Government protection, political connection and management turnover in China

    Government Protection, Political Connection and Management Turnover in China Louis T.W. Cheng, T.Y. Leung PII: DOI: Reference:

S1059-0560(16)30045-4 doi: 10.1016/j.iref.2016.03.010 REVECO 1261

To appear in:

International Review of Economics and Finance

Received date: Revised date: Accepted date:

26 January 2015 22 February 2016 1 March 2016

Please cite this article as: Cheng, L.T.W. & Leung, T.Y., Government Protection, Political Connection and Management Turnover in China, International Review of Economics and Finance (2016), doi: 10.1016/j.iref.2016.03.010

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Government Protection, Political Connection and Management Turnover in China

Government Protection, Political Connection

RI P

T

and Management Turnover in China

MA

NU

SC

Louis T. W. Cheng* School of Accounting and Finance Hong Kong Polytechnic University Hong Kong Tel: (852) 2766 7140 Fax: (852) 2356 9550 E-mail address: [email protected]

AC

CE

PT

ED

T. Y. Leung LSK School of Business and Administration Open University of Hong Kong Hong Kong Tel: (852) 2768 6949 Fax: (852) 2391 9095 E-mail address: [email protected]

* corresponding author Highlights  strategic firms perform better than non-strategic firms  strategic firms employ top managers with different demographic profiles  strategic firms are more likely to have management turnovers  turnovers are less frequent if the chairpersons and CEOs are politically-connected

0

ACCEPTED MANUSCRIPT Abstract

T

Mainland Chinese government employs two related strategies to protect its national

protection to industries of national interest.

It grants government

RI P

and economic interests in the process of financial liberalization.

In addition, it also maintains

We argue these strategic firms with economic and national interests

NU

strategic firms.

SC

political-linkage with certain firms to assert their influence. We term these firms as

demonstrate better performance and higher management turnovers.

Management

MA

turnovers are less frequent if the chairpersons and CEOs are politically-connected.

ED

The strategic firms also rebound better from financial distress than non-strategic firms.

PT

Keywords: political connection; government protection; management turnover; strategic firm

AC

CE

JEL: G32; G34; G38

1

ACCEPTED MANUSCRIPT Government Protection, Political Connection and Management Turnover in China 1. Introduction

T

In the past two decades, China has become a successful case in expanding its economic

RI P

presence through listing its firms on the stock markets in Mainland and overseas.

To privatize

state-own assets that have investment value to investors while maintaining effective control and

The process of privatization can be a lengthy one and

NU

& Poulsen, 2004; Sun & Tong, 2003).

SC

influence is a delicate matter (Jones, Megginson, Nash, & Netter, 1999; Megginson, Nash, Netter,

the financial benefit to the firm is reflected by the trade-off between government guarantee and

MA

performance improvement being completely privatized (Borisova & Megginson, 2011). Economist (2012) suggests that state enterprises in China have become more productive and

ED

the government has become a more sophisticated owner.

Even for non-state firms, the In

PT

government strategically offers its patronage if these private firms appear to be a winner.

short, the Chinese government uses a portfolio of sophisticated techniques to maintain control to

CE

the listed firms, no matter they are state-owned or privately-owned.

AC

Among many tools to control the listed firms, two interesting practices in China related to maintaining political control capture our attention. national security interest as protected industries.

First, China classifies certain industries of Under such a government protection, these

firms get business and financial advantages from government but subject to foreign ownership restriction.

Second, in order to protect national security and economic interests, the Chinese

government deploys political connection to certain firms through the top management and directors, which is another common phenomenon in China.

While politically connected firms

and protected firms possess different characteristics and behave differently due to industry

2

ACCEPTED MANUSCRIPT structure and other factors, these two types of firms are important channels developed by the Chinese government to pursue certain national economic and political interest.1

T

We strongly believe that Chinese central government has a long-term national agenda to

the cooperation of all kinds of firms.

