2.1. Akuntansi Manajemen 3

J. Account.
Public Policy 35 (2016) 567–591
Contents lists available at ScienceDirect
J. Account. Public Policy

journal homepage: www.elsevier.com/locate/jaccpubpol
Full length article
The substantial convergence of Chinese

accounting standards with IFRS and the
managerial pay-for-accounting performance
sensitivity of publicly listed Chinese firms
Bin Ke a,1, Yubo Li b,2, Hongqi Yuan c,⇑
a
Department of Accounting, Business School, National University of Singapore, Mochtar Riady Building, BIZ 1, # 07-53,
15 Kent Ridge Drive, Singapore 119245, Singapore
b
Division of Accounting, SHU-UTS SILC Business School, Shanghai University, 20 Chengzhong Road, JiaDing District,
Shanghai 201800, PR China
c
Department of Accountancy, School of Management, Fudan University, 670 Guoshun Road, Shanghai 200433, PR China
a b s t r a c t
We examine how China’s adoption of a new set of Chinese

Accounting Standards (CAS) that is substantially converged with
the IFRS affects the managerial pay-for-accounting performance
sensitivity of publicly listed Chinese firms. We find that central-
government-controlled firms adopted an economically significant
managerial pay-for-accounting performance sensitivity prior to
the new CAS adoption. However, the pay-for-accounting
performance sensitivity declined significantly after the new CAS
adoption, especially for central-government-controlled firms that
were more significantly affected by the new CAS adoption. We
do not find similar evidence for local-government-controlled firms
whose managers were not subject to economically significant pay-
for-accounting performance sensitivity prior to the new CAS adop-
tion. There is no evidence that the difference in results between the
two types of firms is due to confounding events. Overall, our results
suggest that China’s new CAS adoption reduces the stewardship
usefulness of financial reporting.
Ó 2016 Elsevier Inc. All rights reserved.
⇑ Corresponding author. Fax: +86 21 6564 3203.

