 1–13. doi:10.1023/A:1011941212444.
Gong, Y., and K. C. Ho. 2018.“Does Corporate Social Responsibility Matter for Corporate Stability? Evidence from
China.”Quality & Quantity 52 (2): 2291–2319. doi:10.1007/s11135-017-0665-6.
Graham, J. R., C. R. Harvey, and S. Rajgopal. 2005.“The Economic Implications of Corporate Financial Reporting.”
Journal of Accounting and Economics 40 (1): 3–73. doi:10.1016/j.jacceco.2005.01.002.
Ho, S. S. M., A. Y. Li, K. Tam, and J. Y. Tong. 2016.“Ethical Image, Corporate Social Responsibility, and R&D
Valuation.”Pacific-Basin Finance Journal 40: 35–348. doi:10.1016/j.pacfin.2016.02.002.
Jones, T. M. 1995.“Instrumental Stakeholder Theory: A Synthesis of Ethics and Economics.”The Academy of
Management Review 20 (2): 404–437. doi:10.5465/AMR.1995.9507312924.
Kim, Y., M. Park, and B. Wier. 2012.“Is Earnings Quality Associated With Corporate Social Responsibility?”The
Accounting Review 87 (3): 761–796. doi:10.2308/accr-10209.
Larcker, D. F., and T. O. Rusticus. 2010.“On the Use of Instrumental Variables in Accounting Research.”Journal of
Accounting and Economics 49: 186–2005. doi:10.1016/j.jacceco.2009.11.004.
Li, Y., and C. Foo. 2015.“A Sociological Theory of Corporate Finance: Societal Responsibility and Cost of Equity in
China.”Chinese Management Studies 9 (3): 269–294. doi:10.1108/CMS-12-2014-0232.
Li, Y., J. Zhang, and C. Foo. 2013.“Towards a Theory of Social Responsibility Reporting: Empirical Analysis of 613
CSR Reports by Listed Corporations in China.”Chinese Management Studies 7 (4): 519–534. doi:10.1108/CMS-
09-2013-0167.
Liu, M., Y. Shi, and Z. Wu. 2017.“Does Family Involvement Explain Why Corporate Social Responsibility Affects
Earnings Management?”Journal of Business Research 75: 8–16. doi:10.1016/j.jbusres.2017.02.001.
Marquis, C., and C. Qian. 2014.“Corporate Social Responsibility Reporting in China: Symbol or Substance?”
Organization Science 25 (1): 127–148. doi:10.1287/orsc.2013.0837.
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 19
Moon, J., and X. Shen. 2010.“CSR in China Research: Salience, Focus and Nature.”Journal of Business Ethics 94:
613–629. doi:10.1007/s10551-009-0341-4.
Narayanan, M. P. 1985.“Managerial Incentives for Short-Term Results.”The Journal of Finance 40 (5): 1469–1484.
doi:10.1111/j.1540-6261.1985.tb02395.x.
Orlitzky, M., F. L. Schmidt, and S. L. Rynes. 2003.“Corporate Social and Financial Performance: A Meta-Analysis.”
Organization Studies 24 (3): 403–444. doi:10.1177/0170840603024003910.
Pan, L. H., C. T. Lin, S. C. Lee, and K. C. Ho. 2015.“Information Ratings and Capital Structure.”Journal of
Corporate Finance 31: 17–32. doi:10.1016/j.jcorpfin.2015.01.011.
Phillips, R., R. E. Freeman, and A. C. Wicks. 2003.“What Stakeholder Theory Is Not.”Business Ethics Quarterly 13
(4): 479–502. doi:10.5840/beq200313434.
Porter, M. E., and M. R. Kramer. 2006.“Strategy & Society: The Link between Competitive Advantage and
Corporate Social Responsibility.”Harvard Business Review 84 (12): 78–92. doi:10.1007/s10726-006-9065-3.
Prior, D., J. Surroca, and J. Tribo. 2008.“Are Socially Responsible Managers Really Ethical? Exploring the
Relationship between Earnings Management and Corporate Social Responsibility.”Corporate Governance 16:
160–177. doi:10.1111/j.1467-8683.2008.00678.x.
Riccardo, L. 2002.“Inconsistency of Naive GMM Estimation for QR Models with Endogenous Regressors.”
Economics Letters 75 (2): 179–185. doi:10.1016/S0165-1765(01)00623-1.