RI P

pursue economic stability and stable growth for the country. These agenda sometimes demand We call them strategic firms.

Depending on the nature

We realize that these two groups of firms have different

NU

industry and political connection.

SC

of the national interest, the forms of implementation are mainly through two channels: protected

characteristics. However, we argue that despite these differences, if the national strategy of

MA

managing these firms for national interest does exist, then we should expect certain common impacts on performance, management turnover, and reaction under distress.

We acknowledge

However, this is a worthwhile research questions to be examined for academic to better

PT

theory.

ED

that our classification of these two groups of firms into some category can be controversial in

understand the relationship between Chinese firms and the government.

CE

The idea of industry protection is related to privatization.

In order to protect its natural

AC

resources and state assets through the privatization process, governments of developing countries often make use of industry protection to reduce the impact from foreign competition and to protect ownership. The various effects of privatization are well documented (Jones et al., 1999; Megginson et al., 2004) and recently the issue of political connection has already formed its own literature (Boubakri, Guedhami, Mishra, & Saffar, 2012; Faccio, 2006).

In management

research, industry protection can be interpreted as a special kind of corporate resource under the

1

We understand that this grouping is debatable. Each firm has its own unique feature yet at the same time possesses some common characteristics (e.g., commonly used characteristics to group firms are industry, firm size, market to book). Thus, we can still use these common characteristics to group firms for a certain research issue even though these firms may not be similar in all other aspects.

3

ACCEPTED MANUSCRIPT resource based view of firms (RBV) (Barney, 1991; Teece, Pisano, & Shuen, 1997) while the human capital of networking is documented under the guanxi literature (Park & Luo, 2001).

T

In this study, we argue that governments of developing countries employ two related

RI P

strategies to protect their national security interest and economic interest through listed firms. We define firms as strategic firms if the firms are in the industries under government

SC

protection/regulation or if the firms are politically connected.

First, there are many reasons a

NU

country wants to protect certain industries (Hillman, 1982; Melitz, 2005). Such protected firms enjoy certain business advantages in terms of receiving government contracts and/or lower cost Second, a country can also deploy politically connected CEOs and board members

MA

of capital.

to keep a tight control on certain firms.

As emerging Asian countries, such as China, are

ED

well-known for their relationship-driven business style and strong government influence in

relationship.

PT

business and financial sectors, political connection is regarded as a very important business Thus, companies indeed welcome politically connected executives and board

CE

members even though such relationships come with a price of sub-optimal business decisions

AC

due to government influence at times. While this second strategy may not be as effective and direct as the first approach, it allows the government to pinpoint certain firms instead of the whole industries for strategic reasons. We conjecture that government protection and political connection are being used by developing countries such as China to achieve an overall control in listed firms.

Consequently,

these strategic firms are expected to perform better financially, employ top managers with certain characteristics and skill sets, and demonstrate different turnover patterns compared with non-strategic firms.

4

ACCEPTED MANUSCRIPT Clearly, strategic firms need to appoint top managers who are willing and able to pursue political agenda if needed and increase shareholders’ wealth at the same time.

Such a

T

dual-objective requirement leads to a series of interesting empirical questions: Q1. What kind of

RI P

financial performance do these strategic firms demonstrate? Q2. What is the management turnover for strategic firms compared with non-strategic firms? Q3. When a firm is in financial

SC

trouble, what happen to the top managers of strategic firms? and lastly Q4. What is the

NU

post-turnover financial performance of strategic firms compared with non-strategic firms? Based on the literature and empirical questions stated above, we propose that 1) strategic

non-strategic firms.

MA

firms enjoying business and financial privileges have better financial performance relative to 2) Management turnover of strategic firms is subject to political decisions

ED

of government, leading to a different pattern relative to that of non-strategic firms.

3) Top

PT

managers of strategic firms with political ties are less likely to be replaced during financial distress.

4) Based on the conjecture that the financial well-being of strategic firms is protected

Our sample period covers eight years from 2001 to 2008.