E-mail addresses: bizk@nus.edu.sg (B. Ke), yuboli@shu.edu.cn (Y. Li), yuanhq@fudan.edu.cn (H. Yuan).
1
Fax: +65 6773 6493.
2
Fax: +86 21 6998 0017.
http://dx.doi.org/10.1016/j.jaccpubpol.2016.06.006
0278-4254/Ó 2016 Elsevier Inc. All rights reserved.
568 B. Ke et al. / J. Account. Public Policy 35 (2016) 567–591
1. Introduction
Effective on January 1, 2007, publicly traded Chinese firms were required to adopt a new set of Chi-
nese Accounting Standards (CAS) that is substantially converged with the International Financial
Reporting Standards (IFRS). The objective of this study is to examine how the mandatory adoption
of the new CAS affects the sensitivity of the CEO’s annual cash compensation to reported accounting
earnings for publicly traded Chinese firms. Because the majority of publicly listed Chinese firms are
controlled by the Chinese government, we limit our sample to state-controlled Chinese firms that
are only listed on the main board of the domestic stock exchanges (commonly referred to as A share
firms) to avoid complications associated with mixing firms with different ownership structures.3
It has been long recognized by investors, standard setters, and researchers (Christensen and
Demski, 2003; Watts, 2006; O’Connell, 2007) that the two primary objectives of financial reporting
are to provide financial information that is useful to investors in securities valuation, referred to as
valuation usefulness, or compensation and debt contracting, referred to as stewardship usefulness.
Prior research shows that valuation usefulness and stewardship usefulness are related but distinct
concepts (Gjesdal, 1981; Paul, 1992; Lambert, 2001; Bushman et al., 2006). For example, Lambert
(2001, p.41) provides a detailed discussion about the distinction between the two concepts and shows
analytically that in general the way information is aggregated for valuation purposes is not the same
way this information would be aggregated for incentive contracting purposes. Hence, a change in a
financial reporting system that increases the valuation usefulness does not necessarily increase the
stewardship usefulness or vice versa.
In recent years, both the International Accounting Standards Board (IASB) and the U.S. Financial
Accounting Standards Board (FASB) have gradually shifted their focus to valuation usefulness (e.g., fair
value accounting). In a proposed converged Conceptual Framework for financial reporting announced
in July 2005, the IASB and FASB argued that stewardship is a subset of (rather than being distinct from)
the overall financial reporting objective of decision-usefulness and therefore decided not to designate
stewardship as a separate financial reporting objective (IASB, 2005, para. 24). This decision turned out
to be controversial and generated heated discussions during the process of revising the IASB’s concep-
tual framework for financial reporting (IASB, 2006, 2008). In the final adopted version of the IASB’s
Conceptual Framework taking effect in September 2010 (IASB, 2010), the IASB continues to designate
decision-usefulness as the overall objective of financial reporting but acknowledges the importance of
stewardship in standard setting.
With the worldwide adoption of various modified versions of the IFRS, there is a growing concern
that failing to designate stewardship as a separate financial reporting objective may diminish the
overall usefulness of financial reporting (Ball, 2006; O’Connell, 2007; Sunder, 2009; Watts, 2003,
2006). For example, Watts (2003) argues that existing observed accounting practices in a country
are shaped by multiple institutional forces, including not only the valuation demand but also non-
valuation demands (e.g., debt and compensation contracting). Hence, a narrow focus on the valuation
usefulness of financial reporting in mandatory accounting regulations such as fair value accounting
could produce unintended negative consequences and impose significant costs on investors and the
economy. This concern is expected to be more acute in weak investor protection countries such as
China because financial markets in those countries are usually small and less developed, bank financ-
ing is more prevalent (Lee et al., 2013), and therefore investors’ demand for the valuation usefulness of
financial reporting is lower.
Given that IFRS are principles-based and thus allow managers significant discretion in financial
reporting, there is an additional concern that the implementation and enforcement of IFRS are of
low quality (Henderson and O’Brien, 2016). As a result, the quality of IFRS-based financial reports
may not be high from the perspectives of either valuation usefulness or stewardship usefulness or
3
Ke et al. (1999) show that managerial pay in closely held private firms is more likely based on subjective measures rather than
objective measures such as accounting earnings. Because most non-state-controlled Chinese firms have high ownership
concentration, it is not appropriate to mix non-state-controlled Chinese firms with state-controlled Chinese firms in the same
analysis.
B. Ke et al. / J. Account. Public Policy 35 (2016) 567–591 569
both. Again, this concern is more acute in weak investor protection countries because the legal
enforcement of laws and regulations is much weaker in these countries (Daske et al., 2008; Li, 2010).
There is a large and growing accounting literature devoted to assessing the consequences of
mandatory IFRS adoption around the world. Most studies in this literature focus on the valuation use-
fulness of financial reporting. As noted above, valuation usefulness and stewardship usefulness are
related but not identical concepts. Hence, one cannot use the results on the valuation usefulness of
financial reporting to draw conclusions about the stewardship usefulness of financial reporting or vice
versa. To our best knowledge, there is little published empirical research on the effect of mandatory
IFRS adoption on the stewardship usefulness of financial reporting in weak investor protection coun-
tries (see Section 2 for a detailed review of related research). In addition, the existing published
research on the economic consequences of mandatory IFRS adoption is largely limited to European
countries and not much is known about the economic consequences of mandatory IFRS adoption in
non-European countries, which tend to have weaker investor protection. Even much less is known
about the impact of mandatory IFRS adoption on state-controlled firms. We contribute to the IFRS lit-
erature by examining the effect of China’s new CAS adoption in 2007 on state-controlled Chinese
firms’ managerial pay-for-accounting performance sensitivity (pay-for-performance sensitivity for
brevity).
Besides the fact that it is a representative weak investor protection country, China is interesting to
study in itself because China is the largest developing country that has embraced IFRS. In addition, the
new CAS represent a major overhaul of China’s accounting standards championed by Chinese govern-
ment regulators. Hence, it is of great importance not only to China but also to IASB to understand the
intended and unintended consequences of China’s mandatory new CAS adoption.
Our sample of state-controlled Chinese firms comprises both central-government-controlled firms
and local-government-controlled firms. We analyze the two types of firms separately because we
expect the effect of the mandatory new CAS adoption on the managerial pay-for-performance sensi-
tivity to be stronger for central-government-controlled firms than for local-government-controlled
firms. As we explain in Section 3, central-government-controlled firms have a stronger incentive than
local-government-controlled firms to pursue shareholder value maximization during our sample per-
iod 2005–2009.4 For example, central-government-controlled firms were explicitly required by law to
adopt earnings-based managerial incentive compensation contracts, but local-government-controlled
firms were not. Further, central-government-controlled firms were explicitly required to use reported
accounting earnings to assess managerial performance, but local-government-controlled firms were
not. Therefore, we expect managers of central-government-controlled firms to face a greater pay-for-
performance sensitivity than managers of local-government-controlled firms in the pre-new CAS adop-
tion period 2005–2006. More importantly, consistent with the optimal incentive compensation contract-
ing theory (Holmstrom, 1979; Lambert and Larcker, 1987), if the new CAS adoption increases (decreases)
the stewardship usefulness of financial reporting, we should expect the managerial pay-for-performance
sensitivity to increase (decrease) more significantly for central-government-controlled firms than for
local-government-controlled firms in the post-new CAS adoption period 2008–2009.
Consistent with our predictions, we find that the managerial pay-for-performance sensitivity for
central-government-controlled firms is both statistically and economically significant and it is signif-
icantly larger than the sensitivity for local-government-controlled firms in the pre-new CAS adoption
period (the managerial pay-for-performance sensitivity for central-government-controlled firms is
more than three times as large as that for local-government-controlled firms), suggesting that man-
agers of central-government-controlled firms are more incentivized to maximize shareholder value
than managers of local-government-controlled firms. More importantly, there is a significant decline
in the managerial pay-for-performance sensitivity in the post-new CAS adoption period for central-
government-controlled firms but not for local-government-controlled firms. These findings are consis-
4
We do not wish to imply that shareholder value maximization is the sole objective of central-government-controlled Chinese
firms. In fact, the evidence in Du et al. (2012) suggests that central-government-controlled firms also have to bear some social
responsibilities such as job creation. Nevertheless, as we show later, we do find that central-government-controlled Chinese firms
experienced significant structural reforms around 2003 that have significantly increased managerial incentives to maximize
shareholder value.
tent with the hypothesis that the new CAS adoption reduces the stewardship usefulness of financial
information and therefore firms that are more oriented toward shareholder value maximization
(i.e., central-government-controlled firms) adjust downward the managerial pay-for-performance
sensitivity in the post-new CAS adoption period. Since we do not find similar effects for local-
government-controlled firms, it seems unlikely that the results for central-government-controlled
firms are attributable to confounding macro shocks that affect both types of state-controlled firms.
We next examine whether the observed effect of mandatory new CAS adoption on the managerial
pay-for-performance sensitivity is stronger for central-government-controlled firms that were more
significantly affected by the new CAS adoption. The answer is indeed yes for central-government-
controlled firms but we do not find a similar effect for local-government-controlled firms. These
results provide further corroborative evidence that the observed decline in central-government-
controlled firms’ managerial pay-for-performance sensitivity is likely due to the mandatory new
CAS adoption rather than something else.
We also explicitly identify the major potential confounding macro shocks that occurred during our
sample period and determine whether such confounding events can explain the difference in the man-
agerial pay-for-performance sensitivity between central-government-controlled firms and local-
government-controlled firms over the pre- and post-periods. In particular, we analyze the effect of
the split share structure reform that made all previously non-tradable shares freely tradable, the effect
of a CSRC regulation that permitted domestically listed Chinese firms to adopt stock option compen-
sation schemes, the effect of a central government stimulus plan in response to the 2008 global finan-
cial crisis, and the effect of concurrent increased regulatory enforcement of accounting standards. In
addition, we use the propensity score matching method to better match central-government-
controlled firms with local-government-controlled firms. None of these robustness checks alters our
inferences.
Overall, we conclude that China’s new CAS adoption reduces the managerial pay-for-performance
sensitivity for central-government-controlled Chinese firms. This conclusion is complementary to
Ozkan et al. (2012) who find that the mandatory IFRS adoption increases the managerial pay-for-
performance sensitivity for firms from the Continental Europe, a region with stronger investor protec-
tion. This difference in results suggests that the effects of mandatory IFRS adoption on the stewardship
usefulness of financial reporting are different in weak versus strong investor protection countries.
The rest of the paper is organized as follows. Section 2 discusses the institutional background and
relevant research. Section 3 develops our research hypotheses. Section 4 presents the research design
for the hypotheses. Section 5 describes the sample selection procedures and descriptive statistics. Sec-
tion 6 discusses the regression results. Section 7 performs a battery of sensitivity checks to rule out
confounding macro shocks. Section 8 concludes.
2. Institutional background and related research
2.1. Institutional background
Taking effect on January 1, 2007 for all publicly traded firms listed on the two domestic stock
exchanges, the newly adopted Chinese Accounting Standards (CAS) substantially differ from the old
CAS. The new CAS comprise a basic standard, which is akin to a conceptual framework, and 38 specific
standards that address nearly all the issues covered in the International Financial Reporting Standards
(IFRS). Among the 38 standards, 16 are revisions of previously existing standards and 22 are newly
created. Except for certain modifications that reflect China’s unique circumstances and environment,
the new CAS are substantially converged with IFRS.5 Deloitte Touche Tohmatsu (2006) indicates that
the new CAS are not simply an expansion of the disclosure requirements, but change the primary basis
of accounting in mainland China. A key distinction between the old CAS and the new CAS is the adoption
5
Examples of the major differences between the new accounting standards and IFRS are the definition and disclosure of related
party transactions, business combinations of entities under common control, and reversal of impairment losses for long-lived
assets (see Deloitte Touche Tohmatsu, 2006).
of fair value accounting under the new CAS. In addition, the new CAS provide firm managers with
increased discretion in financial reporting, such as R&D, goodwill, revenue recognition, and fair value
measurement.6
In order to ensure the successful implementation of the new CAS, Chinese regulators introduced a
series of supporting regulations that took effect over the period 2007–2009. The most noteworthy
include the following: the new risk-based auditing standards that are similar to International Stan-
dards on Auditing issued by the International Auditing and Assurance Standards Board, Guides to
Chinese Accounting Standards, and Basic Standards on Firms’ Internal Control. In addition, starting
from 2006 multiple government agencies (including the Ministry of Finance, the China Securities
Regulatory Commission (CSRC), and China Institute of CPA) organized large-scale training on the
new CAS. Furthermore, the CSRC treated listed firms’ implementation of the new CAS as one of its
key examination areas during the years 2006–2008.
Because of the substantial improvement in the quality of accounting standards and concurrent
increased regulatory enforcement, Chinese regulators expressed optimism that the newly adopted
CAS would result in a significant improvement in Chinese firms’ financial reporting quality. The
new CAS have also won some international recognition. In December 2007, the Hong Kong Institute
of CPA (HKICPA) recognized the new CAS as equivalent to Hong Kong Financial Reporting Standards
(HKFRS), which are identical to IFRS. In December 2010, the Hong Kong Stock Exchange decided to
allow mainland-incorporated companies listed in Hong Kong to have an option to present financial
statements using the new CAS and audited by an approved mainland audit firm. In addition, in
November 2008 the European Securities Committee (ESC) voted to permit Chinese issuers to use
the new CAS when they enter the EU market without adjusting financial statements in accordance
with IFRS endorsed by the EU.
However, there are also doubts whether the mandatory new CAS adoption can result in desirable
economic consequences due to mainland China’s weak investor protection and poor record of enforce-
ment of government regulations. The concern is that firm managers may abuse the increased discre-
tion under the new CAS by engaging in opportunistic earnings management. For example, in a
research report on the first-year implementation of the new accounting standards, Ministry of
Finance Accounting Division (2008) identified a series of severe problems, including preparers’ lack
of understanding and intentional violation of the new accounting standards. Furthermore, it is unclear
how an increased emphasis of valuation usefulness in the new CAS (e.g., fair value accounting) would
affect the stewardship usefulness of the new CAS. Overall, it is an open question whether China’s
mandatory new CAS adoption would result in increased stewardship usefulness.
2.2. Related research
Even though more than 100 jurisdictions have either fully adopted various modified versions of
IFRS or are committed to doing so in the near future (see http://www.iasplus.com), the costs and ben-
efits of the global mandatory IFRS adoption are still hotly debated. Proponents claim that the world-
wide mandatory IFRS adoption reduces the costs that multinational firms incur in preparing and
auditing their financial statements, increases firms’ financial reporting quality (e.g., the comparability
of financial information across countries), and decreases firms’ cost of capital. However, opponents
argue that the quality of firms’ financial statements depends on not only high-quality accounting stan-
dards but also firms’ legal and institutional environments and the incentives of managers, investors,
and auditors (Ball et al., 2000; Ball et al., 2003; Hung, 2000). Because a country’s legal and institutional
environments and the incentives of relevant decision makers are often slow to change, opponents pre-
dict that the quality of financial reporting is unlikely to converge around the world even with the glo-
bal mandatory adoption of IFRS. Some commentators even predict that if standard setters ignore such
fundamental economic forces when proposing new accounting standards, severe unintended negative
consequences would result for not only investors and managers but also standard setters (Watts,
2006).
6
See Deloitte Touche Tohmatsu (2006) for a detailed discussion of the differences between the old CAS and new CAS.
In response to this important and ongoing debate, there is a growing accounting literature that ana-
lyzes the consequences of the worldwide mandatory IFRS adoption. Most empirical studies focus on
the valuation usefulness of financial reporting.7 After surveying the existing literature, Brüggemann
et al. (2013) conclude that the effect of mandatory IFRS adoption on the stewardship usefulness of finan-
cial reporting is an under-researched area.
Most recent research has started to investigate the impact of IFRS adoption on the contracting (e.g.,
debt or compensation) usefulness of accounting numbers. For example, using a sample of new debt
issues between 2001 and 2010 in twenty-two IFRS-adopting countries and twenty-one non-IFRS
countries, Ball et al. (2015) find a significant reduction in accounting-based debt covenants following
mandatory IFRS adoption.
With regard to managerial compensation, there are a few recent studies that start to analyze the
effect of mandatory IFRS adoption on managerial compensation.8 Voulgaris et al. (2014) finds that
UK firms reduce the use of Earnings-per-Share (EPS)-based performance measures in CEO equity-
based pay contracts post mandatory IFRS adoption. Wu and Zhang (2011) use a sample of Continental
European firms to examine the effect of mandatory IFRS adoption on the sensitivity of CEO turnover to
foreign peers’ accounting performance. They find that the mandatory IFRS adoption results in an
increased use of relative performance evaluation based on foreign peers’ accounting information, consis-
tent with greater financial reporting comparability associated with the mandatory IFRS adoption. Though
not the focus of their study, Wu and Zhang (2011) find no evidence that mandatory IFRS adoption is sig-
nificantly associated with the sensitivity of CEO turnover to their own firms’ accounting performance. In
contrast to Wu and Zhang (2011), Ozkan et al. (2012) find evidence based on a similar sample of Conti-
nental European firms that mandatory IFRS adoption results in an increased sensitivity of managerial
cash compensation to own firms’ accounting performance for countries with a large difference between
IFRS and their pre-adoption local accounting standards.
One common thread of the preceding studies is that they all cover countries or markets with strong
investor protection. For example, Kaufmann et al.’s (2009) investor protection (rule of law) rating in
2007, the first year of China’s mandatory new CAS adoption, is 1.69 for United Kingdom, 1.44 for Hong
Kong, and 1.75 for the median of the 16 European countries used in Wu and Zhang (2011) and Ozkan
et al. (2012).9 These ratings are high given that Kaufmann et al.’s (2009) investor protection ratings fol-
low a normal distribution with a mean of zero and a standard deviation of one. In contrast, China’s inves-
tor protection rating was only 0.45 in 2007 per Kaufmann et al. (2009). Prior research shows that the
effects of mandatory IFRS adoption are different for weak versus strong investor protection countries
(Daske et al., 2008; Byard et al., 2011). Hence, it is difficult to determine ex ante whether the evidence
from the prior studies based on strong investor protection countries can be generalized to weak investor
protection countries such as China. An important contribution of our study is to provide empirical evi-
dence relevant to the debate on the effect of mandatory IFRS adoption on the stewardship usefulness of
financial reporting in an important emerging market economy. Our research question is important and
timely in light of the ongoing debate on the costs and benefits of mandatory IFRS adoption around the
world.
Academic research on the costs and benefits of China’s mandatory new CAS adoption is still at its
infancy. The results from this research are mixed. For example, Liu et al. (2011) find that China’s new
CAS adoption is associated with decreased earnings management and increased value relevance of
accounting measures. Lee et al. (2013) also find evidence of increase value relevance of earnings post
7
The most common issues examined by this literature include accounting quality (e.g., Cascino and Gassen, 2015; Ahmed et al.,
2013; Christensen et al., 2015; Capkun et al., 2016 users of financial statements such as analysts (e.g., Byard et al., 2011; Tan et al.,
2011) and institutional investors (e.g., DeFond et al., 2011), value relevance of accounting information (e.g., Aharony et al., 2010a;
Barth et al., 2014; Wu and Zhang, 2014; Landsman et al., 2012; Beuselinck et al., 2010), and cost of capital (e.g., Li, 2010; Daske
et al., 2008). Brüggemann et al. (2013) conduct a thorough review of this literature and conclude that the overall evidence is mixed.
8
Wu and Zhang (2009) use a sample of Continental European firms to examine the effect of voluntary IFRS adoption on the
sensitivity of CEO turnover to accounting performance. However, the inference from their study cannot be readily extended to
mandatory IFRS adoption because forced IFRS adopting firms are likely to face different incentives than voluntary adopting firms
(see Ball et al., 2003).
9
Except for Italy (0.41), Greece (0.69), and Portugal (0.95), the rule of law rating for the remaining 13 European countries is
always above 1.
China’s new CAS adoption, but the effect is weaker for firms controlled by the central government. In
contrast, He et al. (2012) find evidence of earnings management to avoid reporting losses by managers
of listed Chinese firms by using the discretion in fair value accounting under the new CAS. To our
knowledge, no study from reputable academic journals has directly examined the effect of China’s
mandatory new CAS adoption on the stewardship usefulness of financial reporting.
Hou et al. (2014) have also examined the effect of the new CAS on managerial pay-for-performance
sensitivity using all listed Chinese firms over 2004–2009 but they reached a conclusion opposite to
ours. Unfortunately, Hou et al. suffer from a few important limitations, making their results difficult
to interpret. First, they use the combined compensation of the top three highest paid executives or
directors as the dependent variable and hence their results could be due to the change in the mix
of the top three highest paid executives or directors over time. Second, they only perform a simple
comparison of the managerial pay-for-performance sensitivity for the entire sample pre versus post
the new CAS adoption. They also do not control for many important time-variant determinants of
the managerial pay-for-performance sensitivity. Third, they use an unbalanced sample of listed Chi-
nese firms that change significantly in the pre and post periods. Therefore, their documented effect
of the new CAS adoption on the managerial pay-for-performance sensitivity is significantly contami-
nated by confounding events. As noted below, our research design avoids all of these limitations.
3. Hypothesis development
In this section we explain why managerial incentives differ for central-government-controlled