Richardson, A. J., M. Welker, and I. R. Hutchinson. 1999.“Managing Capital Market Reactions to Corporate Social
Responsibility.”International Journal of Management Reviews 1 (1): 17–43. doi:10.1111/1468-2370.00003.
Roychowdhury, S. 2006.“Earnings Management through Real Activities Manipulation.”Journal of Accounting and
Economics 42 (3): 335–370. doi:10.1016/j.jacceco.2006.01.002.
Seo, M. H., and Y. Shin. 2016.“Dynamic Panels with Threshold Effect and Endogeneity.”Journal of Econometrics
195 (2): 169–186. doi:10.1016/j.jeconom.2016.03.005.
Stein, J. 1989.“Efficient Capital Market, Inefficient Firms: A Model of Myopic Corporate Behavior.”Quaterly
Journal of Economics 104 (4): 655–669. doi:10.2307/2937861.
Tran, K. C., and E. G. Tsionas. 2013.“GMM Estimation of Stochastic Frontier Model with Endogenous
Regressors.”Economics Letters 118 (1): 233–236. doi:10.1016/j.econlet.2012.10.028.
Wang, X., F. Cao, and K. Ye. 2016.“Mandatory Corporate Social Responsibility (CSR) Reporting and Financial
Reporting Quality: Evidence from a Quasi-Natural Experiment.”Journal of Business Ethics 1: 1–22. doi:10.1007/
s10551-016-3296-2.
Yang, S., F. He, Q. Zhu, and S. Li. 2018.“How Does Corporate Social Responsibility Change Capital Structure?”
Asia-Pacific Journal of Accounting & Economics 25 (3–4): 352–387. doi:10.1080/16081625.2017.1354710.
Zang, A. Y. 2012.“Evidence on the Trade-Offbetween Real Activities Manipulation and Accrual-Based Earnings
Management.”The Accounting Review 87 (2): 675–703. doi:10.2308/accr-10196.
Appendix A. CSR definitions (see Gong and Ho 2018)
Macrocosm (CSR_M)
Theme Series Item
Strategy M1 Responsibility strategy information: including the target and the
procedure of social responsibility strategy, the risks and the challenges of
crucial responsibility identification
M2 Sustainable development information: including the impacts of climate
changes, social problems, and macroeconomic environment changes on
corporate sustainable development
M3 Responsibility strategy and corporations efficient matched information:
including the impacts of products/services supplied by corporations on the
society and the environment.
M4 Management and responsibility strategy: including the corporate owners
(e.g., chairman) and corporate managers’(e.g., CEO) declarations about
social responsibility/sustainable development
M5 Establishment and achievement of social responsibility target: including
the long-term and short-term targets of corporate social responsibility, and
quantitative governance of the targets
Governance M6 Corporation’s fundamental information: including the fundamental
information contained in the firm and the industry that the firm belongs to,
and that in the social environment.
(Continued)
20 Y. GONG AND K.-C. HO
(Continued).
Macrocosm (CSR_M)
Theme Series Item
M7 Ethics value and principle information: including the understanding of
corporate social responsibility, and ethics value and behavioral principle on
the sustainable development.
M8 Social responsibility institutions’information: including the institutional
structure of the departments that supervise social responsibility, and the
staffinformation
M9 Decision procedure and structure information: including corporations’
management methods, procedures, and processes on environmental,
social, and economic issues.
M10 Governance transparency: including the mechanism of information
disclosure
M11 Risk management information: including the corporation evaluation and
risk management, especially the risk-related information about the
sustainable development
M12 Commercial moral governance: including the anti-commercial bribery
administration system
M13 Internal practical information: including the multi-department
participations of social responsibility, and the system and mechanism of
the group and parent company encouraging social responsibility in
subsidiaries.
Stakeholders M14 Identifications of stakeholders: including identifications of all stakeholders
and their importance
M15 Communications of stakeholders: including the mechanism of stakeholders’
communications and expected feedbacks.
M16 Suggestion from stakeholders: including the suggestion from different
stakeholders.
Content (CSR_C)
Economic Efficiency C1 Profits and returns information: including total annual incomes, profits,
and dividend proposal.
C2 Growth rate information: including the growth rate of total incomes, total
profits, and the changes of dividend proposal.
