Vol. 13(2), pp.

59-77, April-June 2021

DOI: 10.5897/JAT2020.0436
Article Number: 8686CB666438
ISSN 2141-6664
Copyright © 2021
Author(s) retain the copyright of this article Journal of Accounting and Taxation
http://www.academicjournals.org/JAT

Full Length Research Paper

The COVID-19 and earnings management:

China’s evidence
He Xiao and Jianqun Xi*
Division of Business and Management, BNU-HKBU United International College, 2000 Jintong Road, Tangjiawan,
Zhuhai, Guangdong, 519000 China.
Received 15 November, 2020; Accepted 23 March, 2021

This study investigated the relationship between the COVID-19 outbreak and the Chinese listed firms’
earnings management practices. It also examined how this relationship was moderated by the Chinese
listed firms’ corporate social responsibility (CSR) and the external corporate governance mechanism.
The data in this study were mainly retrieved from the China Stock Market and Accounting Research
(CSMAR) database and the Chinese Research Data Services Platform (CNRDS). The final sample
contained 2,029 A-share firms listed in the Shanghai and Shenzhen Stock Exchanges, which released
financial reports during the pandemic in 2020. The study applied the performance-adjusted Jones and
the modified Jones model to calculate accrual-based earnings. To estimate the real activity-based
earnings, this study used the following three measurements: The absolute value of the abnormal cash
flow from operations, the absolute value of the abnormal production costs, and the absolute value of
the abnormal discretionary expenditures. The results of this study indicated an increase in accrual-
based earnings management (AEM) and a significant decline in real activity-based earnings
management (REM), in firms in the most severely affected regions. In these regions, both AEM and REM
were less pronounced for the firms with a higher CSR performance than those with a lower CSR
performance. Moreover, firms audited by the Big 10 auditors were less likely to manipulate earnings
through AEM or REM.

Key words: COVID-19, earnings management, corporate social responsibility (CSR), big 10 auditors.

INTRODUCTION

The COVID-19 outbreak has exerted varied impacts on real activity-based earnings management (REM) (Graham
the financial market and firm financial performance (Ruiz et al., 2005; Kim et al., 2018). AEM occurs when
et al., 2020). Given the scenario and firm behavior during managers manipulate the accrual component of earnings,
financial crises, the firms are expected to manipulate their while REM occurs when managers manipulate real
reported earnings to respond to the current negative activities, such as when they reduce expenditures on
market environment (Choi et al., 2011). The manipulation research and development, which directly impacts the
of earning or earnings management is divided into two cash flow of the firms (Cimini, 2015).
streams-accrual-based earnings management (AEM) and The research on earnings management has focused on

*Corresponding author. E-mail: jianqunxi@uic.edu.cn, tigerhxiao@uic.edu.cn.

Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution
License 4.0 International License
60 J. Account. Taxation

AEM and REM, given that earnings are the sum of scenario.
accruals and operating cash flows. The empirical research
has shown substantial evidence that firms manage
earnings by manipulating various operational, investment, LITERATURE REVIEW AND RESEARCH
financial activities (Xu et al., 2007) and discretionary HYPOTHESES
expenditures (Roychowdhury, 2006). For instance, Baber
and Fairfield (1991) and Roychowdhury (2006) have Several studies have examined the managerial
investigated the manipulation of operational and manipulation of earnings (Healy, 1985; DeFond and
investment activities, particularly the manipulation of Jiambalvo, 1994; Subramanyam, 1996; Fields et al.,
discretionary expenditures. Studies have also examined 2001). For instance, Bartov et al. (2001) examined the
firms’ manipulation of production, inventory, and sales to relationship between discretionary accruals and audit
smooth earnings and meet earnings targets (Dhaliwal et qualifications and found a positive relationship between
al., 1994; Roychowdhury, 2006). In this regard, studies audit qualifications and abnormal accruals. In the context
have found that firms can use the income from the sales of financially distressed firms, Dimitropoulos and Asteriou
of long-term assets to smooth earnings and achieve (2010) investigated the influence of corporate governance
earnings forecasts (Bartov, 1993; Herrmann et al., 2003). mechanisms on managers’ behaviour and the accounting
Earnings management also occurs when managers use policies.
judgments in financial reporting (Healy and Wahlen, In the case of China, Chen et al. (2010) discovered that
1999). Dye (2002) defines such structuring of financially distressed companies adopt earnings
transactions as “classification manipulation.” Mittelstaedt management techniques to avoid the monitoring by the
et al. (1995), Comiskey and Mulford (1986), Imhoff and government. In the context of the European Union (EU),
Thomas (1988), and Ayers et al. (2002) are some studies Dimitras et al. (2015)’s study on the relationship between
that have examined earnings management by structuring the financial crisis and earnings management indicated
operational and investment transactions. that financially distressed companies audited by a Big 4
Earnings manipulation through financial activities has auditor exhibit lower discretionary accruals. Similarly,
been examined in regard to stock repurchases (Bens et Cimini (2015) concluded that, during a financial crisis, an
al., 2003; Hribar et al., 2006), the relationship between increase in conditional conservatism enhances the
earnings management and the use of stock options in the earnings quality and impairs earnings management.
compensation package (Matsunaga, 1995; Kimbrough During such crises, the stringent monitoring of auditors
and Louis, 2004; Carter et al., 2006), and financial contributed toward reducing earnings management, and
instruments (Hand et al., 1990; Barton, 2001; Pincus and thereby improved the quality of financial reporting.
Rajgopal, 2002), among others. Additionally, Marquardt According to Kim et al. (2018), managers can
and Wiedman (2005) have provided evidence on how manipulate reported earnings through AEM and REM.
firms engage in earnings management by structuring While AEM occurs before the announcement of financial
financial transactions. statements to inflate (or deflate) reported earnings, REM
The study contributes to the literature in the following occurs throughout the financial year because it requires
perspectives. First, despite the growing literature on changes to the firm’s operations (Zang, 2012). Hence,
market and firm reactions to the COVID-19 (Ding et al., REM is costlier than that of AEM. However, AEM is
2020; Ruiz et al., 2020), this is among the first studies to subject to greater scrutiny from external auditors or
examine Chinese listed firms’ earnings management regulators, and hence AEM is more easily detectable
practices in the context of the ongoing pandemic. than REM (Kim et al., 2018). The adoption of AEM and
Second, the empirical evidence provides inconclusive REM is also determined by the current pandemic
and contradictory views on the relationship between CSR scenario. The COVID-19 has significantly affected
and earnings management (EM) (Prior et al., 2008). Chinese listed firms’ economic activities. Given that REM
There is insufficient and limited research on the CSR-EM focuses on manipulating cash flows through operational,
relationship in China (Islam et al., 2015; Kim et al., 2018). investment, and financial activities throughout the
The current study extends prior research by re-examining financial year, REM is more difficult to conduct than that
the CSR-EM relationship in China, during the pandemic of AEM in the most severely affected regions. Hence, the
crisis. Third, it adds insights to the literature by examining first hypothesis is as follows:
how the relationship between the COVID-19 and the
Chinese listed firms’ earnings management practices is H1: During the COVID-19 outbreak, the listed firms in the
moderated by the CSR and the external corporate most severely affected regions are more likely to engage
governance mechanism (auditing by the Big 10 audit in manipulating accrual-based earnings than real activity-
firms) (DeFond and Subramanyam, 1998). Finally, this based earnings.
study helps Chinese authorities and listed firms better
understand the earnings management and CSR practices Concerning CSR and EM, there is inconclusive and
influencing the quality of financial reporting in the current mixed empirical evidence on the relationship between
 He and Jianqun 61