firms versus local-government-controlled firms during our sample period 2005–2009. Then, we dis-
cuss how such differences may affect the effect of the new CAS adoption on the managerial pay-
for-performance sensitivity for central-government-controlled firms versus local-government-
controlled firms.
Historically all state-owned Chinese firms operated their businesses according to state planning.
Managers of these firms suffered from severe agency problems and the notion of shareholder value
did not exist. Accordingly, there was little managerial pay-for-performance sensitivity (Li, 2000;
Wei, 2000). However, starting from the early nineties, China’s central government began to modernize
state-owned firms by encouraging them to be incorporated and publicly listed on domestic and inter-
national stock exchanges. In particular, since the formation of the State-owned Assets Supervision and
Administration Commission (SASAC), a ministry-level central government agency, as the ultimate con-
trolling shareholder of central-government-controlled firms in 2003, the SASAC has implemented
numerous regulations to strength the governance of central-government-controlled firms in order
to encourage managers to increase shareholder value. One important reform initiative undertaken
by the SASAC during our sample period is the implementation of two regulations that require a rigor-
ous annual evaluation of managerial performance and an explicit link between managerial annual
compensation and reported accounting earnings (SASAC, 2003, 2004). Furthermore, since 2004 the
SASAC has adopted a tough policy of eliminating small and underperforming central-government-
controlled firms through forced mergers and acquisitions. As a result of this policy, the number of
central-government-controlled firms shrank from 196 in 2003 to 129 in 2009. Because of these reform
initiatives, we expect an economically significant managerial pay-for-performance sensitivity for
central-government-controlled firms during the pre-new CAS adoption period 2005–2006. We state
this prediction in the following hypothesis10:
Hypothesis 1 (H1). The managerial pay-for-performance sensitivity is economically significant for

central-government-controlled firms prior to the mandatory new CAS adoption.
10
Of course a cynic may argue that all government-controlled Chinese firms are merely tools controlled by politicians to serve
their own self-interests rather than maximize shareholder value. This argument implies no or weaker managerial pay-for-
performance sensitivity in all government-controlled Chinese firms, opposite to H1. Our empirical results support H1 but not this
alternative prediction.
However, managers of local-government-controlled A share firms are likely to have much weaker
incentives to maximize shareholder value. One important reason is that the reform of local-
government-controlled firms significantly lagged behind the reform of central-government-
controlled firms. Due to the complexities of reforming state-owned enterprises, the central govern-
ment’s many reform initiatives often started with central-government-controlled firms, which were
also strategically more important to the central government (Ma, 2009). In addition, because local-
government-controlled firms were under the direct supervision of local governments rather than
the SASAC, the SASAC did not have significant influence over matters of local-government-
controlled Chinese firms. All of the SASAC’s reform initiatives since 2003 were merely advisory to
local-government-controlled firms (SASAC, 2003). Local governments had the explicit discretion to
design their own regulations that tailor to the specific situations of their local-government-
controlled firms. Even if some central government regulations directly applied to local-government-
controlled firms, such regulations were often difficult to enforce at the local level due to weak rule
of law and the divergence of interests between local governments and the central government (see
Batson, 2010 and the discussion below).11 Hence, while central-government-controlled firms were sub-
ject to the monitoring of multiple central government agencies including the SASAC and the National
Audit Office, Chen et al. (2009) indicate that managers of local-government-controlled firms were usually
subject to no monitoring at all. Therefore, managers of local-government-controlled firms were often
entrenched and had the ability to set their own compensation, which is unlikely to be very sensitive
to firm performance.
Even in Shanghai, which is at the fore front of China’s economic development, the reform of local-
government-controlled firms significantly lagged behind the reform of central-government-controlled
firms. For example, we directly confirmed with an anonymous official working for the local-SASAC of
the Shanghai Municipal Government that Shanghai-government-controlled firms had not followed the
SASAC regulations on managerial performance evaluation (SASAC, 2003, 2004).12 Some non-listed
Shanghai-government-controlled firms hadn’t even established a formal system of managerial perfor-
mance evaluation and compensation. Quite often, many Shanghai-government-controlled firms followed
ad hoc managerial evaluation and compensation practices such as ‘‘different compensation policies for
different firms” and ‘‘different compensation contracts for different managers” (the SASAC of Shanghai
Putuo District, 2006).
Another important reason for the weaker pay-for-performance sensitivity in local-government-
controlled firms is that local government leaders have fixed tenures and their future career prospects
directly depend on the local GDP growth during their tenures (Li and Zhou, 2005; Zhou, 2004). In
addition, local governments have to shoulder a variety of local social responsibilities such as employ-
ment and social welfare (Lin et al., 2004). Hence, local government leaders have a strong incentive to
pressure managers of local-government-controlled firms to pursue the growth of firm size and
employment rather than profit maximization. Consistent with this prediction, prior research shows
that local-government-controlled firms are more prone to overinvestment (Zhang and Wang,
2010), having excess employees on their payrolls (Zeng and Chen, 2006), executing related party
transactions that hurt minority shareholders’ interests (Cheung et al., 2010; Jiang et al., 2010), and
reporting lower accounting earnings (Chen et al., 2009). In order to induce managers of local-
government-controlled firms to follow the preferences of local government leaders, we expect the
latter to have little incentives to tie managerial pay to firm performance in an economically signifi-
cant way.
In summary, the preceding discussion suggests that managers of local-government-controlled
firms should face a weaker pay-for-performance sensitivity than managers of central-government-
controlled firms during the pre-new CAS adoption period. We state this prediction in the following
hypothesis:
11
China’s runaway real estate development and ballooning local government borrowing in recent years illustrate the difficulty
that the central government has in controlling the behavior of local governments (Li, 2009).
12
Even though they share similar names, local-SASACs and the SASAC (a ministry-level central government agency located in
Beijing) are not directly related. Local-SASACs are controlled by local governments and are not under the supervision of the SASAC.
Hypothesis 2 (H2). The managerial pay-for-performance sensitivity is greater for central-

government-controlled firms than for local-government-controlled firms prior to the mandatory
new CAS adoption.
How does the mandatory new CAS adoption affect the managerial pay-for-performance sensitivity
for local-government-controlled firms versus central-government-controlled firms? Assuming that a
firm’s objective is to induce corporate managers to take actions that will maximize shareholder value,
the optimal compensation contracting theory (Holmstrom, 1979; Lambert and Larcker, 1987; Banker
and Datar, 1989) hypothesizes that the relative weight placed on a performance measure in a compen-
sation contract is an increasing function of its ‘‘signal-to-noise” ratio with respect to the unobservable
managers’ actions. Consistent with this hypothesis, Lambert and Larcker (1987) find that U.S. firms
place relatively more weight on market performance (and less weight on accounting performance)
in compensation contracts when the variance of the accounting measure of performance is high rel-
ative to the variance of the market measure of performance. Following this line of reasoning, the effect
of the mandatory new CAS adoption on the managerial pay-for-performance depends on whether the
reported accounting earnings under the new CAS are more informative about managers’ actions (i.e.,
stewardship usefulness) than the accounting earnings under the old CAS.
As noted in the Introduction, there is an ongoing debate on the effect of China’s new CAS adoption
on the stewardship usefulness of reported earnings. Proponents of China’s new CAS argue that the new
CAS is more principles-based and therefore should have a better ability than the old CAS to capture a
firm’s underlying economic fundamentals. Hence, similar to the argument in Ozkan et al. (2012) in the
case of mandatory IFRS adoption in Europe, the new CAS earnings could be more informative about
managers’ actions. On the other hand, opponents counter that many changes reflected in the new
CAS earnings (e.g., fair value accounting) are often beyond managers’ control and therefore reported
new CAS earnings would become noisier as a measure of managers’ actions. There is also a debate
on whether the increased managerial discretion in financial reporting under the more principles-
based new CAS makes reported accounting earnings more informative about managers’ actions. This
concern is especially acute in a market like China with weaker institutions because managers could
abuse the increased discretion available under the new CAS by manipulating reported earnings to ben-
efit themselves at the expense of shareholder value (He et al., 2012). Because of multiple conflicting
forces, the net effect of the mandatory new CAS adoption on the stewardship usefulness of financial
reporting is unclear ex ante.
Our next two hypotheses shed direct light on this debate by empirically examining the impact of
China’s new CAS adoption on the managerial pay-for-performance sensitivity. Recall that H1 predicts
managers of central-government-controlled firms to face an economically significant pay-for-
performance sensitivity. Therefore, if the mandatory new CAS adoption reduces (increases) the stew-
ardship usefulness of financial reporting, the optimal incentive compensation contracting theory
would predict the managerial pay-for-performance sensitivity to decrease (increase) in the post-
new CAS adoption period 2008–2009 for central-government-controlled firms. We state this predic-
tion in the following hypothesis:
Hypothesis 3 (H3). To the extent that the mandatory new CAS adoption reduces (increases) the
stewardship usefulness of financial reporting, the managerial pay-for-performance sensitivity is
expected to decrease (increase) for central-government-controlled firms in the post-new CAS
adoption period.
Our discussions preceding H2 suggest that managers of local-government-controlled firms face lit-
tle economically significant pay-for-performance sensitivity in the pre-new CAS adoption period. In
addition, H2 predicts that managers of local-government-controlled firms face a weaker pay-for-
performance sensitivity than managers of central-government-controlled firms prior to the new
CAS adoption. Therefore, we also expect the mandatory new CAS adoption to have a smaller impact
on the managerial pay-for-performance sensitivity for local-government-controlled firms. We sum-
marize this prediction using the following hypothesis:
Hypothesis 4 (H4). To the extent that the mandatory new CAS adoption reduces (increases) the
stewardship usefulness of financial reporting, the managerial pay-for-performance sensitivity is
expected to decrease (increase) to a greater extent in the post-new CAS adoption period for central-
government-controlled firms than for local-government-controlled firms.
Our hypothesis development is based on the implicit assumption that central-government-