C3 Main products and services information: including the market occupations,
sales, and innovation of products and services.
Employees C4 Employees and employment information: including the general structure
of employees, and the information about each employment relationship.
C5 Employees’professional growth information: including the information
that corporations invest in employees’knowledgement and technique
improvement.
C6 Career healthy and safety information: including the risk management of
career safety and healthy, and the system of employees’safety and healthy
C7 Human rights protection information: including the information about fair
salary, no employment of child labor, and dealt with compliant.
C8 Working environment and social protection information: including the
information disclosure of salary level, holiday arrangement, other benefits
beyond salary, and care of handicapped employees
C9 Social conversations and cares information: including the employee union,
and employee-happiness-related activities and cares from union.
C10 Responsibility education information: including the understanding and
training of sustainable development strategy.
Environment C11 Environment management information: including the environment
management system operated in the corporation, the certificate from
environment management system, the annual aggregated investment on
environment.
C12 Pollution prevention information: including the identification of pollutions
and wastes; the measure, record, and report about pollutions; pollutions
controlling measures.
C13 Sustainable resources utilization information: including the identification
of energy and water resources, resources consumption, energy-saving
measures, and the feasible opportunity of the replacement of non-recycled
energy.
(Continued)
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 21
(Continued).
Macrocosm (CSR_M)
Theme Series Item
C14 Mitigation and adoption to climate change information: including the
measure, record, and report about the usage of greenhouse gases
emission; greenhouse gases reduction measurement; the avoidance of
negative impacts from climate change.
Fair Operation C15 Anti-corruption policy information: including the implementation and
improvement of anti-corruption policy and practice; the encouragement of
anti-corruption behaviors of employees and supply chain partners.
C16 Social responsibility promotion information: including the encouragement
of social responsibility, the promotion of social responsibility
knowledgement, and the investigation of social responsibility commitment.
Consumer C17 Products or services certificate information: including the quality
management system and innovations of products or services.
C18 Consumer management information: including consumer management
system and consumer satisfaction investigation.
C19 Consumer safety and health protection information: including the
qualification rate of products or services, the declaration of the products or
services safety, and the recycled system of products.
C20 Consumer services information: including the convenience of consumer
services, the complaint rate from consumer, and the complaint solution.
C21 Consumer personal information protection: including the authority setting
of personal information, and knowing whether the corporation has their
personal information and the application of information deletion.
C22 Consumer education information: including the consumer knowledgement
about products property, and consumer equity education.
Community C23 Donation information: including the total amount of donation and the
structure of donation (funding, material, and free professional service)
C24 Voluntary activities information: including the scope and performance of
voluntary activities.
C25 Political participation information: including the participation in regional
and industrial association, and the participation in related policies and
regulations settings.
C26 Create job opportunity: including the annual amount of new employees’
information.
C27 Science and Technology Development Information: including the
participation in the central and local government’s science projects and the
cooperation with university and research institution.
C28 Create wealth information: including the collection of community opinions
and supporting community enterprises.
C29 Healthy promotion: including the promotion of health awareness and the
reduction of negative impacts of products or services.
C30 Social investment information: including the screening of investment
environment, and the screening of investment society and governance.
Technique (CSR_T)
Contents Equilibrium T1 Integrity: stakeholders’responsibilities disclosure
T2 Moderate: negative information disclosure or the challenges and barriers
occurred in the procedure of disclosure
Information
Comparison
T3 Congruence: the consistency with past reports
T4 Quantification:the disclosure level of quantified information
Reports Innovation T5 Innovation: including structure innovation, writing and format innovation
T6 Innovation efficiency:Efficient impact of innovation on corporation, and the
probability of promotion at the industry level
T7 Stakeholders’suggestion disclosure
T8 Degree of third-party audit
T9 Authority of third-party institutions
T10 Mechanism efficiency of readers’suggestion and feedbacks
Normative T11 Reports normative: including report time, contents coverage,
announcement period, reality promise, etc.
T12 Reports standardize: including standard selection, and standard
transparency
(Continued)
22 Y. GONG AND K.-C. HO
(Continued).