CSR and earnings management. While Scholtens and and Rama (2006) and Francis and Yu (2009) found that
Kang (2013), Kim et al. (2012), Bozzolan et al. (2015), the Big 4 auditors provide higher quality than that of the
and Martinez-Ferrero et al. (2015) found a negative non-Big 4 auditors. Contrarily, using the Korean data,
relationship between CSR and EM practices, Prior et al. Jeong and Rho (2004) concluded that there is no
(2008) documented a positive relationship, indicating difference in audit quality between the Big 6 and non-Big
management opportunistic behaviour. However, other 6 auditors.
studies have shown mixed results, indicating that the Studies on the relationship between the Big 6 auditing
CSR-EM relationship can be moderated by various firms and accruals have also revealed conservative
institutional factors (Chih et al., 2008; Yip et al., 2011; auditor behaviors (DeFond and Subramanyam, 1998;
Choi et al., 2013). Francis and Krishnan, 1999). For example, Becker et al.
Early studies on the relationship between CSR and EM (1998) found evidence of more conservatism in the
have mainly focused on AEM (Prior et al., 2008). For discretionary accruals of the clients of the Big 6 than
instance, Kim et al. (2012) found a negative relationship those of the non-Big 6 audit firms. Francis and Wang’s
between CSR and both AEM and REM-high CSR (2008) worldwide research on the audit quality provided
engagement reduces both AEM and REM. Lim and Choi by the Big 4 auditors between 1994 and 2004 showed
(2013) asserted that firms with good CSR activities that the Big 4 auditors provide higher earnings quality
constrain REM, focusing on the effect of the ethical and ensure stringent investor protection. Concerning the
implication of CSR on financial reporting. Bozzolan et al. financial crises, Butler et al. (2004) and Johl et al. (2007)
(2015) reported that enhanced CSR is likely to concluded that auditors play an important role in ensuring
discourage REM but encourage AEM because REM high quality in financial reports, especially during the
undermines firms’ long-term value. Conversely, Cohen et crisis periods. For instance, DeAngelo et al. (1994)
al. (2008) and Ewert and Wagenhofer (2005) argued that concluded that managers’ choices to manage earning
enhanced CSR may discourage AEM but encourage reflect their recognition of the financial difficulties of their
REM because the probability of detecting AEM will companies to avoid a violent discovery of losses. Further,
increase as Chinese firms provide more operational according to Chia et al. (2007), the Big 4 auditing
information in their CSR disclosure. companies contribute toward reducing earnings
Kim et al. (2018) discovered that Chinese firms’ management, especially during the financial crisis.
enhanced CSR decreases their EM practices. In China, since the Big 10 audit firms have greater
Conversely, SOEs and firms operating in more incentives to protect their integrity and reputation when
institutionally developed regions are more likely to conducting an audit on listed companies, they maintain a
engage in REM, while increasing their CSR activities. higher audit quality than that of the non-Big 10 audit firms
Therefore, listed firms with enhanced CSR are less likely (Chen et al., 2010). Therefore, the Big 10 audit firms take
to engage in earnings management practices through a more conservative approach when auditing listed
both AEM and REM; thus the second hypothesis as companies’ financial reports in this pandemic scenario.
follows: Thus, the following hypothesis is developed:

H2: During the COVID-19 outbreak, listed firms in the H3: During the COVID-19 outbreak, listed firms audited
most severely affected regions are less likely to by the Big 10 auditors are less likely to engage in
manipulate their earnings through both AEM and REM earnings manipulation through AEM or REM.
when increasing their CSR activities.

Concerning audit quality, the literature documents varied METHODOLOGY

audit quality between the Big 6 and non-Big 6 audit firms
Measurement of earnings management
(DeAngelo, 1981). In line with the study by DeAngelo
(1981) and Craswell et al. (1995) found the significant Based on Kothari et al. (2005) and Kim et al. (2018), the current
premiums earned by the Big 6 audit firms over the non- study used both the performance-adjusted Jones and modified
Big 6 auditing firms to be consistent with the Jones models to calculate the discretionary accruals. According to
differentiation in the audit quality. A large clientele base these models, a firm’s total accruals decompose into non-
and reputed brand name are incentives for the Big 6 discretionary and discretionary accruals. Specifically, discretionary
accruals were estimated by adopting the cross-sectional modified
auditing firms to detect and constrain earnings Jones model adjusted for performance.
management of their clients (MacDonald, 1997). Geiger The non-discretionary accruals were estimated as follows:

(1)

where Accruals = total accruals defined as the change in the non- and amortization; A= the total asset of the firms; ∆Sales = the
cash current assets minus the change in the current liabilities change in sales; ∆AR = the changes in net receivables; PPE = the
excluding the current portion of long-term debt minus depreciation property, plant, and equipment; and ROA = the net income divided
62 J. Account. Taxation

by the lagged total assets. calculate the discretionary accruals, drawn from Dechow et al.
Thus, the discretionary accruals were converted into absolute (1995). The model was rewritten by adding a modified sales change
values considered as a proxy for AEM. A higher AEM indicates that variable defined as , where is the change in the
firms engage in a higher level of EM through discretionary accruals. accounts receivable. The non-discretionary accruals of the modified
The study also applied the modified Jones model (MAEM) to Jones model are estimated as follows:

(2)

Since a firm can also manipulate its earnings through its operations
(Roychowdhury, 2006), the study estimated REM through the
following three REM measurements: the absolute value of the where indicates earnings management practices of the firm i,
abnormal cash flow from operations ( ), the absolute value of which includes two proxies-AEM and REM. indicates the
the abnormal production costs ( , and the absolute value of the number of provincial COVID-19 deaths scaled by the total number
abnormal discretionary expenditures ( , consistent with Kuo et of casualties in China. Consistent with the literature (Kim et al.,
al. (2014) and Bozzolan et al. (2015). The estimations of , , 2018), the study added the firm’s one-year lagged control variables
are detailed in Appendix I. including the firm age, leverage ratio, book-to-market ratio,
To estimate REM, we applied the following equation suggested institutional shareholding percentages, Big 10 shareholder’s
by Kim et al. (2018): ownership percentages, independent director percentages, Big 10
auditors, audit fee, cash flow growth rate, sales growth rate, ROE,
Tobin’s Q, and the natural logarithm of firm size. The study
(3) controlled the province-level characteristic by adding the legal
environment index in our model suggested by Wang et al. (2017).
Consistent with Kim et al. (2018), the study developed an The detailed definitions of variables are shown in Table 1. It also
aggregate proxy by multiplying -1 for CFO (cash flow from included the industry fixed effect and cluster the standard error at
operations) and ADE (abnormal discretionary expenditures) in the industry level to account for the arbitrary serial correlation
Equation 4. A high REM indicates that listed firms engage in a among industries.
higher level of EM through operational changes. The study also considered the heterogeneous impact of the
regional pandemic severity and various firm characteristics on firms’
earnings management practices. This was achieved by inserting
Empirical studies the firm corporate governance features into the following model:
where indicates two measurements-the CSR
To assess the impact of the regional pandemic severity on firm’s scores and the Big 10 auditors from the lagged year. In regions with
earnings management practices, the study applied the following varied pandemic severities, the coefficient of interest- -shows the
regression models to test the hypotheses:

(5)

moderation of the CSR performance or the Big 10 auditors on firm’s auditors. Table 2 lists the summary statistics of all the key variables.
earning management. All control variables in model 5 are
consistent with the control variables defined in model 4.
MAIN RESULTS
Descriptive statistics
The main regression results are presented in Table 3.
Data were mainly retrieved from the following two sources. First, the Columns 1 and 2 of Table 3 show that firms in the most
pandemic data, as a proxy of the relative provincial pandemic severely affected regions increased their discretionary
severity, from the China Stock Market & Accounting Research accruals by 34.1% (AEM) and 37.6% (MAEM), which
(CSMAR) database scaling the number of provincial COVID-19 were calculated by applying the Jones and Modified
deaths by the total number of casualties in China by May 31, 2020.
Jones models, respectively. However, in these regions,
The pandemic death proportion of an average province is around
2.15%. The data on the stock returns and firm characteristics were the REM experienced a statistically significant decline, as
also gathered from the CSMAR database. Second, CSR scores shown in Column 3. Quantitatively, a 1% increase in the
were obtained from the Chinese Research Data Services Platform proportion of pandemic deaths would decrease the REM
(CNRDS), which includes 819 firms’ CSR performances in 2019. by 1%. The finding is consistent with our Hypothesis 1
The major sample contains 2,029 A-share firms listed in the that the COVID-19 outbreak has exerted varied impacts
Shanghai and Shenzhen Stock Exchanges, which released
financial reports during the pandemic in 2020. The AEM, modified
on the earnings management practices of firms in the
AEM, and REM have the following means: 0.0109, 0.0110, and most severely affected regions (Dimitras et al., 2015).
0.0253, respectively. The average CSR score of the sampled firms These results indicate that firms engage in manipulating
is 6.591, and 53.327% of the sampled firms engage with Big 10 their accrual-based earnings by inflating their reported
 He and Jianqun 63

Table 1. Variable’s definitions.