We include all (but

AC

difficulty.

CE

by the government, we expect a speedy financial rebound of strategic firms after their financial

except financial) firms listed on the Shanghai Stock Exchange and Shenzhen Stock Exchange. The data on company returns, financial statements, board characteristics and ownership come from the China Stock Market and Accounting Research (CSMAR) database. and then classify the political connection data from the annual reports.

We hand-collect

There are a total of

7,959 observations in our sample. Our results show that strategic firms perform better financially compared with non-strategic firms.

Next, we examine top management turnovers and find that strategic firms are more

likely to have top management turnovers. We argue that turnover decisions of strategic firms

5

ACCEPTED MANUSCRIPT are somewhat influenced by political agenda of the government.

Thus the more frequent

turnover of strategic firms may be politically-driven and not be related to traditional

T

market-driven rationales such as moving to a higher pay job or being fired due to poor

RI P

performance.

We also find that after management turnover due to financial crisis, strategic firms In addition, there is no relation between the demographic

SC

outperform non-strategic firms.

We conjecture that this result is

NU

profile of top managers and post-turnover performance.

consistent with our argument that strategic firm performance is more related to its government

MA

protection and connection and less related to management leadership and ability when facing crisis.

ED

Following this section is background and hypotheses. Section 3 describes the research We conclude the study in Section 5.

PT

method. The results are presented in Section 4.

CE

2. Background and Hypotheses

AC

We argue that both guanxi and political connection literature supports our conjecture that political connection has been used as a channel for the mainland government to promote corporate China. First, the guanxi literature documents that in China, guanxi is a network of interpersonal ties and social connections that businesses use to exchange favors in order to achieve objectives. Social ties have been found to affect market and financial performance positively in emerging economies (Li, Poppo, & Zhou, 2008). Clearly, guanxi is useful to conduct business and marketing activities in China (Ambler, Styles, & Wang, 1999). political connection is a special type of guanxi.

In fact,

Peng and Luo (2000) define political ties as

informal social connections with officials in various levels of administration in the governments

6

ACCEPTED MANUSCRIPT (central, state or local) and regulatory agencies (tax authorizes, securities commission, stock exchange etc). Therefore, based on the past success of mainland government in employing

T

political influence to its advantage, using politically connected board directors and senior

RI P

management to influence business strategies at firm level would be expected. In addition, Calomiris, Fisman, and Wang (2010) show that the stock market assures the value of political They find that firms led by former government

SC

connection over the benefits of state ownership.

NU

officials experience positive market reaction to the disposal announcement of government shares in China.

MA

Some anecdotal evidence of such practice can be observed.

For instance, The Economist

(2012) reports that the Mainland government, through different means, provide support to

ED

industrial sectors or firms which have potential for success.

Lenovo Group Limited (Lenovo) is

PT

a typical example of a “national champion” being created through the usage of political connection. While Lenovo is not a SOE, its founders are well connected politically because

CE

they are members of a unit of Chinese Academy of Science (i.e., an institution of the State

AC

Council of China), which provides the seed money for the establishment of Lenovo.

In addition,

the leading founder of Lenovo, Liu Chuanzhi, was a representative of National People’s Congress for several years.

When it acquired IBM’s personal-computer division for US$1.25

billion in 2004”, the government was providing support behind the scene. According to Gartner (2013), Lenovo overtook HP as the number one personal computer vendor as of the third quarter sales of 20132, and political connection serves as an important element in the success of Lenovo. In finance, the issue of protected industry is explored under the privatization literature (Jones et al., 1999; Megginson et al., 2004).

The management literature also examines similar

2

Source: “Gartner says worldwide PC shipments in the third quarter of 2013 declined 8.6 percent”, Gartner, October 2013, http://www.gartner.com/newsroom/id/2604616

7

ACCEPTED MANUSCRIPT empirical issue on protected industries but is labeled under the resource based view of firms (RBV) (Barney, 1991; Teece et al., 1997) while the human capital of networking is documented

T

under the guanxi literature (Park & Luo, 2001).