controlled Chinese firms care about the impact of the new CAS adoption on managerial incentives.
We find anecdotal evidence consistent with this assumption. Specifically, in early 2008 prior to the
fiscal year 2007 (the year of the new CAS adoption) annual report filing deadline, the SASAC issued
a supplemental regulation that explicitly permitted but did not mandate central-government-
controlled firms to exclude earnings related to trading securities and available-for-sale securities from
the performance benchmarks in the annual managerial performance evaluation (SASAC, 2008). This
evidence suggests that the SASAC was concerned about the potential negative impact of China’s
mandatory new CAS adoption on the stewardship usefulness of financial reporting. On the other hand,
we find no evidence based on interviewing a few senior managers working for local-government-
controlled firms in Shanghai that the mandatory new CAS adoption had a significant impact on man-
agerial compensation for local-government-controlled firms.
4. Research design
We use the following baseline firm fixed effects regression model to test the effect of China’s new
CAS adoption on the managerial pay-for-performance sensitivity for state-controlled Chinese firms
(firm and year subscripts are omitted for brevity):
LnðCOMPÞ ¼ b0 þ b1 RET þ b2 ROA þ b3 CONTROL þ b4 RET CONTROL þ b5 ROA
CONTROL þ c0 POST þ c1 POST RET þ c2 POST ROA þ c3 POST
CONTROL þ c4 POST RET CONTROL þ c5 POST ROA CONTROL þ e ð1Þ
See the Appendix A for all variable definitions.
The dependent variable COMP includes the CEO’s annual cash compensation (salary and bonus).13
Prior U.S. research (e.g., Core et al., 2003) indicates that a typical U.S. firm CEO’s pay-for-performance
sensitivity comes from three major sources: (a) annual cash compensation; (b) equity holdings; and
(c) turnover. Typically a U.S. CEO’s pay-for-performance sensitivity from equity holdings dwarfs the
CEO’s pay-for-performance sensitivity from cash compensation. Hence, a narrow focus on a U.S. CEO’s
cash compensation would understate the CEO’s total pay-for-performance sensitivity and may result
in erroneous inferences. We do not suffer from this problem in our setting because due to legal restric-
tions, equity-based incentives such as stock option compensation are rarely used in state-controlled A
share firms during our sample period. In addition, Ke et al. (2012) find no evidence that state-
controlled A share firms’ CEO turnover is sensitive to firm performance. Therefore, we can simply rely
on model (1) to capture the bulk of the CEO’s pay-for-performance sensitivity for state-controlled A share
firms.14
The key variables of interest are ROA and POST ROA. As noted in the Introduction, reported
accounting earnings were explicitly used in state-controlled Chinese firms’ managerial compensation
contracts during our sample period, consistent with the importance of stewardship role of accounting
performance measures. The coefficient on ROA captures the managerial pay-for-performance sensitiv-
ity for the pre-new CAS adoption period. The coefficient on POST ROA captures the incremental
change in the coefficient on ROA in the post-new CAS adoption period.
13
In addition to the CEO, another important player in government-controlled Chinese firms is the Chairman of the Board.
However, the compensation for a significant portion of the board chairmen is missing because their compensation is paid by the
parent company of the listed firm and therefore is not required to be disclosed in annual reports.
14
State-controlled Chinese firm executives often enjoy significant perks (Cai et al., 2011). We do not include perks in COMP
because data on perks are not readily available. In addition, the provision of perks is typically associated with an executive’s job
title rather than his performance. Therefore, we believe that including perks should not significantly affect our inferences.
As detailed in Sections 6 and 7, we adopt several complementary approaches to demonstrate the

causal effect of the new CAS adoption on the managerial pay-for-performance sensitivity of state-
controlled Chinese firms. First, we compare the managerial pay-for-performance sensitivity before
versus after the new CAS adoption for central-government-controlled firms versus local-
government-controlled firms. As noted in the previous section, local-government-controlled firms’
managerial pay-for-performance sensitivity is expected to be less affected by the new CAS adoption.
On the other hand, both types of state-controlled firms may face similar confounding macro shocks in
our sample period. Hence, we can use the difference in the coefficient on POST ROA for central-
government-controlled firms versus local-government-controlled firms to identify the effect of the
new CAS adoption and control for confounding macro shocks that affect both types of firms. Second,
to demonstrate more directly the effect of the new CAS adoption, we examine whether the predicted
effects of H3 and H4 are stronger for firms more significantly affected by the new CAS adoption. Third,
we explicitly identify the major macro events that occurred around the new CAS adoption and
examine whether such macro events can explain the predicted effects of H3 and H4. Finally, we
re-estimate our primary regression results using a smaller but better matched sample of central-
government-controlled firms and local-government-controlled firms using the propensity score
matching method.
Consistent with U.S. research, we include stock return (RET) as an additional control for firm per-
formance. Prior research suggests that the CEO’s pay-for-performance sensitivity varies with both
time-invariant and time-variant firm characteristics. Hence, we include CONTROL and its interactions
with RET and ROA. Following prior research (Davilla and Penalva, 2006; Leone et al., 2006; Ke et al.,
2012), CONTROL contains ln(ASSETS), GROWTH, LEV, and LARGEST_OWN.15 We allow all the coeffi-
cients to vary with POST because the accounting variables in the pre- and post-IFRS adoption periods
are based on different accounting standards and thus the coefficients on the regression variables may
change across the two periods. To reduce the influence of outliers, all continuous variables are win-
sorized at the top and bottom one percentile. To avoid multicollinearity and facilitate the interpretation
of the regression coefficients, we demean all the continuous explanatory variables for all the regression
models.
Finally, we include firm fixed effects in the regression to control for unobservable correlated omit-
ted variables. Due to inherent differences in the nature of the business across firms, the type of hired
CEO and accordingly the level of the CEO’s pay likely differ across firms. Hence, it is important to use
firm fixed effects to control for such cross-firm differences in order to increase the estimation effi-
ciency of the regression model and avoid the correlated omitted variable problem.16
5. The sample and descriptive statistics
Table 1 details our sample selection procedures for state-controlled A share firms. A listed firm is
defined to be state-controlled if the immediate largest shareholder of the listed firm is a state-owned
enterprise or a government agency. The immediate largest shareholder of listed Chinese firms is typ-
ically the controlling shareholder because Chen et al. (2009) show that most Chinese firms are dom-
inated by a single largest shareholder whose ownership far exceeds that of the second largest
shareholder. We further divide the sample of state-controlled firms into those controlled by the cen-
tral government and those controlled by local governments.
We limit our sample to 2005–2009 because the SASAC’s regulations on managerial compensation
of central-government-controlled firms were the same during this period. For example, reported
accounting earnings were explicitly required as a primary performance measure in all executive com-
15
We also include a dummy variable for negative ROA as part of CONTROL and obtain similar inferences.
16
An alternative approach to controlling for firm fixed effects is to estimate regression model (1) using a change (first difference)
specification (Wooldridge, 2002). We do not follow this alternative approach because our panel data set is unbalanced and
therefore a change specification would result in a significant loss of approximately 40% of the unique firms used in regression
model (1). Untabulated regression results for the change specification show that the coefficient on POST DROA is still
significantly negative for central SOEs while the coefficient on POST DROA is insignificant for local SOEs. However, we find no
significant difference in the coefficient on POSTDROA for central SOEs versus local SOEs.
Table 1
Sample selection procedures for state-controlled A share firms.
Initial sample of A share firms over the period 2005–2006 and 2008–2009 (firm years) 5999
Exclude non-state-controlled firms and firms with missing ultimate controlling shareholder data (2276)
Exclude firms in the financial industry (55)
Exclude firms dually listed in Hong Kong (163)
Exclude firm years with CEO turnover (700)
Exclude firm years with missing data (145)
Exclude firms that do not have data in both the pre- and post-new CAS adoption periods (401)
Exclude firms whose actual controlling shareholder have changed in the pre- and post-new CAS (67)
adoption periods
Final sample (firm years) 2192
Central-government controlled firm years 600
Local-government controlled firm years 1592
pensation contracts for central-government-controlled firms. Starting from 2010, the SASAC required
central-government-controlled firms to use EVA (economic value added) to assess managerial perfor-
mance in executive compensation contracts. Hence, we do not include the firm years subsequent to
2009 in our sample. In addition, 2005 is the first year when the CSRC required listed firms to disclose
data on individual directors’ compensation. We exclude 2007 from the sample because it is the first
year of new CAS adoption but inferences are similar if 2007 is included in the post period. We exclude
financial firms because they are subject to different government regulations, including managerial
compensation, especially during the period after the 2008 financial crisis. We exclude the firm years
with CEO turnover because firm performance during the CEO turnover years likely reflects the contri-
bution of both the departing and incoming CEOs. In addition, the CEO cash compensation for such
years is not for the whole year and thus it is not comparable with the compensation in other firm
years. We require each firm to have at least one year of data in both the pre- and post-IFRS adoption
periods in order to make sure that any documented changes in the CEO’s pay-for-performance sensi-
tivity are not due to the change in the mix of firms over the two periods. Finally, we exclude the firms
whose ownership type (i.e., central-government-controlled, local-government-controlled, or non-
state-controlled) changed during our sample period.
Most of the financial data used in the study are obtained from the following commercial databases:
RESSET, CSMAR, and CCER. For the firm years with missing data, to the extent possible, we also hand
collected the missing data from annual reports.
Table 2 reports the descriptive statistics for our key regression variables for central-government-
controlled firms and local-government-controlled firms separately (Panel A) and correlations (Panel
B). There are a few interesting empirical regularities. First, the CEO’s annual total cash compensation
is economically significant relative to an average worker’s annual compensation for both types of
state-controlled firms.17 In addition, the CEO’s annual cash compensation is significantly higher for
central-government-controlled firms. Second, stock market performance RET is comparable for the
two types of firms but there is weak evidence that ROA is significantly higher for central-government-
controlled firms. Third, the two types of firms differ on several dimensions. The mean (the median) of
ASSETS is significantly larger (smaller) for central-government-controlled firms than for local-
government-controlled firms. Central-government-controlled firms experience faster growth (GROWTH)
and have lower financial leverage (LEV). The two types of firms also differ in terms of stock ownership by
the largest shareholder (LARGEST_OWN) but the economic significance of the difference in LARGES-
T_OWN does not appear to be big. It is also important to note that the mean and median values of LAR-
GEST_OWN are high, suggesting that share ownership is concentrated in the hands of the central
government for central-government-controlled firms and the local government for local-government-
controlled firms.
17
According to China National Statistics Bureau, the average worker’s annual pay for state-owned enterprises is RMB$35,053 in
2009.
Table 2
Descriptive statistics and correlation matrix for the regression variables.
Variable name Central-government- Local government- Two-tailed p value Two-tailed p value from
controlled firms controlled firms from the two-sample t the two-sample ranksum
(N = 600) (N = 1592) test test
Mean (median) [S.D.]
Panel A. Descriptive statistics for state-controlled Chinese firms
COMP 413,023 335,326 0.000*** 0.000***
(329,550) (269,200)
[331,201] [275,928]
ROA 0.039 0.035 0.220 0.077*
(0.032) (0.029)
[0.066] [0.069]
RET 0.014 0.052 0.211 0.243
(0.019) (0.000)
[0.608] [0.644]
ASSETS 4680 4210 0.096* 0.030**
(2110) (2410)
[6950] [5430]
GROWTH 0.190 0.140 0.002*** 0.000***
(0.153) (0.101)
[0.351] [0.325]
LEV 0.493 0.517 0.010*** 0.003***
(0.501) (0.536)
[0.193] [0.184]
LARGEST_OWN 40.63 39.27 0.070* 0.090*
(41.64) (38.71)
[16.56] [15.29]
COMP OPROA RET POST SIZE GROWTH LEV
Panel B. The Pearson correlation matrix for the regression variables
OPROA 0.2871***
RET 0.1554*** 0.1360***
POST 0.2904*** 0.0388* 0.3589***
SIZE 0.4283*** 0.1958*** 0.1208*** 0.3589***
GROWTH 0.0532** 0.2979*** 0.1021*** 0.1208*** 0.0944***
LEV 0.0092 0.3604*** 0.0594*** 0.1021*** 0.2749*** 0.0589***
LARGEST_OWN 0.0373* 0.1913*** 0.0026 0.0594*** 0.1759*** 0.1006*** 0.0954***
See the Appendix A for variable definitions.