Variable Definitions
AEM Performance-adjusted Discretionary Accruals based on Jones Model (Kothari et al., 2005)
MAEM Performance-adjusted Discretionary Accruals based on Modified Jones Model (Kothari et al., 2005)
REM Real activity-based earnings management
AEM_Choi Performance-adjusted Discretionary Accruals based on Modified Jones Model (Choi et al., 2011)
MAEM_ Choi Performance-adjusted Discretionary Accruals based on Modified Jones Model (Choi et al., 2011)
Spread_3 The firm’s bid-ask spread over event window [0, 3], as in Corwin and Schultz (2012)
Spread_30 The firm’s bid-ask spread over event window [0, 30], as in Corwin and Schultz (2012)
CSR The firm's CSR scores from the CNRDS
Big 10 A dummy variable that equals to one if the firm's auditor is a big ten auditing firm and zero otherwise
CFO Operating cash flows divided by lagged total assets
Sales growth Net sales growth rate
Cash growth Cash flow growth rate
Leverage The firm's total debt versus total asset
Book-to-Market ratio The firm's book value to its market value
Firm Size The natural logarithm of the firm's total asset
Tobin Q The ratio between a firm's physical asset's market value and its replacement value
ROA The firm's net income versus total asset
Age No. of years since the establishment of the firm
Board size Size of the board of the firm
Top 10 ownership The firm's largest ten shareholders' ownership percentage
Institutional ownership The firm's institutional investor share percentage
IND The no. of independent directors divided by the total number of directors
Audit fee The natural logarithm of annual audit fees that firms paid
Mortality No. of the provincial COVID-19 mortality scaled by the total death in China by May 31st, 2020.
No. of the provincial COVID-19 confirm cases scaled by the total confirm cases in China by May 31st,
Confirmed
2020
Legal environment Legal environment index of a province

earnings right before the announcement of the financial al., 2018). Columns 4 to 6 of Table 4 suggest that firms
statements. Since the outbreak of COVID-19 was an audited by the Big 10 auditors in the previous year are
unexpected event, the managers may have manipulated less likely to engage in manipulating earnings through
their firms’ earnings using AEM instead of REM, given either AEM or REM, which confirms Hypothesis 3. The
that that REM requires managers to manipulate the findings of the current study are in line with the research
operational, financial, and investment activities throughout conducted by Becker et al. (1998), who found similar
the financial year. Furthermore, the outbreak of COVID- evidence of a higher accounting conservatism of the Big
19 has severely affected the economy, and hence the 10 auditing firms than that of the non-Big 10 auditing
manipulation of cash flows through operational, financial, firms. They argue that the Big 10 auditing firms have
and investment activities in the most severely affected greater incentives to protect their integrity and reputation
regions became more difficult and expensive and easily against any threat of litigations when auditing listed
detectable. companies.
Table 4 exhibits the empirical results of two moderating
factors-CSR and external Big 10 auditors. Columns 1 to 3
show that, in the most severely affected regions, the Robustness check
earnings management practices (e.g., AEM and REM)
are less pronounced for the firms with a higher CSR Alternative proxy for earnings management and
performance, as measured by the CSR scores. This pandemic severities
finding is consistent with Hypothesis 2, which highlights
the influence of CSR on earnings management To confirm whether the results are robust, the study
manipulation. It is also supported by the finding that firms adopted the alternative proxies for measuring accrual-
with a higher CSR performance are less likely to based earnings management, consistent with Kim et al.
manipulate earnings through both AEM and REM (Kim et (2018). Thus, it estimated accruals by calculating the
64 J. Account. Taxation

Table 2. Summary statistics.

Variable Obs Mean Std. Dev. Min Max

AEM 2,029 0.109 0.448 0 13.634
MAEM 2,029 0.110 0.475 0 15.120
AEM_CHOI 2,029 0.092 0.574 0 17.393
MAEM_ CHOI 2,029 0.092 0.625 0 19.680
REM 2,029 -0.025 0.056 -0.738 0.473
Spread_3 2,029 0.002 0.009 -0.025 0.063
Spread_30 2,029 0.003 0.004 -0.012 0.027
Mortality 2,029 0.018 .128 0 0.972
Confirmed 2,029 0.024 .106 0 0.809
Leverage 2,029 0.440 .241 0.014 5.681
Book-to-Market Ratio 2,029 1.429 1.788 0.024 19.322
Sales growth 2,029 -7.300 43.260 -1510.222 0.039
institutional ownership 2,029 0.449 .244 0 0.974
IND 2,029 0.379 0.058 0.250 0.800
Big 10 2,029 0.533 0.499 0 1.000
Top 10 ownership 2,029 0.597 0.155 0 0.972
Firm Size 2,029 22.524 1.441 17.954 28.636
Age 2,029 3.000 .291 1.574 4.122
Cash growth 2,029 -0.157 .658 -13.122 0.917
Tobin's Q 2,029 1.870 1.854 0.692 41.971
ROA 2,029 0.033 0.228 -7.700 4.707
Audit fee 2,029 14.030 0.710 12.206 18.240
Legal environment 2,029 8.738 4.383 0.448 14.854
CSR 819 6.591 2.508 0 11.000

Table 3. The impact of regional pandemic severity on firms’ earnings management.

(1) (2) (3)

Variable
AEM MAEM REM
Mortality 0.341*** (0.022) 0.376*** (0.024) -0.01** (0.004)
Leverage 0.092** (0.037) 0.115*** (0.031) 0.053*** (0.004)
Book-to-Market ratio -0.008 (0.006) -0.008 (0.007) 0.002 (0.001)
Sales growth 0.001 (0.000) 0.000 (0.001) 0.001** (0.000)
Institutional ownership 0.050* (0.025) 0.052* (0.028) 0.021* (0.008)
IND 0.110 (0.129) 0.105 (0.136) -0.010 (0.040)
Big 10 -0.001 (0.016) 0.001 (0.017) -0.001 (0.003)
Top 10 ownership -0.134** (0.052) -0.140** (0.058) -0.043*** (0.008)
Firm Size -0.102*** (0.020) -0.104*** (0.022) -0.004* (0.002)
Age 0.034 (0.023) 0.035 (0.024) 0.008 (0.005)
Cash growth -0.024 (0.035) -0.033 (0.040) -0.018** (0.006)
Tobin Q 0.067** (0.030) 0.076** (0.034) -0.006** (0.002)
ROA 0.131*** (0.033) 0.149*** (0.028) -0.040** (0.011)
Audit fee 0.059** (0.028) 0.057* (0.028) 0.000 (0.000)
Legal environment -0.002 (0.001) -0.002 (0.002) -0.000 (0.001)
Industry FE Yes Yes Yes
Observations 2,029 2,029 2029
R-squared 0.239 0.245 0.134
Industry fixed effects are included in all estimations. Standard errors, clustered at the industry level, are shown in
brackets. The detailed definitions of variables are in Table I. *, **, and *** denote significance at the 10, 5 and 1%
level, respectively.
 He and Jianqun 65

Table 4. The impact of CSR performance and Big 10 auditor on firms’ earnings management.