Macrocosm (CSR_M)
Theme Series Item
T13 Reports seriousness: including mispronounced characters
Availability & information
transmission efficiency
T14 Reports in multi-language version
T15 Reports availability, considering disabled people
T16 Reports promotion by designing and formatting
T17 Graphic analysis of quantified information
Industry index (CSR_I)
Mining industry I1 Research and development and application of clean coal technology
I2 Restoration and management of ecological environment
I3 Research and development of renewable energy or new energy
I4 Establishment of fire prevention, explosion-proof management system
Communication and cultural
industry
I5 Sustainable development promotion through media
I6 Friendly information about media environment
Electricity, gas and water production
and supply industry
I7 Identification of the areas where the sustainable development policy
established.
I8 Security incident handling mechanism
I9 Transportation safety insurance mechanism
I10 Negative information on environmental violations
Electronic Industry I11 Heavy metal emissions, recycling management information
I12 Land pollution avoidance management information
I13 Solid waste, waste residue recycling management information
I14 Surrounding ecological environment governance information
Real estate industry I15 The policy of affordable housing development
I16 Compliance demolition information
I17 The policy of green real estate development
I18 Construction waste management
I19 The policy of noisy and light pollution
I20 Protection of migrant workers’rights and interests
Textiles, clothing, and fur
manufacturing industry
I21 Concerned about animal welfare information.
I22 Storage of hazardous chemicals
I23 Research and development of environmental protection equipment
I24 Research and development of low-carbon fabric
Machinery, equipment, and
instrument manufacturing
industry
I25 Storage of hazardous chemicals
I26 Solid waste, waste residue recycling management information
I27 Product safety consideration in the product design process
I28 Product recall mechanism
Construction industry I29 Project protection rate
I30 Qualification of construction enterprises
I31 Engineering responsibility subcontracting information
I32 Protection of natural habitats, wetlands, forests, wildlife corridors, agricultural
land in construction process
I33 Construction and demolition materials recycling and re-use information and
policy
Financial and insurance industry I34 The policy of fair designs and sales of financial products and services
I35 Products and services information compliance disclosure
I36 Anti-money laundering mechanism
I37 Evaluation of environmental and social risks
Transportation and storage industry I38 Sustainable transportation mechanism, transport mode conversion, and
transportation route planning.
I39 The purchase and use of environmentally friendly transportation
I40 Noisy management
I41 The use of renewable energy and measures to improve energy efficiency
Metal manufacturing industry I42 Hazardous chemicals management
I43 Wasted water, wasted gas, and solid waste recycling mechanism
I44 SO
2
, COD, hydroxide, and smoke dust pollutions and reduction information
I45 Noisy management
(Continued)
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 23
(Continued).
Macrocosm (CSR_M)
Theme Series Item
Wood furniture manufacturing
industry
I46 Reduction of chemical production, formaldehyde, and other harmful
substances
I47 Environmental certificate
I48 Sustainable development of forestry industry
I49 Raw materials recycling, reuse mechanism
Agriculture, forestry, animal
husbandry and fishery industry
I50 Land protection policy
I51 Livestock manure, fertilizers and pesticides management
I52 Research and development and application of Ecological agriculture, multi-
functional agriculture, renewable agricultural technology
I53 Support for agricultural machinery technology, chemical and biotechnology
innovation system
Wholesale, retail trade industry I54 Fair trade insurance mechanism
I55 Products information compliance disclosure
I56 Energy savings information and measures in storage logistics
I57 Problem product handling mechanism
Social services industry I58 High-quality services insurance system
I59 Services for disabled people
I60 Services safety information
I61 Expense transparency information
Petroleum, chemical, and plastics
industry
I62 Hazardous chemicals management
I63 Wasted gas, wasted water, and solid waste recycling system
I64 Toxic waste emission management
I65 Major chemical leakage accidents
Food and beverage industry I66 Proportion of products certified by a third party in line with the
internationally accepted food safety management standard system
I67 Products safety emergency response mechanism
I68 Problem food handling mechanism
I69 Raw materials safety and health controlling mechanism
I70 Products label and advertisement compliance information
Information technology industry I71 Technology innovation mechanism and achievements
I72 Services innovation mechanism and achievements
I73 Consideration of human rights in the information disclosure process and
promotion
Medicine and biological products
industry
I74 Research and development management
I75 Clinical trial management
I76 New medicine research and development information
I77 Medicine recycling mechanism
Paper industry I78 Water prevention and fire prevention mechanism
I79 Hazardous chemicals management
I80 Wasted water and solid wastes recycling measures and mechanism
24 Y. GONG AND K.-C. HO
Appendix B. Subsample Robustness Checks
B-1. Results excluding large firms
In this table, we exclude the largest 20% of firms from our sample to re-estimate our main regression specified in
equation (5). Panel A presents the results for all CSR disclosure firms, Panel B presents the results for mandatory
CSR disclosure firms, and Panel C presents the results for voluntary CSR disclosure firms. AQ is the accrual quality.