***
Indicates statistical significance at the 1% level.
**
*
6. Regression results
6.1. The baseline regression results for central-government-controlled firms versus local-government-
controlled firms
Panel A of Table 3 shows the firm fixed effects regression result of CEO cash compensation for
central-government-controlled firms. The reported p values in all tables are based on standard errors
clustered by firm.18 We also checked for and found no evidence of severe multicollinearity for the
regression models in Table 3. The highest VIF is 5.52 in Panel A of Table 3 and 4.19 in Panel B of Table 3.
Consistent with H1, the coefficient on ROA in Panel A of Table 3 is significantly positive, suggesting
that the CEO’s cash pay is positively associated with ROA in the pre-new CAS adoption period for
18
All of our inferences are robust to computing the p values using the two-way clustering by year and firm, but we prefer to
report our main results using the one-way clustering by firm because we only have five years of panel data. Thompson (2011)
shows that when the number of clusters is small, the two-way clustering could lead to an over rejection of the null hypothesis.
Table 3
Firm fixed effects regression results of CEO annual cash compensation for state-controlled Chinese firms.
Coefficient Two-tailed p value

Panel A. Central-government-controlled firms
Test variables
POST 0.316*** 0.000
RET 0.096* 0.100
ROA 4.136*** 0.000
POST RET 0.116 0.126
POST ROA 2.527*** 0.002
Control variables
LN(ASSETS) 0.319*** 0.000
POST LN(ASSETS) 0.034 0.396
GROWTH 0.054 0.432
POST GROWTH 0.082 0.518
LEV 0.220 0.455
POST LEV 0.471 0.147
LARGEST_OWN 0.001 0.723
POST LARGEST_OWN 0.002 0.502
LN(ASSETS) RET 0.008 0.895
POST LN(ASSETS) RET 0.007 0.924
LN(ASSETS) ROA 1.240* 0.055
POST LN(ASSETS) ROA 0.978 0.133
GROWTH RET 0.002 0.989
POST GROWTH RET 0.153 0.332
GROWTH ROA 0.264 0.786
POST GROWTH ROA 0.276 0.824
LEV RET 0.183 0.681
POST LEV RET 0.581 0.171
LEV ROA 1.020 0.688
POST LEV ROA 4.735* 0.050
LARGEST_OWN RET 0.003 0.531
POST LARGEST_OWN RET 0.002 0.667
LARGEST_OWN ROA 0.082** 0.043
POST LARGEST_OWN ROA 0.039 0.306
Within R2 0.489
N 600
Panel B. Local-government-controlled firms

Test variables
POST 0.391*** 0.000
RET 0.082*** 0.004
ROA 1.275*** 0.004
POST RET 0.145*** 0.000
POST ROA 0.126 0.763
Control variables
LN(ASSETS) 0.271*** 0.000
POST LN(ASSETS) 0.037 0.175
GROWTH 0.026 0.577
POST GROWTH 0.005 0.948
LEV 0.639*** 0.001
POST LEV 0.191 0.207
LARGEST_OWN 0.004** 0.039
POST LARGEST_OWN 0.000 0.906
LN(ASSETS) RET 0.050* 0.054
POST LN(ASSETS) RET 0.130*** 0.000
LN(ASSETS) ROA 0.792** 0.032
POST LN(ASSETS) ROA 0.088 0.787
GROWTH RET 0.001 0.985
POST GROWTH RET 0.035 0.791
GROWTH ROA 0.387 0.577
POST GROWTH ROA 1.392* 0.095
Table 3 (continued)

LEV RET 0.196 0.171
POST LEV RET 0.320* 0.089
LEV ROA 2.905** 0.040
POST LEV ROA 1.615 0.242
LARGEST_OWN RET 0.000 0.843
POST LARGEST_OWN RET 0.000 0.922
LARGEST_OWN ROA 0.026 0.251
POST LARGEST_OWN ROA 0.022 0.355
Within R2 0.441
N 1592
Two-tailed p value
Ho: coefficient on ROA across the two panels is identical 0.001***
Ho: coefficient on POST ROA across the two panels is identical 0.008***
The dependent variable is ln(COMP). See the Appendix A for variable definitions. See Section 4 for the regression model. The
reported p values allow heteroskedasticity and any type of correlation for observations of the same firm but assume inde-
pendence for observations across different firms (Rogers, 1993). All the continuous explanatory variables are demeaned to avoid
multicollinearity and facilitate coefficient interpretation. To reduce the influence of outliers, all continuous variables are
winsorized at the top and bottom one percentile.
***
**
*
central-government-controlled firms. Consistent with H3, the coefficient on POST ROA is signifi-
cantly negative, suggesting that the CEO’s pay-for-performance sensitivity declines significantly in
the post-new CAS adoption period for central-government-controlled firms. These results are consis-
tent with the notion that the mandatory new CAS adoption reduces the stewardship usefulness of
financial reporting and therefore the optimal weight of accounting performance in the CEO’s cash
compensation is reduced in the post period.
The coefficients on ROA and POST ROA are also economically significant. The coefficient on ROA
implies that a one standard deviation increase in ROA would increase the CEO’s cash compensation by
27% (RMB112,745 evaluated at the mean COMP) in the pre-new CAS adoption period. The coefficient
on POST ROA implies that the new CAS adoption results in a 17% reduction in the CEO’s cash com-
pensation for every one standard deviation increase in ROA.
The coefficient on RET is significantly negative at 10% level and the coefficient on POST RET is
positive but insignificant at the 10% two-tailed level. Hence, there is little evidence that CEO’s pay
is positively associated with stock price performance for central-government-controlled firms. This
evidence is not surprising because China’s stock prices are volatile and hence do not serve as a reliable
indicator of managerial performance.
The coefficients on the interactions between firm performance (RET and ROA) and CONTROL are
generally insignificant. Notable exceptions are the significant coefficients on ln(ASSETS) ROA and
LARGEST_OWN ROA. The positive coefficient on ln(ASSETS) ROA suggests that the managerial
pay-for-performance sensitivity is stronger for larger firms. One interpretation of this finding is that
larger firms have more growth options and thus require a stronger pay-for-performance sensitivity
(Leone et al., 2006). The negative coefficient on LARGEST_OWN ROA implies that the managerial
pay-for-performance sensitivity is lower in firms with higher control shareholder ownership. This
finding is consistent with optimal contracting as argued in Ke et al. (1999).
Panel B of Table 3 shows the firm fixed effects regression result of CEO cash compensation for local-
government-controlled firms. The coefficient on ROA is significantly positive. More importantly, the
coefficient on ROA is significantly smaller for local-government-controlled firms than for central-
government-controlled firms (two-tailed p = 0.001). The magnitude of the managerial pay-for-
performance sensitivity for central-government-controlled firms is more than three times as large as
the magnitude of the managerial pay-for-performance sensitivity for local-government-controlled
firms (4.136 vs. 1.275). For local-government-controlled firms in the pre-new CAS adoption period, a
one standard deviation increase in ROA would increase the CEO’s cash compensation by only 9%
(RMB29,500 evaluated at the mean COMP), which is not as economically significant as for central-
government-controlled firms. Overall, this evidence is consistent with H2 that managers of local-
government-controlled firms face a much weaker pay-for-performance sensitivity than managers of
central-government-controlled firms prior to the new CAS adoption.
The coefficient on POST ROA is insignificant for local-government-controlled firms. More impor-
tantly, consistent with H4, the coefficient on POST ROA is significantly smaller for local-
government-controlled firms than for central-government-controlled firms (two-tailed p = 0.008).
The coefficient on POST ROA for local-government-controlled firms implies that the new CAS adop-
tion results in a 0. 9% reduction in the CEO’s cash compensation for every one standard deviation
increase in ROA, which is not economically significant. These results are consistent with the hypoth-
esis that local-government-controlled firms adopted little economically significant managerial pay-
for-performance sensitivity in the pre-new CAS adoption period and therefore the mandatory new
CAS adoption had little impact on these firms’ managerial pay-for-performance sensitivity.
Overall, the evidence in Table 3 suggests that China’s mandatory new CAS adoption reduces the
stewardship usefulness of financial reporting; therefore, central-government-controlled Chinese
firms, which are more oriented toward shareholder value maximization, altered their managerial
incentive compensation contracts to avoid the negative consequences of the new CAS adoption. Since
we find different results for central-government-controlled firms and local-government-controlled
firms as predicted, it seems unlikely that the results in Table 3 are due to unobservable confounding
macro effects that affect both types of state-controlled firms.
6.2. The cross-sectional variation in the impact of the new CAS adoption
We next examine whether the documented negative effects of H3 and H4 are stronger for state-
controlled firms that are more significantly affected by the new CAS adoption. To test this supplemen-
tal hypothesis, we first develop a measure of the impact of the new CAS adoption on accounting earn-
ings based on the absolute difference in net income under the new CAS and the old CAS for the
transition year 2006, scaled by the average total assets in year 2006 based on the old CAS (denoted
as ABS_DIFF). Then, we define a dummy variable GAP that equals one if ABS_DIFF is greater than
the median ABS_DIFF and zero otherwise.19 Please note that GAP is defined using 2006 only because
the two sets of net income are available for only the transition year. We allow the coefficients on POST,
RET, ROA, POST RET, and POST ROA in the baseline regression model to vary with GAP. The coeffi-
cient on GAP POST ROA is our variable of interest.
Table 4 shows the regression results of the baseline regression model augmented with GAP and its
interactions with POST, RET, ROA, POST RET, and POST ROA. For brevity, we do not tabulate the
coefficients on the control variables in Table 4. Note the coefficient on GAP is subsumed because
GAP is a firm fixed effect and therefore the coefficient on GAP cannot be separately identified in a firm
fixed effect regression.20
The key variable of interest is the coefficient on GAP POST ROA. Consistent with our prediction,
the coefficient on GAP POST ROA in Panel A of Table 4 is significantly negative for central-
government-controlled firms. For local-government-controlled firms in Panel B of Table 4, the coeffi-
cient on GAP POST ROA is insignificant. This finding is not surprising because as noted in Table 3,
managers of local-government-controlled firms face a weak pay-for-performance sensitivity to start
with. In addition, we find that the coefficient on GAP POST ROA is significantly different for
central-government-controlled firms and local-government-controlled firms (two-tailed p
value = 0.011). Overall, the evidence in Table 4 suggests that the negative effect of mandatory IFRS
adoption on the managerial pay-for-performance sensitivity is stronger for central-government-
controlled firms whose accounting earnings are more significantly affected by the mandatory new
19
Inferences are qualitatively similar if we use ABS_DIFF instead.
20
We also checked for and found no evidence of severe multicollinearity for the regression models in Table 4. The highest VIF is
12.75 in Panel A of Table 4 and 7.22 in Panel B of Table 4.
Table 4
Firm fixed effects regression results of CEO annual cash compensation for state-controlled Chinese firms: conditional on the
magnitude of the new CAS adoption’s impact on accounting earnings.