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

Variable
AEM MAEM REM AEM MAEM REM
-0.010*** -0.010*** -0.047*** - - -
Mortality×CSR
(0.002) (0.003) (0.007) - - -

- - - -0.545*** -0.623*** -0.014***

Mortality×Big 10
- - - (0.028) (0.030) (0.004)

0.113*** 0.089*** -0.029 0.091** 0.113*** -0.029

Leverage
(0.032) (0.027) (0.018) (0.037) (0.030) (0.004)

0.001 -0.000 0.003** -0.008 -0.009 0.003*

Book-to-Market ratio
(0.003) (0.003) (0.002) (0.006) (0.007) (0.002)

0.000 -0.001 -0.001** 0.001 0.000 -0.001**

Sales growth
(0.000) (0.001) (0.000) (0.000) (0.001) (0.000)

Institutional -0.026 -0.005 0.014* 0.043 0.043 0.014*

ownership (0.026) (0.023) (0.008) (0.025) (0.028) (0.007)

-0.012 0.019 -0.003 0.094 0.087 -0.005

IND
(0.045) (0.050) (0.027) (0.128) (0.135) (0.026)

-0.012 -0.005 -0.027** -0.108** -0.110** -0.027**

Top 10 ownership
(0.034) (0.036) (0.013) (0.041) (0.043) (0.013)

-0.041*** -0.038*** -0.002 -0.101*** -0.103*** -0.001

Firm Size
(0.006) (0.006) (0.003) (0.020) (0.022) (0.003)

0.017 0.014 0.004 0.035 0.036 0.005

Age
(0.011) (0.011) (0.004) (0.023) (0.025) (0.004)

0.006 0.001 -0.017*** -0.022 -0.031 -0.017***

Cash growth
(0.005) (0.007) (0.006) (0.033) (0.038) (0.006)

0.049*** 0.044*** -0.005*** 0.067** 0.075** -0.005***

Tobin's Q
(0.016) (0.012) (0.001) (0.030) (0.034) (0.001)

0.089** 0.076*** -0.105*** 0.131*** 0.148*** -0.107***

ROA
(0.033) (0.026) (0.029) (0.033) (0.028) (0.028)

-0.002* -0.001* -0.000 -0.002 -0.002 -0.000

Legal environment
(0.001) (0.001) (0.000) (0.002) (0.002) (0.000)

0.014 0.011 -0.000 0.057* 0.055* -0.001

Audit fee
(0.010) (0.010) (0.004) (0.028) (0.030) (0.004)

0.004 0.005 0.000 NA NA NA

Big 10
(0.008) (0.007) (0.003) NA NA NA

Industry FE Yes Yes Yes Yes Yes Yes

Observations 815 815 815 2,029 2,029 2,029
R-squared 0.563 0.600 0.528 0.245 0.253 0.226
Industry fixed effects are included in all estimations. Standard errors, clustered at the industry level, are shown in brackets. The detailed
definitions of variables are in Table I. *, **, and *** denote significance at the 10, 5 and 1% level, respectively.
66 J. Account. Taxation

Table 5. Robustness Check 1: Alternative proxy for earnings management for the main results.

(1) (2)
Variable
AEM_ CHOI MAEM_ CHOI
0.535*** 0.620***
Mortality
(0.105) (0.119)

0.303 0.348
Leverage
(0.197) (0.218)

-0.017* -0.017*
Book-to-market ratio
(0.008) (0.009)

0.007 0.007
Sales growth
(0.005) (0.006)

0.033 0.027
Institutional ownership
(0.034) (0.037)

-0.389* -0.406*
IND
(0.209) (0.216)

-0.021 -0.021
Big 10
(0.019) (0.021)

-0.126* -0.138*
Top 10 ownership
(0.067) (0.075)

-0.111*** -0.112***
Firm Size
(0.038) (0.040)

0.020 0.019
Age
(0.026) (0.029)

-0.074 -0.086
Cash growth
(0.065) (0.074)

0.150 0.171
Tobin Q
(0.104) (0.121)

0.116 0.140
ROA
(0.132) (0.152)

0.075 0.073
Audit fee
(0.048) (0.049)

-0.002 -0.002
Legal environment
(0.002) (0.002)

Industry FE Yes Yes

Observations 2,029 2,029
R-squared 0.253 0.255

difference between a firm’s net income and net cash flow discretional accruals. The results shown in Tables 5 and
(Kim et al., 2018). Subsequently, this was followed by 6 (the alternative estimations of EM) are consistent with
inserting the newly estimated accruals into both the the main findings of the current study.
Jones and Modified Jones models to compute new Next, the study used the confirmed proportion of
 He and Jianqun 67

Table 6. Robustness Check 2: Alternative proxy for earnings management for the heterogeneous effect.

(1) (2) (3) (4)

Variable
AEM_CHOI MAEM_ CHOI AEM_CHOI MAEM_ CHOI
-0.019*** -0.019*** - -
Mortality×CSR
(0.005) (0.005) - -

- - -1.131*** -1.322***
Mortality×Big 10
- - (0.044) (0.047)

0.084* 0.079* 0.296 0.338

Leverage
(0.044) (0.044) (0.190) (0.209)

-0.007 -0.008 -0.016* -0.016

Book-to-market ratio
(0.006) (0.005) (0.009) (0.010)

-0.003 -0.002 0.007 0.008

Sales growth
(0.003) (0.003) (0.006) (0.006)

0.002 0.005 0.013 0.003

Institutional ownership
(0.024) (0.025) (0.043) (0.050)

-0.040 -0.029 -0.394* -0.413*

IND
(0.066) (0.062) (0.214) (0.223)

0.009 0.007 -0.091* -0.098*

Top 10 ownership
(0.027) (0.030) (0.051) (0.056)

-0.032*** -0.032*** -0.113*** -0.114***

Firm size
(0.006) (0.006) (0.039) (0.041)

0.016 0.017 0.019 0.019

Age
(0.013) (0.014) (0.026) (0.029)

-0.006 -0.008 -0.072 -0.083

Cash growth
(0.011) (0.011) (0.062) (0.070)

0.000 0.001 0.149 0.170

Tobin Q
(0.003) (0.003) (0.104) (0.120)

0.059 0.062* 0.116 0.139

ROA
(0.035) (0.032) (0.130) (0.149)

0.002 0.003 0.075 0.072

Audit fee
(0.010) (0.010) (0.048) (0.049)

-0.002 -0.002* -0.003 -0.002

Legal environment
(0.001) (0.001) (0.002) (0.002)

0.008 0.006 NA NA
Big 10
(0.008) (0.008) NA NA

Industry FE Yes Yes Yes Yes

Observations 815 815 2029 2029
R-squared 0.347 0.345 0.267 0.270
Industry fixed effects are included in all estimations. Standard errors, clustered at the industry level, are shown in brackets. The
detailed definitions of variables are in Table I. *, **, and *** denote significance at the 10, 5 and 1% level, respectively.
68 J. Account. Taxation

provincial COVID-19 deaths as an alternative proxy of the likely to engage in REM during the COVID-19 outbreak,
regional pandemic seriousness. This was consistent with which is consistent with the findings from Ding et al.
the previous findings of the current study as exhibited in (2007).
Tables 7 and 8.