RI P

In many developing countries (such as Brazil, India, and Malaysia)3 especially China, certain industries (e.g., energy, raw materials, petrochemicals) are under government protection.

SC

Nee, Opper, and Wong (2007) argue that the Chinese central government retains control over the

telecommunications, and energy.

NU

firms in the strategically and politically important industries, which include utilities, Bass and Chakrabarty (2014) develop the resource security

MA

theory to argue that ownership affects resource acquisition.

They suggest that

government-owners are most concerned with securing their country’s future and focus more on The firms in the protected industries are usually large monopolies, able to

ED

resource exploration.

PT

receive subsidies and special treatment (e.g., less stringent profitability requirement in new equity issues) and under direct control of the State Council (Aharony, Lee, & Wong, 2000).

In

CE

addition, there are some industries which are under special government regulation with In short, putting firms in protected

AC

restriction on share ownership (Sun & Tong, 2003).

industries and regulated industries can be viewed as a government strategy and an effective way to protect a country’s security and economic interests. In fact, the literature has clearly documented that politically connected firms enjoy certain competitive advantages.

There are a number of studies showing that politically connected firms

have better performance (e.g., Cooper, Gulen, & Ovtchinnikov, 2010; Fisman, 2001). Besides, firms having political connections can also enjoy other advantages including securing government contracts and property right protection; getting preferential tax treatment; and access

3

For example, automobile industry is protected in Malaysia; computer industry is protected in Brazil in the 1990s; telecommunication, heavy machinery and renewable energy are protected in India.

8

ACCEPTED MANUSCRIPT to cheaper finance through political networks (Claessens, Feijen, & Laeven, 2008; Goldman, Rocholl, & So, 2009).

Also, Su, Fung, Huang, and Shen (2014) report that politically

T

connected firms pay higher cash dividends because they can obtain more government support.

RI P

According to the RBV theory in management, the ties through “political connections” should be considered as a valuable resource to attain competitive advantage. Boddewyn and

In the managerial networking literature, Burt (1997) suggests that

NU

difficult for others to imitate.

SC

Brewer (1994) argue that the capability to develop ties with political actors is a resource which is

social ties and contacts of top executives with external entities can reduce transaction costs

MA

through exchange of resources, information and knowledge.

In addition, Holburn and Zelner

(2010) show that firms with political resources and capabilities can lobby policy for their own

ED

benefits and protect their investments against adverse policy changes and political rent-seeking

PT

behavior of competitors.

Specifically, a few studies examine Chinese data and show these advantages as well.

Li

CE

and Atuahene-Gima (2001) argue that political networking is a unique resource in the

AC

transitional economy of China. Peng and Luo (2000) find that managerial networks with local and government officials allow the new ventures and firms to have better performance in China. Park and Luo (2001) show that the Chinese firms use “guanxi” as a strategic mechanism to overcome the competitive disadvantage by exchanging favors with government authorities. There have been a number of studies examining the relation between management turnover and performance.

In the finance and accounting literature, human resources are regarded as

firm (intangible) assets (Lajili & Zéghal, 2006). to the disposal of a corporate asset.

Therefore, dismissal of an employee is similar

According to the agency theory, top management is

9

ACCEPTED MANUSCRIPT accountable for firm performance and replacement of poorly performing staff is an essential element of effective governance (Fama, 1980; Shleifer & Vishny, 1997).