Panel A. Central-government controlled firms
POST 0.341*** 0.000
RET 0.062 0.377
ROA 3.719*** 0.003
POST RET 0.049 0.626
POST ROA 0.180 0.886
GAP Subsumed
GAP POST 0.036 0.609
GAP RET 0.045 0.615
GAP ROA 0.553 0.652
GAP POST RET 0.238* 0.063
GAP POST ROA 3.170*** 0.009
Within R2 0.506
N 600
Panel B. Local-government controlled firms

POST 0.395*** 0.000
RET 0.140*** 0.001
ROA 1.493** 0.013
POST RET 0.228*** 0.000
POST ROA 0.170 0.755
GAP Subsumed
GAP POST 0.009 0.859
GAP RET 0.133*** 0.007
GAP ROA 0.574 0.441
GAP POST RET 0.186*** 0.009
GAP POST ROA 0.203 0.752
Within R2 0.446
N 1592
Two-tailed p value
Ho: coefficient on GAP POST ROA across the two panels is 0.011**
identical
The dependent variable is ln(COMP). See the Appendix A for variable definitions. The model in this table is the baseline
regression model in Section 4 augmented with GAP and its interactions with POST, RET, ROA, POST RET, and POST ROA. For
brevity, the coefficients on the other explanatory variables are omitted. The reported p values allow heteroskedasticity and any
type of correlation for observations of the same firm but assume independence for observations across different firms (Rogers,
1993). All the continuous explanatory variables are demeaned to avoid multicollinearity and facilitate coefficient interpretation.
To reduce the influence of outliers, all continuous variables are winsorized at the top and bottom one percentile.
***
**
*
CAS adoption. This evidence provides further support for the hypothesis that China’s new CAS adop-
tion reduces the stewardship usefulness of financial reporting.
As noted in the Introduction, one of the controversies surrounding the mandatory IFRS adoption
debate is fair value accounting. Hence, we also examine whether the documented negative effects of
H3 and H4 are stronger for firms that more heavily rely on fair value accounting in the post-new
CAS adoption period. Specifically, we define a new dummy variable FV that is equal to one if the earn-
ings attributable to tradable and available-for-sale financial assets in any of the post-new CAS adoption
years is non-zero and zero otherwise. Then, we run the regression model in Table 4 by replacing GAP
with FV. As shown in Table 5, the coefficient on FV POST ROA is negative but insignificant for
central-government-controlled firms (two-tailed p value = 0.301) while the coefficient FV POST
ROA is significantly positive for local-government-controlled firms (two-tailed p value = 0.066). More
importantly, the coefficient on FV POST ROA is significantly different for central-government-
Table 5
Firm fixed effects regression results of CEO annual cash compensation for state-controlled Chinese firms: conditional on the effect
of fair value accounting in the post-new CAS adoption period.

Panel A. Central-government-controlled firms
POST 0.325*** 0.000
RET 0.005 0.934
ROA 3.198*** 0.000
POST RET 0.017 0.834
POST ROA 1.812** 0.030
FV Subsumed
FV POST 0.014 0.845
FV RET 0.228** 0.028
FV ROA 1.670 0.146
FV POST RET 0.197 0.181
FV POST ROA 1.171 0.301
Within R2 0.500
N 592
Panel B. Local-government-controlled firms

POST 0.449*** 0.000
RET 0.111*** 0.006
ROA 2.147*** 0.001
POST RET 0.178*** 0.001
POST ROA 0.799 0.101
FV Subsumed
FV POST 0.083* 0.086
FV RET 0.048 0.355
FV ROA 1.436* 0.065
FV POST RET 0.056 0.449
FV POST ROA 1.196* 0.066
Within R2 0.450
N 1578
Two-tailed p value
Ho: coefficient on FV POST ROA across the two panels is 0.062*
identical
The dependent variable is ln(COMP). FV is a dummy variable that equals one if the earnings attributable to tradable and
available-for-sale financial assets in any of the post-new CAS adoption years is non-zero. See the Appendix A for all other
variable definitions. The regression model in this table is the same as in Table 4 except that CAP is replaced with FV. For brevity,
the coefficients on the other explanatory variables are omitted. The reported p values allow heteroskedasticity and any type of
correlation for observations of the same firm but assume independence for observations across different firms (Rogers, 1993).
All the continuous explanatory variables are demeaned to avoid multicollinearity and facilitate coefficient interpretation. To
reduce the influence of outliers, all continuous variables are winsorized at the top and bottom one percentile.
***
**
*
controlled firms and local-government-controlled firms (two-tailed p value = 0.062). This evidence
weakly suggests that the use of fair value accounting under the new CAS reduces the stewardship use-
fulness of financial reporting for state-controlled Chinese firms.
7. Potential confounding macro shocks
Several major confounding macro shocks occurred around the same time as China’s new CAS adop-
tion. In this section we explicitly identify these major potential confounding macro shocks and then
discuss whether they could explain the documented results in Tables 3 and 4. Furthermore, we use
the propensity score matching method to better match the central-government-controlled firms with
the local-government-controlled firms. None of our hypothesis tests are qualitatively affected in the
following robustness checks for both Tables 3 and 4. Hence, for brevity, we tabulate and discuss the
results for Table 3 only below.
7.1. The split share structure reform
We first examine whether our regression results in Tables 3 and 4 could be explained by China’s
split share structure reform. During our sample period the CSRC initiated the split share structure
reform that made all non-tradable shares freely tradable. Prior to the split share structure reform,
all A share firms had two types of common stocks: non-tradable shares and tradable shares. Non-
tradable shares are largely owned by a controlling shareholder while tradable shares are listed on
one of the two domestic stock exchanges and can be owned by Chinese citizens, domestic institutions
and qualified foreign institutional investors. While the CSRC announced the first pilot batch of four
companies for the reform in April 2005, more than 90% of the state-controlled A share firms in our
sample completed the reform by the end of 2006, which coincided with China’s new CAS adoption.
Because the primary objective of the split share structure reform is to help align the interests
between non-tradable shareholders and tradable shareholders, the reform may help strengthen the
managerial pay-for-performance sensitivity, opposite to our documented effects in Table 3. In addi-
tion, this incentive alignment effect, if it exists, is expected to be larger for firms whose controlling
shareholders own a greater percentage of previously non-tradable shares (i.e., LARGEST_OWN).
The baseline regression model (1) has already included the interactions between LARGEST_OWN
and the two firm performance measures (RET and ROA) that are allowed to vary with POST. Hence,
our inferences in Tables 3 and 4 have already controlled for the effect of the split share structure
reform. With regard to the incentive alignment effect of the split share structure reform, as shown
in Table 3, the coefficients POST LARGEST_OWN RET and POST LARGEST_OWN ROA are
insignificant in both Panel A and Panel B. Hence, there is mixed evidence suggesting that the reform
helps strengthen the incentive alignment between non-tradable shareholders and tradable sharehold-
ers. More importantly, our inferences for the hypotheses are robust to controlling for the split share
structure reform effect.
7.2. Managerial compensation reform
We next examine the impact of managerial compensation reform on our inferences for the
hypotheses. In early 2006 the CSRC issued rules that allowed A shares to adopt stock option compen-
sation. Hence, it is possible that the observed decline in the managerial pay-for-performance sensitiv-
ity post the mandatory IFRS adoption for state-controlled A share firms reflects a substitution between
cash compensation and equity compensation for the firms that adopted stock option schemes during
our sample period. To rule out this alternative explanation, we identify our sample firms that adopted
stock option plans during 2005–2009. There are 57 unique firms that implemented stock option plans
during 2005–2009 (11 unique central-government-controlled firms and 46 unique local-government-
controlled firms).21 We include in CONTROL of regression model (1) a dummy variable that equals one
for the firms that adopted stock option schemes during our sample period and obtain similar inferences
for the hypotheses (see Panel A of Table 6).
7.3. The 2008 economic stimulus plan
We also examine the impact of China’s 2008 economic stimulus plan on our inferences for the
hypotheses. In response to the 2008 global financial crisis, the Chinese central government imple-
mented a massive RMB four trillion economic stimulus plan in November 2008. We identified seven
types of firms most affected by the stimulus plan (involving 173 unique firms representing 534 firm
years in our sample), including firms in the environmental industry, firms in the high-tech industry,
21
Due to the significant decline and volatility of China’s overall stock market in 2008–2010, many listed firms suspended the
initiation of managerial equity incentive schemes in the post-new CAS adoption period.
Table 6
Firm fixed effects regression results of CEO annual cash compensation for state-controlled Chinese firms: controlling for potential
confounding macro events.
Panel A. Managerial compensation reform

Central-government-controlled firms
ROA 4.195*** 0.000
POST ROA 2.582*** 0.002
Within R2 0.492
N 600
Local-government-controlled firms
ROA 1.248*** 0.006
POST ROA 0.143 0.736
Within R2 0.442
N 1592
Two-tailed p value
Ho: coefficient on POST ROA across the two panels is identical 0.007***
Panel B. The 2008 economic stimulus plan

ROA 5.233*** 0.000
POST ROA 3.005*** 0.009
Within R2 0.506
N 600
ROA 1.084** 0.021
POST ROA 0.004 0.994
Within R2 0.446
N 1592
Two-tailed p value
Ho: coefficient on POST ROA across the two panels is identical 0.012**
Panel C. Propensity score matched sample

ROA 4.220*** 0.000
POST ROA 2.595*** 0.004
Within R2 0.488
N 563
ROA 1.977*** 0.007
POST ROA 0.014 0.984
Within R2 0.506
N 601
Two-tailed p value
Ho: coefficient on ROA across the two panels is identical 0.037**
Ho: coefficient on POST ROA across the two panels is identical 0.019**
The dependent variable is ln(COMP). See the Appendix A for variable definitions. See Section 4 for a description of the baseline
regression model. For Panel A’s regression model, we include in CONTROL an extra dummy variable that equals one for the firms
that implemented stock option plans during 2005–2009. For Panel B’s regression model, we include in CONTROL an extra
dummy variable that equals one for the firms most affected by the 2008 economic stimulus plan. For Panel C’s regression model,
we use the propensity score matching method described in Section 7 to identify a matched sample of central-government-
controlled firms and local-government-controlled firms. For brevity, the coefficients on the other explanatory variables are
omitted for all the regression models in the table. The reported p values allow heteroskedasticity and any type of correlation for
observations of the same firm but assume independence for observations across different firms (Rogers, 1993). All the
continuous explanatory variables are demeaned to avoid multicollinearity and facilitate coefficient interpretation. To reduce the
influence of outliers, all continuous variables are winsorized at the top and bottom one percentile.
⁄
***
**
firms in the infrastructure industry, firms in the agriculture industry, firms in the health care and cul-
ture industries, real estate firms, and the firms in the Sichuan earthquake disaster zone. To control for
the effect of the stimulus plan, we include in CONTROL of regression model (1) a dummy variable that
equals one for the 173 unique firms most affected by the stimulus plan and obtain similar inferences
for the hypotheses (see Panel B of Table 6).
7.4. The concurrent increased enforcement of accounting standards
We next assess the impact of increased enforcement of accounting standards on our inferences for
the hypotheses. As discussed in Section 2, the Chinese government implemented several supporting
measures over the period 2007–2009 that aimed to strengthen the enforcement of the new CAS. Since
it is difficult to adequately control for the confounding effects of increased regulatory enforcement,
any improved financial reporting quality and associated economic consequences post the mandatory
IFRS adoption could be attributable to either the improved quality of the new accounting standards or
the improved enforcement of accounting standards (Christensen et al., 2013). Fortunately, our study is
immune from this alternative explanation because, if the increased regulatory enforcement resulted in
the increased stewardship usefulness of financial reporting, the managerial pay-for-performance sen-
sitivity for central-government-controlled firms should have increased rather than decreased in the
post-new CAS adoption period.
7.5. Propensity score matching
Rather than explicitly analyzing the effects of potential confounding macro shocks on our infer-
ences, we next use the propensity score matching method to identify a matched sample of central-
government-controlled firms and local-government-controlled firms and then re-estimate the regres-
sion models in Table 3. Specifically, we first run the following logistic regression model using the full
sample in 2005–2009 (firm and year subscripts are omitted for brevity):
CENTRAL SOE ¼ b0 þ b1 ROA þ b2 LNðASSETSÞ þ b3 LEV þ b4 GROWTH þ b5 LARGEST OWN