The effect of the release of the financial report on the

Difference-in-differences approach information asymmetry during the pandemic period

To control for the temporal effect, the study applied a It is worth considering the effect of the release of the
difference-in-differences approach (DID) by including the financial report on information asymmetry during the
firm’s AEM and REM in the previous year (that is, 2019) COVID-19 outbreak, since previous findings of the
and comparing them with those made during the current study have shown that the firms have attempted
pandemic period. Specifically, the following DID model different earnings management practices in regions with
was used: different pandemic severities. Based on the previous
studies, such as Fong et al. (2017) and Nagar et al.
(2019), the study adopted the daily bid-ask spread as the
(6) proxy for information asymmetry, calculation of daily bid-
ask spread, drawn from Corwin and Schultz (2012).
where is a dummy variable and equals one if the Columns 1 and 2 of Table 12 show that, after the
financial report is announced in 2020, and zero financial disclosure, the short-term (3 days) information
otherwise. The industry fixed effect ( , province fixed asymmetry levels shrink and the longer term (30 days)
effects ( ), and year fixed effects ( ) are also included information asymmetry levels increase in firms in the
to relieve the possible time-invariant province most severely affected areas. These findings indicate the
characteristics and time-varying economic conditions short- and long-term effects of the EM practices on the
during the sampled periods. It must be noted that market-based information asymmetry levels during the
and are absorbed by the time and province fixed pandemic, which is in line with the conclusions in Choi et
effects, respectively, and are thus omitted in Equation 6. al. (2011). Columns 3 to 6 of Table 12 confirm that the
The standard error is clustered at the industry level. As CSR performance and Big 10 auditors can reduce both
presented in Table 9, the DID estimates (Postt × Deathi) the short- and long-term information asymmetries of firms
are still significant after controlling for the province and after the outbreak of a pandemic. This finding is supported
year fixed effects. Table 10 shows that all the previous by studies such as Cui et al. (2018) and Pittman and
moderating effects from CSR performance and Big 10 Fortin (2004).
auditor are robust to this alternative setting. Overall, the
baseline results hold when considering unobserved
regional and time fixed effects. Conclusion

This study investigated the relationship between the

DISCUSSION COVID-19 outbreak and the Chinese listed firms’
earnings management practices. It also examined how
The effect of ownership structure on the earnings this relationship was moderated by the listed firms’ CSR
management during the pandemic period and the external corporate governance mechanism.
Since the COVID-19 outbreak was unexpected, studies
The study explored the heterogeneous effect of the on COVID-19 have mainly focused on the market
ownership structure on the earnings management reactions to the pandemic. Currently, there is limited
practices of the listed firms. Columns 1 to 3 of Table 11 research on the relationship between the COVID-19
show that the firms with higher ownership concentration outbreak and earnings management practices, especially
are more likely to engage in AEM and less likely to use in China. Further, research conducted in China provides
REM during the COVID-19 outbreak, which is supported inconclusive and contradictory conclusions on the CSR-
by the findings from Kim and An (2018). This indicates EM relationship. In China, there is limited study on how
that the controlling shareholders’ cash flow rights are auditors moderate the impact from external shocks on
more concentrated and that they can expropriate the earnings management. To fill these gaps in the extant
value of minority shareholders based on higher AEM. literature, the current study provided some meaningful
However, REM might lead to more distortions in firms’ thoughts to the policymakers and academics for future
operations, and hence the controlling shareholders are considerations.
more likely to reduce the use of REM as the disparity is The results revealed that listed firms in the most
perceived to grow. In Table 11, Columns 4 to 6 show that severely affected regions were more likely to engage in
the SOEs are more likely to engage in AEM and less AEM and less likely to engage in REM. The additional
 He and Jianqun 69

Table 7. Robustness Check 3: Alternative proxy for the regional pandemic severity for the main results.

(1) (2) (3)

Variable
AEM MAEM REM
0.412*** 0.456*** -0.013***
Confirmed
(0.027) (0.029) (0.004)

0.093** 0.116*** 0.055***

Leverage
(0.037) (0.031) (0.013)

-0.009 -0.009 0.003*

Book-to-Market ratio
(0.006) (0.007) (0.002)

0.001 0.000 -0.001***

Sales growth
(0.000) (0.001) (0.000)

0.048* 0.050* 0.020**

Institutional ownership
(0.025) (0.028) (0.009)

0.105 0.099 -0.013

IND
(0.128) (0.135) (0.027)

-0.003 -0.000 -0.001

Big 10
(0.016) (0.017) (0.003)

-0.130** -0.136** -0.041**

Top 10 ownership
(0.051) (0.056) (0.016)

-0.101*** -0.103*** -0.004

Firm Size
(0.020) (0.022) (0.004)

0.035 0.035 0.009**

Age
(0.023) (0.024) (0.004)

-0.024 -0.033 -0.018***

Cash growth
(0.035) (0.040) (0.005)

0.067** 0.075** -0.007**

Tobin Q
(0.030) (0.034) (0.003)

0.131*** 0.149*** -0.052*

ROA
(0.033) (0.028) (0.028)

-0.003 -0.002 0.000

Legal environment
(0.001) (0.002) (0.000)

0.059** 0.057* -0.001

Audit fee
(0.028) (0.029) (0.005)

Industry FE Yes Yes Yes

Observations 2,029 2,029 2029
R-squared 0.240 0.246 0.355

analysis on the moderating effect of CSR and external auditors, on the baseline finding demonstrates that, in the
corporate governance, measured by the Chinese Big 10 most severely affected regions, both AEM and REM are
70 J. Account. Taxation

Table 8. Robustness Check 4: Alternative proxy for the regional pandemic severity for the heterogeneous effect.

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

Variable
AEM MAEM REM AEM MAEM REM
-0.012*** -0.012*** -0.057*** - - -
Confirmed×CSR
(0.003) (0.003) (0.009) - - -

- - - -0.663*** -0.759*** -0.016***

Confirmed×Big 10
- - - (0.038) (0.041) (0.005)

0.113*** 0.089*** -0.029 0.091** 0.113*** -0.029

Leverage
(0.032) (0.027) (0.018) (0.037) (0.030) (0.018)

0.001 -0.000 0.003** -0.008 -0.009 0.003*

Book-to-market ratio
(0.003) (0.003) (0.002) (0.006) (0.007) (0.002)

0.000 -0.001 -0.001** 0.001 0.000 -0.001**

Sales growth
(0.000) (0.001) (0.000) (0.000) (0.001) (0.000)

-0.026 -0.005 0.014* 0.043 0.044 0.014*

Institutional ownership
(0.026) (0.023) (0.008) (0.025) (0.028) (0.007)

-0.013 0.018 -0.003 0.093 0.086 -0.005

IND
(0.045) (0.050) (0.027) (0.128) (0.135) (0.026)

-0.012 -0.005 -0.027** -0.108** -0.110** -0.027**

Top 10 ownership
(0.034) (0.036) (0.013) (0.041) (0.043) (0.013)

-0.041*** -0.038*** -0.002 -0.102*** -0.104*** -0.001

Firm size
(0.006) (0.006) (0.003) (0.020) (0.022) (0.003)

0.017 0.014 0.004 0.035 0.036 0.005

Age
(0.011) (0.011) (0.004) (0.023) (0.025) (0.004)

0.006 0.001 -0.017*** -0.022 -0.031 -0.017***

Cash growth
(0.005) (0.007) (0.006) (0.033) (0.038) (0.006)

0.049*** 0.044*** -0.005*** 0.067** 0.075** -0.005***

Tobin Q
(0.016) (0.012) (0.001) (0.030) (0.034) (0.001)

0.089** 0.076*** -0.105*** 0.130*** 0.148*** -0.107***

ROA
(0.033) (0.026) (0.029) (0.033) (0.028) (0.028)

-0.002* -0.001* -0.000 -0.003* -0.002 -0.000

Legal environment
(0.001) (0.001) (0.000) (0.002) (0.002) (0.000)

0.014 0.011 -0.000 0.058* 0.055* -0.001

Audit Fee
(0.010) (0.010) (0.004) (0.029) (0.030) (0.004)

0.004 0.005 0.000 NA NA NA

Big 10
(0.008) (0.007) (0.003) NA NA NA

Industry FE Yes Yes Yes Yes Yes Yes

Observations 1,642 4,065 4,065 1,623 4,026 4,026
R-squared 0.186 0.215 0.213 0.279 0.233 0.231
 He and Jianqun 71

Table 9. Robustness Check 5: Difference-in-differences approach for the main results.