T

Many studies have shown that poor performance is related to forced turnovers, particularly

RI P

in countries where there is strong law enforcement (DeFond & Hung, 2004; Huson, Parrino, & Starks, 2001). CEO turnover is negatively related to firm performance measured by stock price

SC

and accounting performance indicators and the sensitivity between performance and turnover is

NU

the greatest in smaller firms (Murphy, 1999). Jenter and Kanaan (2015) find a higher probability of forced CEO turnover when there are declines in market and industry performance. Parrino,

MA

Sias and Starks (2003) show a negative relation between institutional ownership changes and likelihood of forced CEO turnover. Besides past performance, expected performance is also

ED

relevant in CEO turnover decision. Farrell and Whidbee (2003) show CEO replacement is related

PT

to forecasted 5-year EPS growth and dispersion in analyst forecasts. There are some studies showing that after forced turnover, there is improved performance (Huson, Malatesta, & Parrino,

CE

2004; McNeil, Niehaus, & Powers, 2004).

Yuan (2011) shows there is a lower sensitivity between

AC

management turnover in China.

Several studies have been conducted on top

turnover and firm performance for politically connected CEOs in SOEs.

Most studies show

that managerial turnover (chairperson or CEO) is related to pre-turnover firm performance (Chang & Wong, 2009; Firth, Fung, & Rui, 2006). performance, the results are mixed.

In the examination of post-turnover

Groves, Hong, McMillan, and Naughton (1995) find an

increase in productivity after managerial turnover while Firth et al. (2006) provide no evidence of improvement in post-turnover performance.

Chang and Wong (2009) show that only

loss-making firms, but not profit-making firms, have post-turnover profitability improvement. In short, based on the literature on protected industry, political connection, and management

10

ACCEPTED MANUSCRIPT turnover, we construct four hypotheses to test our empirical questions stated earlier.

T

2.1 Hypothesis Development

RI P

In this study, we argue that government protection and appointing politically connected top managers/board members together form a unified proactive strategy of the government in A country can protect its natural

SC

protecting its economic interests through listed firms.

NU

resources from being exploited by foreign countries through barriers of ownership and trade. By doing so, these protected firms also enjoy certain business advantages in terms of receiving

MA

government contracts and/or lower cost of capital.

Similar protection practice is also observed

in industries that are related to technology relevant to national defense (e.g., telecommunications,

ED

airlines). Another category of protected industry is those facing difficulty in competing against

PT

more mature and bigger foreign firms.

As shown in previous studies, government protection status is commonly granted for natural

However, political connection can be employed for firms of both national security

AC

interest.

CE

resources and technology intensive industries when a country mainly wants to protect its security

interest and economic interest.

The literature on political connection has documented clearly

the possibility of better performance and enjoying competitive advantage.

Consequently,

politically connected firms could perform better than non-connected firms. Our approach of consolidating government protection and political connection into the government’s strategic positioning framework is consistent with the ‘‘grabbing hand’’ argument of Shleifer and Vishny (1998) that politicians extract resources from listed SOEs under their control

to

fulfill

certain

political

agenda

11

other

than

maximizing

firm

value.

ACCEPTED MANUSCRIPT Government-owners are most concerned with securing their country’s future and focus more on resource exploration as a national strategy (Bass & Chakrabarty, 2014).

T

Fan, Wong, and Zhang (2007) show that in China, the right of the government to appoint These strategic firms may

RI P

CEOs significantly impairs firm performance and board governance.

not pursue value maximization all the time due to the boundary of national security and

SC

economic interests comes first when the two objectives lead to conflicting decisions.

NU

Nevertheless, strategic firms get business advantages (Agrawal & Knoeber, 2001) as well as cheaper and preferential financing as a reward/compensation from the government (Boubakri et Politically connected firms can lobby policy for their own benefits

MA

al., 2012; Cull & Xu, 2005).

and protect their investments against adverse policy changes and political rent-seeking behavior

ED

of competitors (Holburn & Zelner, 2010).

These competitive advantages exist in the forms of

PT

securing government contracts and property right protection; getting favorable tax treatment; and obtaining cheaper and preferential financing (Boubakri et al., 2012; Claessens et al., 2008; Cull

CE

& Xu, 2005; Goldman et al., 2009). Based on our argument that strategic firms enjoying

AC

business and financial privileges, we hypothesize that:

H1: Strategic firms demonstrate better financial performance relative to non-strategic firms.