REM1 is the first index of real earnings management and REM2 is the second index of real earnings management.
CSR is the rating score of corporate social responsibility, ranging from 1 to 100. CSRs is the standardized CSR. SIZE
is the natural logarithms of total assets. CF is cash flow from operations to total assets and CFV is the standard
deviation of cash flow. DEBT is firm’s leverage ratio. LOSS is a dummy variable, equaling to 1 if operating income
is negative, 0 otherwise. AZ is the Altman’s Z-score. MB is the market to book ratio and ROA is return on assets.
T-statistics is reported in parentheses. ***, **, * denote statistical significance at the 1%, 5%, and 10% levels, based
on industry fixed effect and year fixed effect, respectively.
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Panel A: All CSR disclosure firms
constant −0.0214 −0.0296 0.3660*** 0.3199*** 0.7356*** 0.6858***
(−0.55) (−0.75) (3.75) (3.24) (7.57) (6.98)
CSR −0.0002* −0.0011*** −0.0011***
(−1.70) (−3.80) (−4.08)
CSRs −0.0023* −0.0129*** −0.0139***
(−1.71) (−3.80) (−4.12)
SIZE 0.0038*** 0.0038*** −0.0125*** −0.0125*** −0.0251*** −0.0251***
(2.41) (2.41) (−3.19) (−3.20) (−6.44) (−6.46)
CF 0.0083 0.0083 −0.3758*** −0.3759*** −0.9562*** −0.9562***
(0.48) (0.48) (−7.84) (−7.85) (−20.07) (−20.07)
CFV 0.0346*** 0.0347*** 0.0002 0.0003 −0.0130 −0.0129
(6.60) (6.60) (0.01) (0.02) (−0.86) (−0.85)
DEBT −0.0228*** −0.0228*** −0.0026 −0.0026 −0.0560*** −0.0560***
(−3.44) (−3.44) (−0.15) (−0.15) (−3.26) (−3.26)
LOSS 0.0088*** 0.0088*** 0.0241*** 0.0242*** 0.0031 0.0032
(2.42) (2.42) (2.47) (2.48) (0.32) (0.33)
AZ 0.0029*** 0.0029*** 0.0005 0.0005 0.0029*** 0.0029***
(9.08) (9.08) (0.58) (0.58) (3.28) (3.28)
MB 0.0002 0.0002 −0.0144*** −0.0144*** −0.0030 −0.0029
(0.18) (0.18) (−6.05) (−6.03) (−1.25) (−1.24)
RET −0.0170 −0.0170 −0.0954*** −0.0954*** −0.0272 −0.0272
(−1.63) (−1.63) (−3.42) (−3.42) (−0.98) (−0.98)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.14 0.14 0.25 0.25 0.37 0.37
Observations 1,785 1,785 2,223 2,223 2,223 2,223
Panel B: Mandatory CSR disclosure firms
constant −0.0845** −0.0974*** 0.0789 0.0214 0.3875*** 0.3241***
(−2.17) (−2.48) (0.62) (0.17) (3.18) (2.63)
CSR −0.0003*** −0.0014*** −0.0015***
(−2.98) (−4.36) (−4.88)
CSRs −0.0036*** −0.0165*** −0.0181***
(−2.95) (−4.31) (−4.89)
SIZE 0.0060*** 0.0060*** 0.0004 0.0002 −0.0101** −0.0102**
(3.93) (3.92) (0.08) (0.05) (−2.09) (−2.12)
CF 0.0168 0.0167 −0.3774*** −0.3778*** −0.9324*** −0.9327***
(0.94) (0.94) (−6.45) (−6.45) (−16.52) (−16.53)
CFV 0.0829*** 0.0830*** 0.0041 0.0046 0.0018 0.0023
(5.78) (5.78) (0.08) (0.10) (0.04) (0.05)
DEBT −0.0063 −0.0062 0.0223 0.0223 −0.0091 −0.0090
(−0.99) (−0.98) (1.07) (1.07) (−0.45) (−0.45)
LOSS 0.0071* 0.0071* 0.0300*** 0.0298*** 0.0229** 0.0228*
(1.93) (1.93) (2.47) (2.46) (1.96) (1.95)
AZ 0.0017*** 0.0017*** 0.0023** 0.0023** 0.0043*** 0.0043***
(Continued)
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 25
B-2. Results excluding SOEs
In this table, we exclude SOEs from our sample to re-estimate our main regression specified in equation (5).