(1) (2) (3)

Variable
AEM MAEM REM
0.447*** 0.480*** 0.011**
Mortality*Post
(0.021) (0.013) (0.005)

0.241*** 0.187*** 0.040

Leverage
(0.061) (0.045) (0.025)

0.004 0.002 0.009***

Book-to-Market ratio
(0.005) (0.005) (0.003)

0.000 0.000 -0.000

Sales growth
(0.000) (0.000) (0.000)

0.032* 0.029 0.054***

Institutional ownership
(0.018) (0.021) (0.018)

0.118 0.149 -0.032

IND
(0.104) (0.119) (0.040)

-0.007 -0.002 -0.006

Big 10
(0.016) (0.014) (0.006)

-0.169*** -0.111*** -0.081**

Top 10 ownership
(0.040) (0.032) (0.031)

-0.138*** -0.119*** -0.001

Firm size
(0.017) (0.014) (0.005)

0.036 0.035 0.021**

Age
(0.021) (0.022) (0.009)

0.002 -0.005 -0.020***

Cash growth
(0.007) (0.005) (0.003)

0.055*** 0.058*** -0.014***

Tobin Q
(0.014) (0.016) (0.004)

0.195*** 0.238*** -0.730***

ROA
(0.047) (0.053) (0.118)

Industry FE Yes Yes Yes

Time FE Yes Yes Yes
Province FE Yes Yes Yes
Observations 4,028 4,028 4028
R-squared 0.237 0.253 0.159
Industry, Month×Year, and Province fixed effects are included in all estimations. Standard
errors, clustered at the industry level, are shown in brackets. The detailed definitions of
variables are in Table I. *, **, and *** denote significance at the 10, 5 and 1% level,
respectively.

less pronounced for the firms engaged in a high level of auditors would be less likely to manipulate earnings
CSR activities. Finally, firms audited by the Big 10 through either AEM or REM.
72 J. Account. Taxation

Table 10. Robustness Check 6: Difference-in-differences approach for the heterogeneous effect.

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

Variable
AEM MAEM REM AEM MAEM REM
-0.029*** -0.008*** -0.008*** - - -
Mortality×Post×CSR
(0.001) (0.001) (0.003) - - -

- - - -0.762*** -0.795*** -0.329***

Mortality×Post×Big 10
- - - (0.038) (0.025) (0.017)

0.191 0.138 0.125*** 0.240*** 0.186*** 0.084***

Leverage
(0.222) (0.051) (0.041) (0.062) (0.046) (0.019)

0.009* 0.006 0.003 0.004 0.002 0.013***

Book-to-Market ratio
(0.002) (0.005) (0.004) (0.005) (0.005) (0.003)

0.000 -0.000** -0.000 0.000 0.000 -0.002*

Sales growth
(0.000) (0.000) (0.001) (0.000) (0.000) (0.001)

-0.010* -0.011 0.030 0.030 0.027 0.086***

Institutional ownership
(0.003) (0.020) (0.036) (0.019) (0.021) (0.020)

-0.171 -0.012 -0.104* 0.114 0.144 -0.025

IND
(0.127) (0.037) (0.059) (0.103) (0.119) (0.048)

-0.002 0.006 -0.015 -0.025 -0.012 -0.008

Big 10
(0.007) (0.002) (0.013) (0.023) (0.017) (0.006)

-0.110 0.003 -0.045 -0.161*** -0.100*** -0.139***

Top 10 ownership
(0.084) (0.006) (0.045) (0.039) (0.027) (0.034)

-0.105 -0.080 0.012 -0.138*** -0.120*** -0.013

Firm Size
(0.061) (0.034) (0.009) (0.017) (0.014) (0.008)

0.000 0.014 0.013 0.036 0.035 0.021**

Age
(0.026) (0.006) (0.013) (0.021) (0.022) (0.009)

0.020 0.002 -0.026** 0.002 -0.005 -0.024***

Cash growth
(0.014) (0.001) (0.011) (0.007) (0.005) (0.004)

0.049*** 0.045*** -0.025*** 0.055*** 0.058*** -0.015***

Tobin Q
(0.003) (0.002) (0.009) (0.014) (0.016) (0.005)

-0.094 0.120 -0.241* 0.193*** 0.236*** -0.203*

ROA
(0.063) (0.049) (0.122) (0.046) (0.053) (0.116)

0.062 0.040 -0.023** 0.067*** 0.049*** 0.000

Audit fee
(0.044) (0.023) (0.011) (0.020) (0.018) (0.011)

Industry FE Yes Yes Yes Yes Yes Yes

Time FE Yes Yes Yes Yes Yes Yes
Province FE Yes Yes Yes Yes Yes Yes
Observations 1,639 1,639 1,639 4,028 4,028 4,028
R-squared 0.259 0.321 0.161 0.240 0.257 0.106
 He and Jianqun 73

Table 11. Discussion 1: The mitigating effect from ownership structure.

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

Variable
ADC1 ADC3 REM ADC1 ADC3 REM
-3.298*** -3.423*** 0.140*** - - -
Mortality×Post×Top 10 ownership
(0.053) (0.061) (0.039) - - -

- - - 0.756*** 0.831*** -0.020*

Mortality×Post ×SOE
- - - (0.027) (0.020) (0.003)

0.199 0.185* 0.090*** 0.235*** 0.180*** 0.037

Leverage
(0.151) (0.056) (0.018) (0.062) (0.045) (0.056)

0.006 0.001 0.014*** 0.004 0.002 0.009

Book-to-Market ratio
(0.010) (0.010) (0.003) (0.005) (0.005) (0.008)

0.000* 0.000 -0.001 0.000* 0.000 -0.001

Sales growth
(0.000) (0.000) (0.001) (0.000) (0.000) (0.001)

0.037 0.026 0.089*** 0.036 0.040 0.057

Institutional ownership
(0.020) (0.022) (0.019) (0.032) (0.028) (0.047)

0.118* 0.139* -0.024 0.125 0.155 -0.013

IND
(0.029) (0.037) (0.048) (0.105) (0.121) (0.004)

-0.006 0.002 -0.009 -0.006 0.000 -0.011

Big 10
(0.007) (0.007) (0.006) (0.016) (0.014) (0.010)

-0.137** -0.119*** -0.014* -0.138*** -0.119*** -0.013

Firm Size
(0.028) (0.009) (0.008) (0.017) (0.013) (0.011)

0.036** 0.036*** 0.023** 0.040* 0.041* 0.004

Age
(0.006) (0.003) (0.009) (0.021) (0.023) (0.001)

0.010 -0.004** -0.025*** 0.002 -0.005 -0.021

Cash growth
(0.010) (0.001) (0.005) (0.007) (0.005) (0.004)

0.054** 0.057** -0.017*** 0.054*** 0.057*** -0.013

Tobin Q
(0.009) (0.013) (0.005) (0.014) (0.016) (0.005)

0.253** 0.231* -0.109 0.194*** 0.236*** -0.193

ROA
(0.037) (0.064) (0.070) (0.047) (0.053) (0.067)

0.070*** 0.052** -0.001 0.068*** 0.050*** 0.007

Audit fee
(0.005) (0.008) (0.011) (0.018) (0.016) (0.007)

2.079* 1.809** 0.298*** 2.003*** 1.807*** 0.262

Constant
(0.517) (0.309) (0.087) (0.218) (0.168) (0.195)

Industry Fixed Effect Yes Yes Yes Yes Yes Yes

Time Fixed Effect Yes Yes Yes Yes Yes Yes
Province Fixed Effect Yes Yes Yes Yes Yes Yes
Observations 4,028 4,028 4,028 4,028 4,028 4,028
R-squared 0.243 0.264 0.104 0.241 0.260 0.084
Robust standard errors in parentheses. ***p<0.01, **p<0.05, *p<0.1. Industry fixed effects are included in all estimations. Standard
errors, clustered at the industry level, are shown in brackets. The detailed definitions of variables are in Table I. *, **, and *** denote
significance at the 10, 5 and 1% level, respectively.
74 J. Account. Taxation

Table 12. Discussion 2: The impact of pandemics on information asymmetry level.