Fan et al. (2007) suggest that firms appointing politically connected top managers and directors may exhibit rent-seeking behavior, their appointment and possibly turnover decisions for chairperson/CEO can be driven by different criteria (i.e., more than maximizing shareholder’s wealth).

In addition, Shen, Lin and Wang (2015) find it is more likely for

politically connected firms to demonstrate poor governance practice.

12

Clearly the CEO/board

ACCEPTED MANUSCRIPT member appointment in the strategic firms under the heavy influence of the government is a form of strategic decision.

We expect that management turnover of strategic firms is subject to

T

political decisions of government, which is exogenous and not always related to performance,

RI P

therefore:

SC

H2a: Strategic firms demonstrate a different pattern of management turnover relative to that of

NU

non-strategic firms.

H2b: Top managers of strategic firms with political ties are less likely to be replaced even during

MA

financial distress.

ED

Our third hypothesis is constructed to explore the possible difference in post-turnover

PT

performance of strategic and non-strategic firms facing financial distress. One of the major reasons for forced top management turnover is poor performance (DeFond & Hung, 2004; Huson These firms aim to improve performance after the replacement of

CE

et al., 2001).

Huson et al. (2004), among others, have shown that there

AC

poorly-performing top management.

is improved performance after forced turnover using the US data. Also, Chang and Wong (2009) show post-turnover profitability improvement in loss-making firms.

Therefore, based

on the conjecture that the financial well-being of strategic firms is protected by the government, we expect a speedier financial rebound of strategic firms after their financial difficulty for the strategic sample relative to their counterpart.

H3: Post-turnover performance improvement for financially distress firms is better for strategic firms than for non-strategic firms.

13

ACCEPTED MANUSCRIPT

3. Sample and Variables

T

Our study covers a sample period from 2001 to 2008. The sample includes all firms

RI P

(except those in the finance sector) listed on the Shanghai Stock Exchange and Shenzhen Stock Exchange. We obtain our data on company returns, financial statements, board characteristics

SC

and ownership data from the China Stock Market and Accounting Research (CSMAR) database.

NU

We hand-collect the data of political connections from the annual reports. Our poor performing firm sample comes from two sources. We define poorly performing

MA

firms as those firms which are a) under the special treatment (ST) listing status due to regulation violation as well as inappropriate corporate behavior4 and b) have poor performance of negative

ED

returns, negative return on assets and negative change in return on assets in the turnover year.

PT

The special treatment announcement is a unique mandatory reporting requirement imposed by China Securities Regulatory Commission (CSRC) on January 1, 1998 that when the firms have

CE

abnormality in the financial conditions, the Shanghai Stock Exchange and Shenzhen Stock

AC

Exchange have to announce that the firms are under special treatment and their short stock names are to be tagged with “ST” to alert the investors that the stocks may have the risk of delisting. The ST data come from the Corporate Governance file of CSMAR.

3.1 Defining Protected Industry and Political Connection In China, a number of industries (e.g., energy, raw materials, petrochemicals) are under government protection.

Nee et al. (2007) argue that the Chinese central government retains

4

Cheng, Leung, and Yu (2014, p. 71) provide detailed description of inappropriate corporate behaviors, which include two consecutive years of losses and/or asset shrinkage, ,adverse audit opinion on financial statements, a shrinkage of rights and interest, operation impairment or halt, major lawsuits, bankruptcy, violation of information disclosure rule, and investment risk.

14

ACCEPTED MANUSCRIPT control over the firms in the strategically and politically important industries, which include utilities, telecommunications and energy. We term these industries as protected industries.

T

The firms in the protected industries are usually large monopolies, able to receive subsidies and

direct control of the State Council (Aharony et al., 2000).