Panel A reports the results for all CSR disclosure firms, Panel B presents the results for mandatory CSR disclosure
firms, and Panel C presents the results for voluntary CSR disclosure firm. AQ is the accrual quality. REM1 is the
first index of real earnings management and REM2 is the second index of real earnings management. CSR is the
rating score of corporate social responsibility, ranging from 1 to 100. CSRs is the standardized CSR. SIZE is the
natural logarithms of total assets. CF is cash flow from operations to total assets and CFV is the standard deviation
of cash flow. DEBT is firm’s leverage ratio. LOSS is a dummy variable, equaling to 1 if operating income is negative,
0 otherwise. AZ is the Altman’s Z-score. MB is the market to book ratio and ROA is return on assets. T-statistics is
reported in parentheses. ***, **, * denote statistical significance at the 1%, 5%, and 10% levels, based on industry
fixed effect and year fixed effect, respectively.
(Continued).
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
(5.33) (5.34) (2.12) (2.12) (4.19) (4.20)
MB 0.0010 0.0010 −0.0131*** −0.0130*** −0.0028 −0.0027
(1.10) (1.12) (−4.61) (−4.58) (−1.01) (−0.98)
RET −0.0088 −0.0089 −0.1862*** −0.1865*** −0.0405 −0.0407
(−0.68) (−0.68) (−4.43) (−4.44) (−1.00) (−1.00)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.17 0.17 0.34 0.34 0.46 0.46
Observations 1,285 1,285 1,422 1,422 1,422 1,422
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Panel A: CSR disclosure firms
constant 0.0154 0.0051 0.2576*** 0.2214*** 0.5032*** 0.4621***
(0.50) (0.16) (3.16) (2.65) (6.12) (5.48)
CSR −0.0002*** −0.0008*** −0.0009***
(−2.63) (−3.38) (−3.67)
CSRs −0.0028*** −0.0098*** −0.0109***
(−2.56) (−3.32) (−3.67)
SIZE 0.0020* 0.0020 −0.0080*** −0.0081*** −0.0156*** −0.0156***
(1.65) (1.63) (−2.49) (−2.50) (−4.79) (−4.78)
CF 0.0205 0.0205 −0.2674*** −0.2675*** −0.7947*** −0.7947***
(1.50) (1.50) (−6.73) (−6.73) (−19.82) (−19.82)
CFV 0.0296*** 0.0296*** 0.0013 0.0015 −0.0045 −0.0044
(6.33) (6.34) (0.09) (0.10) (−0.32) (−0.31)
DEBT −0.0193*** −0.0193*** −0.0125 −0.0124 −0.0598*** −0.0598***
(−3.13) (−3.13) (−0.74) (−0.74) (−3.51) (−3.50)
LOSS 0.0114*** 0.0114*** 0.0300*** 0.0301*** 0.0116 0.0117
(3.21) (3.22) (3.01) (3.02) (1.15) (1.16)
AZ 0.0007** 0.0007** −0.0017* −0.0017 0.0050*** 0.0050***
(2.05) (2.05) (−1.64) (−1.64) (4.65) (4.65)
MB 0.0020** 0.0020** −0.0175*** −0.0174*** −0.0075*** −0.0074***
(2.13) (2.15) (−7.40) (−7.38) (−3.13) (−3.12)
RET 0.0063 0.0063 −0.1393*** −0.1394*** −0.0885*** −0.0884***
(0.68) (0.67) (−5.29) (−5.29) (−3.33) (−3.33)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.13 0.13 0.26 0.26 0.37 0.37
Observations 1,820 1,820 2,293 2,293 2,293 2,293
Panel B: Mandatory CSR disclosure firms
constant −0.0450 −0.0575* −0.0104 −0.0568 0.1598 0.1041
(−1.44) (−1.79) (−0.10) (−0.54) (1.58) (1.01)
CSR −0.0003*** −0.0011*** −0.0012***
(−3.54) (−4.06) (−4.64)
(Continued)
26 Y. GONG AND K.-C. HO
(Continued).