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

Variable
Spread_3 Spread_30 Spread_3 Spread_30 Spread_3 Spread_30
-0.013*** 0.008*** - - - -
Mortality
(0.005) (0.002) - - - -

- - -0.011*** -0.003*** - -
Mortality*CSR
- - (0.004) (0.001) - -

- - - - -0.048*** -0.011***
Mortality*Big 10
- - - - (0.007) (0.004)

0.009 0.005 0.016 -0.006 0.009 0.005

Leverage
(0.010) (0.005) (0.025) (0.008) (0.010) (0.005)

-0.002 -0.001 0.001 -0.001* -0.002 -0.001

Book-to-Market ratio
(0.002) (0.001) (0.002) (0.001) (0.002) (0.001)

-0.000 -0.000 -0.000 0.000* -0.000 -0.000

Sales growth
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)

-0.019* -0.007 -0.033 0.005 -0.019* -0.007

Institutional ownership
(0.010) (0.006) (0.024) (0.008) (0.010) (0.006)

-0.034 0.019 -0.043 0.005 -0.035 0.019

IND
(0.034) (0.028) (0.042) (0.021) (0.034) (0.028)

-0.003 -0.003* -0.008 -0.003 -0.002 -0.003

Big 10
(0.004) (0.002) (0.006) (0.004) (0.004) (0.002)

-0.009 0.011 -0.015 -0.007 -0.007 0.012

Top 10 ownership
(0.017) (0.008) (0.027) (0.012) (0.016) (0.008)

0.002 0.003** 0.001 0.003 0.002 0.003**

Firm Size
(0.003) (0.001) (0.004) (0.002) (0.003) (0.001)

-0.007 0.004 -0.015 0.004 -0.007 0.004

Age
(0.008) (0.003) (0.013) (0.005) (0.008) (0.003)

-0.008** -0.000 -0.001 -0.002 -0.008** -0.000

Cash growth
(0.003) (0.002) (0.007) (0.005) (0.003) (0.002)

-0.000 0.000 0.004*** 0.003*** -0.000 0.000

Tobin Q
(0.002) (0.001) (0.001) (0.001) (0.002) (0.001)

0.025*** -0.003 0.036** -0.004 0.025*** -0.003

ROA
(0.006) (0.003) (0.017) (0.006) (0.006) (0.003)

0.006 -0.003 0.000 -0.002 0.006 -0.003

Audit fee
(0.005) (0.002) (0.005) (0.003) (0.005) (0.002)

0.001 0.000 0.001 -0.000 0.001 0.000

Legal environment
(0.001) (0.000) (0.001) (0.000) (0.001) (0.000)

Industry FE Yes Yes Yes Yes Yes Yes

Observations 2,029 2,029 851 851 2,029 2,029
R-squared 0.141 0.063 0.200 0.162 0.142 0.063
Industry fixed effects are included in all estimations. Standard errors, clustered at the industry level, are shown in brackets. The
detailed definitions of variables are in Table I. *, **, and *** denote significance at the 10, 5 and 1% level, respectively.
 He and Jianqun 75

The findings of this research could help Chinese and industry specializations. Journal of Accounting and Economics
20(3):297-322.
authorities, listed firms, and market investors gain more
Cui J, Jo H, Na H (2018). Does corporate social responsibility affect
understanding about earnings management practices information asymmetry?. Journal of Business Ethics 148(3):549-572.
during a negative shock and under various internal and DeAngelo L (1981). Auditor size and audit quality. Journal of Accounting
external factors influencing the quality of financial and Economics 3(3):183-199.
DeAngelo H, DeAngelo L, Skinner DJ (1994). Accounting choice in
reporting. In addition, the findings may also provide
troubled companies. Journal of Accounting and Economics
significant implications for earnings management for 17(1/2):113-143.
academics interested in other emerging markets. Dechow P, Sloan R, Sweeney A (1995). Detecting earnings
management. The Accounting Review 70:193-225.
DeFond M, Jiambalvo J (1994). Debt covenant violation and
manipulation of accruals. Journal of Accounting and Economics
CONFLICT OF INTERESTS 17(1/2):145-176.
DeFond ML, Subramanyam KR (1998). Auditor changes and
The authors have not declared any conflict of interests. discretionary accruals. Journal of Accounting and Economics
25(1):35-67.
Dhaliwal DS, Frankel M, Trezevant R (1994). The taxable and book
income motivations for a LIFO layer liquidation. Journal of Accounting
REFERENCES Research 32(2):278-289.
Dimitras AI, Kyriakou MI, Iatridis G (2015). Financial crisis, GDP
Ayers BC, Lefanowicz C, Robinson J (2002). Do firms purchase the variation and Earnings Management in Europe. Research in
pooling method? Review of Accounting Studies 7(l):5-32. International Business and Finance 34(2):338-354.
Baber WR, Fairfield PM (1991). The effect of concern about reported Dimitropoulos PE, Asteriou D (2010). The effect of board composition
income on discretionary spending decisions: The case of research on the informativeness and quality of annual earnings: Empirical
and development. The Accounting Review 66(6):818-829. evidence from Greece. Research in International Business and
Barton J (200l). Does the use of financial derivatives affect earnings Finance 24:190-205.
management decisions? The Accounting Review 76(1):1- 26. Ding H, Fan H, Lin S (2020). COVID-19, firm exposure, and firm value:
Bartov E (1993). The timing of asset sales and earnings manipulation. A tale of two lockdowns. Available at:
The Accounting Review 68(4):840-855. https://ssrn.com/abstract=3574401
Bartov E, Gul F, Tsui J (2001). Discretionary-accruals models and audit Ding Y, Zhang H, Zhang J (2007). Private vs state ownership and
qualifications. Journal of Accounting and Economics 30:421-452. earnings management: evidence from Chinese listed companies.
Becker C, DeFond M, Jiambalvo J, Subramanyam K (1998). The effect Corporate Governance: An International Review 15(2):223-238.
of audit quality on earnings management. Contemporary Accounting Dye RA (2002). Classifications manipulation and Nash accounting
Research 15(1):1-24. standards. Journal of Accounting Research 40(4):1125-1162.
Bens DA, Nagar A, Wong MHF (2003). Employee stock options, EPS Ewert R, Wagenhofer A (2005). Economic effects of tightening
dilution, and stock repurchase. Journal of Accounting and Economics accounting standards to restrict earnings management. The
36(1-3):51-90. Accounting Review 80(4):1101-1124.
Bozzolan S, Fabrizi M, Mallin CA, Michelon G (2015). Corporate social Fields T, Lys T, Vincent L (2001). Empirical research on accounting
responsibility and earnings quality: International evidence. The choice. Journal of Accounting and Economic 31(1/3):255-307.
International Journal of Accounting 50(4):361-396. Fong KY, Holden CW, Trzcinka CA (2017). What are the best liquidity
Butler M, Leone A, Willenborg M (2004). An empirical analysis of proxies for global research?. Review of Finance 21(4):1355-1401.
auditor reporting and its association with abnormal accruals. Journal Francis JR, Krishnan J (1999). Accounting accruals and auditor
of Accounting and Economic 37:139-165. reporting conservatism. Contemporary Accounting Research
Carter M, Lynch L, Tuna I (2006). The role of accounting in the design 16(1):135-165.
of CEO equity compensation, working paper, University of Francis J, Wang D (2008). The Joint Effect of Investor Protection and
Pennsylvania and University of Virginia. Big 4 Audits on Earnings Quality around the World. Contemporary
Chen Y, Chien C, Huang S (2010). An appraisal of financially distressed Accounting Research 25(1):157-191.
companies’ earnings management: evidence from listed companies Francis J, Yu M (2009). Big 4 Office Size and Audit Quality. The
in China. Pacific Accounting Review 22(1):22-41. Accounting Review 84(5):1521-1552.
Chia YM, Lapsley I, Lee HW (2007). Choice of auditors and earnings Geiger M, Rama D (2006). Audit Firm Size and Going-Concern
management during the Asian Financial Crisis. Managerial Auditing Reporting Accuracy. Accounting Horizons 20(1):1-17.
Journal 22(2):177-196. Graham JR, Harvey CR, Rajgopal S (2005). The economic implications
Chih H, Shen C, Kang F (2008). Corporate social responsibility, investor of corporate financial reporting. Journal of Accounting and Economics
protection, and earnings management: Some international evidence. 40(1):3-73.
Journal of Business Ethics 79(1/2):179-198. Hand J, Hughes PJ, Sefcik SE (1990). In substance defeasances:
Choi JH, Kim JB, Lee JJ (2011). Value relevance of discretionary Security price reactions and motivation. Journal of Accounting and
accruals in the Asian Financial Crisis of 1997–1998. Journal of Economics 13(1):47-89.
Accounting and Public Policy 30(2):166-187. Healy P (1985). The effect of bonus schemes on accounting decisions.
Choi BB, Lee D, Park Y (2013). Corporate social responsibility, Journal of Accounting and Economics 7(1/3):85-113.
corporate governance and earnings quality: Evidence from Korea. Healy PM, Wahlen JM (1999). A review of the earnings management
Corporate Governance: An International Review 21(5):447-467. literature and its implications for standard setting. Accounting
Cimini R (2015). How has the financial crisis affected earnings Horizons 13(4):365-383.
management? A European study. Applied Economics 47(3):302-317. Herrmann D, Inoue T, Thomas WB (2003). The sale of assets to
Cohen DA, Dey A, Lys TZ (2008). Real and accrual‐based earnings manage earnings in Japan. Journal of Accounting Research
management in the pre and post‐Sarbanes‐Oxley Periods. The 41(1):89-108.
Accounting Review 83(3):757-787. Hribar P, Jenkins NT, Johnson WB (2006). Stock repurchases as an
Comiskey E, Mulford CW (1986). Investment decisions and the equity earnings management device. Journal of Accounting and Economics
accounting standards. The Accounting Review 61(3):519-525. 41(1-2):3-27.
Corwin SA, Schultz P (2012). A simple way to estimate bid‐ask spreads Imhoff EA, Thomas JK (1988). Economic consequences of accounting
from daily high and low prices. The Journal of Finance 67(2):719-760. standards: The lease disclosure rule change. Journal of Accounting
Craswell A, Francis J, Taylor S (1995). Auditor brand name reputation and Economics 10(4):277-310.
76 J. Account. Taxation