þ b6 RET þ b7 CAPEX þ b8 RPT þ b9 MONOP þ e ð2Þ
CENTRAL_SOE is a dummy variable that equals one for central-government-controlled firms and
zero for local-government-controlled firms. To our knowledge, there is no prior research on the deter-
minants of CENTRAL_SOE. Therefore, we include CAPEX, RPT, and MONOP based on prior research in
addition to the six explanatory variables included in model (1). Zhang and Wang (2010) find that Chi-
nese SOEs are prone for overinvestment, hence we include CAPEXP, defined as the annual capital
expenditure scaled by the average beginning and ending total assets. Jiang et al. (2010) find that
the related party transactions are lower in local SOEs than in central SOEs. Hence, we also include
RPT, defined as the sum of the related party sales and related party purchases divided by the sum
of total sales revenue and total operating costs in a year (Aharony et al., 2010b). Because many Chinese
SOEs operate in monopoly industries, we also include MONOP, defined as a dummy variable equal to
one for the following industries: energy, telecommunications, military, media, aerospace, mining,
public utilities and public transportations, and tobacco.
Next, we obtain the predicted probabilities from model (2) and find the closest match of local-
government-controlled firm without replacement for each central-government-controlled firm in
the last available year of the pre-new CAS adoption period. To ensure that the observations are closely
matched, we require the difference in the predicted probabilities to be less than 1%. The final matched
sample contains 181 unique central-government-controlled firms (563 firm years) and 181 unique
local-government-controlled firms (601 firm years).
Panel C of Table 6 shows the regression results of model (1) based on the firm years of the matched
sample of central-government-controlled firms and local-government-controlled firms over our sam-
ple period 2005–2009. The untabulated results of regression model (2) show that the coefficients on ln
(ASSETS), GROWTH, CAPEX, and RPT are significantly different from zero, while the coefficients on
ROA, RET, LEV, LARGEST_OWN, and MONOP are insignificant. As shown in Panel C of Table 6, our infer-
ences for the hypotheses are qualitatively similar based on the smaller but better matched samples.
8. Conclusion
The objective of this study is to examine how Chinese Accounting Standards’ substantial conver-
gence with IFRS in 2007 (referred as China’s new CAS adoption) affects state-controlled A share firms’
sensitivity of the CEO’s cash compensation to reported accounting earnings, one important aspect of
the stewardship usefulness of financial reporting. Our sample of state-controlled Chinese firms com-
prises central-government-controlled firms and local-government-controlled firms. We analyze the
two types of firms separately because managers of central-government-controlled firms have a stron-
ger incentive than managers of local-government-controlled firms to maximize shareholder value.
Therefore, we expect the effect of the mandatory new CAS adoption on the managerial pay-for-
performance sensitivity to be stronger for central-government-controlled firms than for local-
government-controlled firms. For central-government-controlled firms, we find that the managerial
pay-for-performance sensitivity is significantly positive and economically significant in the pre-new
CAS adoption period, but the sensitivity declines significantly post the mandatory new CAS adoption,
especially for firms that are more significantly affected by the new CAS adoption. We do not find sim-
ilar effects for local-government-controlled firms. Specifically, while the managerial pay-for-
performance sensitivity for local-government-controlled firms is significantly positive in the pre-
new CAS adoption period, the sensitivity is much smaller than that for central-government-
controlled firms and is not economically significant. In addition, we find no evidence that the manage-
rial pay-for-performance sensitivity for local-government-controlled firms declines significantly post
the mandatory new CAS adoption, even for firms that are more significantly affected by the new CAS
adoption. We also rule out the possibility that the above documented results are due to confounding
macro shocks that occurred around the same time as the new CAS adoption.
Overall, our results support the hypothesis that China’s mandatory new CAS adoption reduces the
stewardship usefulness of financial reporting. As a result, central-government-controlled Chinese
firms, which are more oriented toward shareholder value maximization, altered their managerial
incentive compensation contracts to avoid the negative consequences of the new CAS adoption. Our
findings should be of significant interest to accounting standard setters (e.g., IASB, China’s Ministry
of Finance) who wish to adopt accounting standards that help improve firms’ resource allocation.
Our study suggests a few avenues for future research. First, we have only demonstrated the overall
effect of China’s new CAS adoption on the pay-for-performance sensitivity. Future research may iden-
tify the specific accounting standards in the new CAS that are responsible for the reduced usefulness of
financial accounting. Such detailed knowledge is useful to standard setters who wish to improve exist-
ing accounting standards. Second, to the extent that the managerial pay-for-performance sensitivity in
the pre-new CAS adoption period represents optimal contracting, the reduced managerial pay-for-
performance sensitivity resulting from China’s mandatory new CAS adoption would push state-
controlled A share firms’ managerial compensation away from the initial optimal equilibrium (i.e., a
loss to shareholder value). Hence, it would be interesting for future research to examine whether
shareholders of central-government-controlled A share firms can identify alternative cost effective
monitoring mechanisms in order to bring the managerial compensation structure closer to the initial
optimal equilibrium.
Acknowledgements
We wish to thank two anonymous reviewers, Joseph Fan (CAPANA Conference discussant), Shiheng
Wang (AAA annual meeting discussant), and workshop participants at the Beijing University, China
Europe International Business School, London School of Economics, Nanyang Business School, National
Taiwan University, Shanghai Jiaotong University, Tsinghua University, The CAPANA annual conference,
and the AAA annual meeting for helpful comments. Yuan acknowledges the financial support of the
National Natural Science Foundation of China (No. 71272074).
Appendix A. Variable definitions
Variable name Definition