AQ REM1 REM2
Dependent variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
CSRs −0.0034*** −0.0127*** −0.0147***
(−3.42) (−3.96) (−4.65)
SIZE 0.0041*** 0.0041*** 0.0035 0.0034 −0.0014 −0.0013
(3.33) (3.30) (0.87) (0.85) (−0.35) (−0.34)
CF 0.0245* 0.0245* −0.2241*** −0.2242*** −0.7446*** −0.7445***
(1.87) (1.87) (−5.05) (−5.05) (−16.95) (−16.95)
CFV 0.0387*** 0.0388*** 0.0057 0.0059 0.0211 0.0212
(4.57) (4.57) (0.20) (0.20) (0.74) (0.74)
DEBT −0.0048 −0.0047 0.0117 0.0118 −0.0115 −0.0113
(−0.80) (−0.80) (0.60) (0.60) (−0.59) (−0.58)
LOSS 0.0138*** 0.0138*** 0.0309*** 0.0309*** 0.0218* 0.0219*
(3.89) (3.89) (2.58) (2.58) (1.84) (1.85)
AZ 0.0003 0.0003 −0.0024** −0.0023** 0.0042*** 0.0042***
(0.86) (0.86) (−2.11) (−2.10) (3.80) (3.81)
MB 0.0025*** 0.0025*** −0.0160*** −0.0159*** −0.0063** −0.0063**
(2.86) (2.90) (−5.75) (−5.72) (−2.30) (−2.28)
RET 0.0287*** 0.0285*** −0.2725*** −0.2729*** −0.1668*** −0.1669***
(2.58) (2.56) (−7.44) (−7.44) (−4.60) (−4.60)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.19 0.19 0.34 0.34 0.44 0.44
Observations 1,398 1,398 1,586 1,586 1,586 1,586
Panel C: Voluntary CSR disclosure firms
constant 0.1102 0.1111 0.4814*** 0.4643*** 0.8086*** 0.7810***
(1.11) (1.11) (2.70) (2.58) (4.40) (4.20)
CSR 0.0000 −0.0004 −0.0006
(0.10) (−0.61) (−0.98)
CSRs 0.0002 −0.0046 −0.0073
(0.05) (−0.62) (−0.97)
SIZE −0.0011 −0.0011 −0.0188*** −0.0187*** −0.0292*** −0.0292***
(−0.26) (−0.26) (−2.58) (−2.58) (−3.90) (−3.90)
CF −0.0066 −0.0065 −0.3670*** −0.3669*** −0.9731*** −0.9730***
(−0.14) (−0.14) (−3.80) (−3.80) (−9.77) (−9.77)
CFV 0.0280*** 0.0280*** −0.0024 −0.0024 −0.0168 −0.0167
(3.77) (3.77) (−0.14) (−0.14) (−0.95) (−0.94)
DEBT −0.0658*** −0.0659*** −0.0383 −0.0383 −0.1205*** −0.1206***
(−3.46) (−3.46) (−1.13) (−1.13) (−3.43) (−3.43)
LOSS 0.0078 0.0078 0.0213 0.0214 −0.0135 −0.0133
(0.83) (0.83) (1.17) (1.18) (−0.72) (−0.71)
AZ 0.0025* 0.0025* 0.0015 0.0015 0.0067*** 0.0067***
(1.74) (1.74) (0.54) (0.54) (2.34) (2.34)
MB −0.0045 −0.0045 −0.0176*** −0.0176*** −0.0082* −0.0083*
(−1.36) (−1.36) (−3.72) (−3.72) (−1.68) (−1.69)
RET −0.0125 −0.0125 −0.0287 −0.0287 −0.0109 −0.0108
(−0.62) (−0.62) (−0.68) (−0.68) (−0.25) (−0.25)
Year FE YES YES YES YES YES YES
Industry FE YES YES YES YES YES YES
Adj. R
2
0.08 0.08 0.13 0.13 0.26 0.26
Observations 422 422 707 707 707 707
ASIA-PACIFIC JOURNAL OF ACCOUNTING & ECONOMICS 27