Islam MA, Haque S, Dissanayake T, Leung P, Handley K (2015). Nagar V, Schoenfeld J, Wellman L (2019). The effect of economic
Corporate disclosure in relation to combating corporate bribery: A policy uncertainty on investor information asymmetry and
case study of two Chinese telecommunications companies. management disclosures. Journal of Accounting and Economics
Australian Accounting Review 25(3):309-326. 67(1):36-57.
Jeong S, Rho J (2004). Big Six auditors and audit quality: The Korean Pincus M, Rajgopal S (2002). The interaction between accrual
evidence. The International Journal of Accounting 39:175-196. management and hedging: Evidence from oil and gas firms. The
Johl S, Jubb C, Houghton K (2007). Earnings management and the Accounting Review 77(1):127-160.
audit opinion: evidence from Malaysia. Managerial Auditing Journal Pittman JA, Fortin S (2004). Auditor choice and the cost of debt capital
22(7):688-715. for newly public firms. Journal of Accounting and Economics
Kim Y, Park MS, Weir B (2012). Is earning quality associated with 37(1):113-136.
corporate social responsibility?. The Accounting Review 87(3):761- Prior D, Surroca J, Tribó JA (2008). Are socially responsible managers
796. really ethical? Exploring the relationship between earnings
Kim SH, An Y (2018). The effect of ownership-control disparity on the management and corporate social responsibility. Corporate
Chinese firm’s real activity earnings management. Pacific Accounting Governance: An International Review 16(3):160-177.
Review 30(4):482-499. Roychowdhury S (2006). Earnings management through real activities
Kim SH, Udawatte P, Yin J (2018). The effects of corporate social manipulation. Journal of Accounting and Economics 42(3):335-370.
responsibility on real and accrual‐based earnings management: Ruiz Estrada MA, Koutronas E, Lee M (2020). Stagpression: The
Evidence from China. Australian Accounting Review 29(3):580-594. economic and financial impact of Covid-19 Pandemic. Available at:
Kimbrough M, Louis H (2004). Financial reporting incentives, corporate https://papers.ssrn.com/sol3/papers.cfm
governance, and the trade-off among alternative forms of executive Scholtens B, Kang F (2013). Corporate social responsibility and
compensation, Working paper, Harvard Business School. earnings management: Evidence from Asian economics. Corporate
Kothari P, Leone A, Wasely C (2005). Performance matched Social Responsibility and Environmental Management 20(1):95-112.
discretionary accrual measures. Journal of Accounting and Subramanyam K (1996). The pricing of discretionary accruals. Journal
Economics 39(1):163-197. of Accounting and Economics 22(1):249-281.
Kuo J, Ning L, Song X (2014). The Real and Accrual-based earnings Wang X, Fan G, Yu J (2017). Marketization index of China's provinces:
management behaviors: Evidence from the split share structure NERI report 2016. Social Sciences Academic Press: Beijing, China.
reform in China. The International Journal of Accounting 49(1):101- Xu R, Taylor G, Dugan M (2007). Review of real earnings management
136. literature. Journal of Accounting Literature 26:195-228.
MacDonald E (1997). More accounting firms are dumping risky clients. Yip E, Staden C, Cahan S (2011). Corporate social responsibility
Wall Street Journal, 25 April. reporting and earnings management: The role of political costs.
Martinez-Ferrero J, Gallego-Alvarez I, Garcia-Sanchez ´ IM (2015). A Australasian Accounting Business and Finance 5(3):17-34.
bidirectional analysis of earnings management and corporate social Zang A (2012). Evidence on the trade-off between real activities
responsibility: The moderating effect of stakeholder and investor manipulation and accrual-based earnings management. The
protection. Australian Accounting Review 75(25):359-371. Accounting Review 87(2):675-703.
Marquardt C, Wiedman C (2005). Earnings management through
transaction structuring: Contingent convertible debt and diluted
earnings per share. Journal of Accounting Research 43(2):205-243.
Matsunaga SR (1995). The effect of financial reporting costs on the use
of employee stock options. The Accounting Review 70(1):1-26.
Mittelstaedt F, Nichols W, Regier P (1995). SFAS No. 106 and benefit
reductions in employer sponsored retiree health care plans. The
Accounting Review 70(4):535-556.
 He and Jianqun 77

APPENDIX I

Estimating the abnormal cash flow from operations

The following equation is used to estimate the normal level of cash flow from operations (CFO), based on all the firm-
year observations in the same industry.

( ) ( )

where CFO, TA, Sales, and represent a firm’s cash flow from operations, the total assets, net sales, and the
change in the net sales, respectively. Thus, abnormal CFO (that is, ab_CFO) is calculated by subtracting the actual
CFO from the normal CFO using the above estimation model.

Estimating the abnormal production levels

The following equation is used to estimate the normal production cost (PC), based on all the firm-year observations in
the same industry.

( ) ( ) ( ) ( )

Thus, abnormal PC (that is, ab_PC) is calculated by subtracting the actual PC from the normal PC using the above
estimation model.

Estimating the abnormal levels of discretionary expenditure

The following equation is used to estimate the normal level of discretionary expenditure (DE), based on all the firm-year
observations in the same industry.

( ) ( )

Thus, abnormal DE (that is, ab_DE) is calculated by subtracting the actual DE from the normal DE using the
aforementioned estimation model.