State-control A listed firm is defined to be state-controlled if the immediate largest
shareholder of the listed firm is a state-owned enterprise or a government
agency
COMP The total annual cash compensation (salary and bonus) paid to the CEO (in
RMB)
ROA Annual operating income divided by the average total assets
RET The fiscal-year market-adjusted abnormal stock return. The market return is the
Shanghai and Shenzhen composite index return for all A share firms
ASSETS The year-end total assets (in millions of RMB)
GROWTH The annual sales growth rate
LEV Total debts divided by total assets at the year-end.
LARGEST_OWN Ownership percentage of the largest shareholder of the listed firm
GAP A dummy variable that equals one if ABS_DIFF, defined as the natural logarithm
of one plus the absolute difference between the year 2006 net income based on
the old CAS and the year 2006 net income based on the new CAS scaled by the
average total assets based on the old CAS, is greater than the median ABS_DIFF,
and zero otherwise. Note that GAP is a firm fixed effect and does not vary over
time
POST A dummy variable that equals one for the post-IFRS adoption period 2008–
2009, and zero for the pre-IFRS adoption period 2005–2006
FV A new dummy variable that is equal to one if the earnings attributable to
tradable and available-for-sale financial assets in any of the post-new CAS
adoption years is non-zero and zero otherwise
References
Aharony, J., Barniv, R., Falk, H., 2010a. The impact of mandatory IFRS adoption on equity valuation of accounting numbers for
security investors in the EU. Eur. Account. Rev. 19 (3), 35–78.
Aharony, J., Wang, J., Yuan, H., 2010b. Tunneling as an incentive for earnings management during the IPO process in China. J.
Account. Public Pol. 29 (1), 1–26.
Ahmed, A.S., Neel, M., Wang, D., 2013. Does mandatory adoption of IFRS improve accounting quality? Preliminary evidence.
Contemp. Account. Res. 30 (4), 1344–1372.
Ball, R., 2006. International Financial Reporting Standards (IFRS): pros and cons for investors. Account. Bus. Res. 36, 5–27,
Special Issue.
Ball, R., Kothari, S.P., Robin, A., 2000. The effect of international institutional factors on properties of accounting earnings. J.
Account. Econ. 29 (1), 1–51.
Ball, R., Li, X., Shivakumar, L., 2015. Contractibility and transparency of financial statement information prepared under IFRS:
evidence from debt contracts around IFRS adoption. J. Account Res. 53 (5), 915–963.
Ball, R., Robin, A., Wu, J., 2003. Incentives versus standards: properties of accounting income in four East Asian countries. J.
Account. Econ. 36 (1–3), 235–270.
Banker, R., Datar, S., 1989. Sensitivity, precision, and linear aggregation of signals for performance evaluation. J. Account. Res. 27
(1), 21–39.
Barth, M.E., Landsman, W.R., Young, D., Zhuang, Z., 2014. Relevance of differences between net income based on IFRS and
domestic standards for European firms. J. Bus. Financ. Account. 41 (3–4), 297–327.
Batson, A. 2010. China seeks a new self through an old method. The Wall Street J. online, October 13.
Beuselinck, C.A.C., Joos, P.P.M., Khurana, I.K., van der Meulen, S. 2010. Mandatory IFRS Reporting and Stock Price
Informativeness. (CentER Discussion Paper; vol. 2010-82). Tilburg: Accounting.
Brüggemann, U., Hitz, J., Sellhorn, T., 2013. Intended and unintended consequences of mandatory IFRS adoption: extant
evidence and suggestions for future research. Eur. Account. Rev. 22 (1), 1–37.
Bushman, R., Engel, E., Smith, A., 2006. An analysis of the relation between the stewardship and valuation roles of earnings. J.
Account. Res. 44 (1), 53–83.
Byard, D., Li, Y., Yu, Y., 2011. The effect of mandatory IFRS adoption on financial analysts information environment. J. Account.
Res. 49 (1), 69–96.
Cai, H.B., Fang, H.M., Xu, L.C., 2011. Eat, drink, firms and government: an investigation of corruption from entertainment and
travel costs of Chinese firms. J. Law Econ. 54 (1), 55–78.
Capkun, V., Collins, D.W., Jeanjean, T., 2016. The effect of IAS/IFRS adoption on earnings management (smoothing): a closer look
at competing explanations? J. Account. Public Pol. 5 (5).
Cascino, S., Gassen, J., 2015. Mandatory IFRS adoption and accounting comparability. Rev. Account. Stud. 20 (1), 242–282.
Chen, G., Firth, M., Xu, L., 2009. Does the type of ownership control matter? Evidence from China’s listed companies. J. Bank.
Financ. 33 (1), 171–181.
Cheung, Y., Rau, P.R., Stouraitis, A., 2010. Helping hand or grabbing hand? Central vs. local government shareholders in Chinese
listed firms. Rev. Financ. 14 (4), 669–694.
Christensen, H.B., Hail, L., Leuz, C., 2013. Mandatory IFRS reporting and changes in enforcement. J. Account. Econ. 56 (2–3), 147–
177.
Christensen, H.B., Lee, E., Walker, M., Zeng, C., 2015. Incentives or standards: what determines accounting quality changes
around IFRS adoption? Eur. Account. Rev. 24 (1), 31–61.
Christensen, J., Demski, J., 2003. Accounting Theory: An Information Content Perspective. McGraw-Hill Irwin.
Core, J., Guay, W., Verrecchia, R., 2003. Price versus non-price performance measures in optimal CEO compensation contracts.
Account. Rev. 78 (4), 957–981.
Daske, H., Hail, L., Leuz, C., Verdi, R., 2008. Mandatory IFRS reporting around the world: early evidence on the economic
consequences. J. Account. Res. 46 (5), 1085–1142.
Davilla, A., Penalva, F., 2006. Governance structure and the weighting of performance measures in CEO compensation. Rev.
Account. Stud. 11 (4), 463–493.
DeFond, M., Hu, X.S., Hung, M.Y., Li, S.Q., 2011. The impact of mandatory IFRS adoption on foreign mutual fund ownership: the
role of comparability. J. Account. Econ. 51 (3), 240–258.
Deloitte Touche Tohmatsu, 2006. China’s New Accounting Standards: A Comparison with Current PRC GAAP and IFRS. <http://
www.iasplus.com/country/china.htm>.
Du, F., Tang, G., Young, S.M., 2012. Influence activities and favoritism in subjective performance evaluation: evidence from
Chinese state-owned enterprises. Account. Rev. 87 (5), 1555–1588.
Gjesdal, F., 1981. Accounting for stewardship. J. Account. Res. 19 (1), 208–231.
He, X., Wong, T.J., Young, D., 2012. Challenges for implementation of fair value accounting in emerging markets: evidence from
china. Contemp. Account. Res. 29 (2), 538–562.
Henderson, D., O’Brien, P.C., 2016. The Standard-Setters’ Toolkit: Can Principles Prevail over Bright Lines? Work. Pap. Wilfrid
Laurier University and University of Waterloo.
Holmstrom, B., 1979. Moral hazard and observability. Bell J. Econ. 10 (1), 74–91.
Hou, Q., Jin, Q., Wang, L., 2014. Mandatory IFRS adoption and executive compensation: evidence from China. China J. Account.
Res. 7 (1), 9–29.
Hung, M., 2000. Accounting standards and value relevance of financial statement: an international analysis. J. Account. Econ. 30
(3), 401–420.
International Accounting Standards Board (IASB), 2005. Agenda Paper 8A, 19 October 2005, Information for Observers <http://
www.iasb.org/Current-Projects/IASB-Projects/Conceptual-Framework/Meeting-Summaries-and-Observer-Notes/October-
2005.htm>.
International Accounting Standards Board (IASB), 2006. Discussion Paper-Preliminary views on an Improved Conceptual
Framework for Financial Reporting: the Objective of Financial Reporting and Qualitative Characteristics of Decision-Useful
Financial Reporting Information. IASB, London, U.K.
International Accounting Standards Board (IASB), 2008. Exposure Draft of An improved Conceptual Framework for Financial
Reporting, The Objective of Financial Reporting, Qualitative Characteristics and Constraints of Decision-Useful Financial
Reporting Information. IASB, London, U.K.
International Accounting Standards Board (IASB), 2010. The Conceptual Framework for Financial Reporting. IASB, London, U.K.
Jiang, G., Lee, C., Yue, H., 2010. Tunneling through inter-corporate loans: the China experience. J. Financ. Econ. 98 (1), 1–20.
Kaufmann, D., Kraay, A., Mastruzzi, M., 2009. Governance Matters VIII: Aggregate and Individual Governance Indicators 1996–
2008. Work. Pap., World Bank Policy Research.
Ke, B., Petroni, K., Safieddine, A., 1999. Ownership Concentration and Sensitivity of Executive Pay to Accounting Performance
Measures: Evidence from Publicly and Privately-held Insurance Companies. J. Account. Econ. 28 (2), 185–209.
Ke, B., Oliver, R., Yu, W., 2012. Hong Kong stock listing and the sensitivity of managerial compensation to firm performance in
state-controlled Chinese firms. Rev. Account. Stud. 17 (1), 166–188.
Lambert, R.A., 2001. Contracting theory and accounting. J. Account. Econ. 32 (1–3), 3–87.
Lambert, R.A., Larcker, D.F., 1987. An analysis of the use of accounting and market measures of performance in executive
compensation contracts. J. Account. Res. 25 (1), 85–125.
Landsman, W.R., Maydew, E.L., Thornock, J.R., 2012. The information content of annual earnings announcements and mandatory
adoption of IFRS. J. Account. Econ. 53 (1–2), 34–54.
Lee, E., Walker, M., Zeng, C., 2013. Does IFRS Convergence Affect Financial Reporting Quality in China? ACCA Research Report
131.
Leone, A.J., Wu, J.S., Zimmerman, J.L., 2006. Asymmetric sensitivity of CEO cash compensation to stock returns. J. Account. Econ.
42 (1–2), 167–192.
Li, S. 2009. Beijing Bid to Cool Property Prices Ignored. South China Morning Post, December 31, 2009.
Li, Z., 2000. Incentives and performance. Account. Res. 1, 24–30 (in Chinese).
Li, H., Zhou, L., 2005. Political turnover and economic performance: the incentive role of personnel control in China. J. Public
Econ. 89 (9–10), 1743–1762.
Li, S., 2010. Does mandatory adoption of International Financial Reporting Standards in the European Union reduce the cost of
equity capital? Account. Rev. 85 (2), 607–636.
Lin, Y.F., Liu, M.X., Zhang, Q., 2004. Policy burdens and the soft-budget constraint: evidence from China. Manage. World 8, 81–89
(in Chinese).
Liu, C., Yao, L.J., Hu, N., Liu, L., 2011. The impact of IFRS on accounting quality in a regulated market: an empirical study of China.
J. Account. Audit. Financ. 26 (4), 659–676.
Ma, H., 2009. China Regulates the Compensation of Heads of Business Enterprises: First Central-Government-Controlled Firms,
Followed by Local-Government-Controlled Firms and Other Types of Chinese Firms (in Chinese). October 23 <www.
chinanews.com>.
Ministry of Finance Accounting Division, 2008. Research report on publicly listed Chinese firms’ implementation of the new
accounting standards for fiscal year 2007. Account. Res. 6, 19–30 (in Chinese).
O’Connell, V., 2007. Reflections on stewardship reporting. Account. Horiz. 21 (2), 215–227.
Ozkan, N., Singer, Z., You, H., 2012. Mandatory IFRS adoption in Europe and the contractual usefulness of accounting information
in executive compensation. J. Account. Res. 50 (4), 1077–1107.
Paul, J., 1992. On the efficiency of stock-based compensation. Rev. Financ. Stud. 5 (3), 471–502.
Rogers, W.H., 1993. Sg17: Regression Standard Errors in Clustered Samples. Stata Tech. Bullet. 13, 19–23.
State-owned Asset Supervision and Administration Commission (SASAC), 2003. Interim Measures for the Assessment of
Management Performance of Heads of Central Government-Owned Enterprises (in Chinese). December 25.
State-owned Asset Supervision and Administration Commission (SASAC), 2004. Interim Measures for the Administration of
Compensation for Heads of Central Government-Owned Enterprises (in Chinese). June 2.
State-owned Asset Supervision and Administration Commission (SASAC), 2008. Supplemental Provisions for the Assessment of
Management Performance of Heads of Central Government-Owned Enterprises (in Chinese). February 25.
State-owned Asset Supervision and Administration Commission (SASAC) of Shanghai Putuo District, 2006. Interim Measures for
the Assessment of Management Performance of heads of Shanghai Putuo District Government-owned Enterprises (in
Chinese). October 30.
Sunder, S., 2009. IFRS and the accounting consensus. Account. Horizons 23 (1), 101–111.
Tan, H., Wang, S., Welker, M., 2011. Analyst following and forecast accuracy after mandated IFRS adoptions. J. Account. Res. 49
(5), 1307–1357.
Thompson, S.B., 2011. Simple formulas for standard errors that cluster by both firm and time. J. Financ. Econ. 99 (1), 1–10.
Voulgaris, G., Stathopoulos, K., Walker, M., 2014. IFRS and the use of accounting-based performance measures in executive pay.
Int. J. Account. 49 (4), 479–514.
Watts, R.L., 2003. Conservatism in accounting, Part I: Explanations and implications. Account. Horiz. 17 (3), 207–221.
Watts, R.L., 2006. What has the invisible hand achieved? Account. Bus. Res. 36 (Supplement 1), 51–61.
Wei, G., 2000. Incentives for top-management and performance of listed companies. Econ. Rev., 32–39 (in Chinese)
Wooldridge, J.M., 2002. Econometric Analysis of Cross Section and Panel Data. The MIT Press, Cambridge, Massachusetts.
Wu, J., Zhang, I., 2009. The voluntary adoption of internationally recognized accounting standards and firm internal
performance evaluation. Account. Rev. 84 (4), 1281–1309.
Wu, J., Zhang, I., 2011. Accounting Integration and Comparability: Evidence from Relative Performance Evaluation around IFRS
Adoption. University of Rochester, Work. Pap.
Wu, J., Zhang, I., 2014. The adoption of internationally recognized accounting standards: implications for the credit markets. J.
Account. Audit. Financ. 29 (2), 95–128.
Zeng, Q.S., Chen, X.Y., 2006. Government control, excess employees, and labor cost. Econ. Res. 5, 74–86 (in Chinese).
Zhang, H., Wang, Z., 2010. Government intervention, government objectives and state-owned listed companies’ overinvestment.
Nankai Bus. Rev. 13 (3), 101–108 (in Chinese).
Zhou, L.A., 2004. The incentive and cooperation of government officials in the political tournaments: an interpretation of the
prolonged local protectionism and duplicative investments in China. Econ. Res. 3, 33–40 (in Chinese).

2.1. Akuntansi Manajemen 3

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

2.1. Akuntansi Manajemen 3

Hochgeladen von

Copyright:

Verfügbare Formate

J. Account.

Public Policy 35 (2016) 567–591

Contents lists available at ScienceDirect

J. Account. Public Policy

Full length article

The substantial convergence of Chinese

We examine how China’s adoption of a new set of Chinese

⇑ Corresponding author. Fax: +86 21 6564 3203.

2. Institutional background and related research

2.1. Institutional background

2.2. Related research

In this section we explain why managerial incentives differ for central-government-controlled

Hypothesis 1 (H1). The managerial pay-for-performance sensitivity is economically significant for

Hypothesis 2 (H2). The managerial pay-for-performance sensitivity is greater for central-

Our hypothesis development is based on the implicit assumption that central-government-

As detailed in Sections 6 and 7, we adopt several complementary approaches to demonstrate the

5. The sample and descriptive statistics

See the Appendix A for variable definitions.

Coefficient Two-tailed p value

Panel B. Local-government-controlled firms

Coefficient Two-tailed p value

Coefficient Two-tailed p value

Panel B. Local-government controlled firms

Coefficient Two-tailed p value

Panel B. Local-government-controlled firms

7. Potential confounding macro shocks

7.1. The split share structure reform

7.2. Managerial compensation reform

7.3. The 2008 economic stimulus plan

Coefficient Two-tailed p value

Panel A. Managerial compensation reform

Panel B. The 2008 economic stimulus plan

Panel C. Propensity score matched sample

7.4. The concurrent increased enforcement of accounting standards

7.5. Propensity score matching

CENTRAL SOE ¼ b0 þ b1 ROA þ b2 LNðASSETSÞ þ b3 LEV þ b4 GROWTH þ b5 LARGEST OWN

Appendix A. Variable definitions

Variable name Definition

Das könnte Ihnen auch gefallen