RI P

special treatment (e.g., less stringent profitability requirement in new equity issues) and under In addition, there are some industries

SC

which are under special government regulation with restriction on share ownership (Sun & Tong,

NU

2003). These industries include utilities, energy, telecommunication and finance. Although these industries are under more government regulation and intervention, the firms in these

industries as regulated industries.

MA

industries also receive more government support and have less competition. We term these In short, we group the firms in protected industries and

ED

regulated industries as firms possessing national security interest.

PT

Holburn and Zelner (2010) define political resources as stocks of available political factors to which the firm gains access, primarily ties with pivotal political actors such as government

CE

officials and interest groups that seek to influence them. Faccio (2006, p. 369) classifies a firm

AC

as politically connected if at least one of its large shareholders (holding at least 10 percent of voting shares directly or indirectly) or one of its top directors (chairperson, CEO, president, vice-president or secretary) is a member of parliament, or a minister, or is closely related to a top politician or party.5

Fan et al. (2007) define politically connected CEOs as those who serve as

current or former government bureaucrats (current or former officers of central or local governments or the military).

In this study, we follow a similar sprit to define a connected firm

but employ a more diversified measure by requiring a certain political connection for three board members.

5

Close relationship is defined to include friendship, former heads of state or prime ministers, past directorship held, foreign politicians or long-standing relationships with political parties.

15

ACCEPTED MANUSCRIPT By political connection, since chairpersons, CEOs and directors are all full-time positions in a firm, these executives are defined as having political connections if they were (not currently are,

T

as they cannot hold two full-time positions) formerly government officials. In addition, they also

RI P

have political connections if they are current members of People’s Congress and/or People Political Consultative Conference (PPCC) (these are honorary and part-time positions).

We

716 politically connected firms.

This definition of political connection yields

NU

chairperson or CEO being politically connected.

SC

categorize a firm as politically connected if the firm has at least two directors with either the

By changing this definition to at least three directors with

MA

either the chairperson or CEO being connected and at least one director with either chairperson or CEO being connected; the overall sub-sample sizes with political connection become 382 and We determine that in order to identify firms with sufficient political

ED

1,266, respectively.

PT

connection strength, the criterion of at least one director with either chairperson or CEO being connected may not be strong enough while at least three directors with either chairperson or

CE

CEO being connected results in too small a sub-sample size for analysis, especially for the forced

AC

turnover analysis focusing on poor performing firms in Tables 4 and 5. Consequently, we adopt the criterion of at least two directors with either chairperson or CEO being connected as our definition of political connection. Using this approach, we have a total of 716 politically connected firms out of 7,959 observations in our sample. Table 1 shows the descriptive statistics of the variables. which 1,590 are strategic firms.

We have 7,959 observations, of

Further decomposition of the strategic firm sample shows

1,095 protected firms and 716 connected firms, within which, 222 firms are both protected and connected.

The chairperson and CEO have a mean tenure of 3.88 years and 3.27 years,

respectively. The chairpersons (mean age = 49.83) are older than the CEOs (mean age = 45.94).

16

ACCEPTED MANUSCRIPT The chairpersons and CEOs are similar in their education levels, with 3.64 and 3.59 years of schooling, respectively.

The firms have a mean board size of 9.8 directors. The board has a

T

median of about one-third of independent directors. The directors hold small shareholding

1,579 are central SOEs and 4,033 are local SOEs.

RI P

percentage (1.58%). There are 1,081 firms with CEO duality. Out of 7,959 observations, In these 5,612 SOE observations, only 600

SC

(10.7%) are classified as connected firms (as the remaining 116 connected firms are non-SOEs).

NU

There are 1,366 and 1,791 firms replacing their chairpersons and CEOs, respectively. In our subsample of 896 poor performing firm observations, 410 firms experience top For these poorly performing firms, the mean ROA∆, mean Ret∆ and

MA

management turnovers.

mean MB∆ are -0.1085, -0.1098 and -0.9305, respectively.

4. Empirical Results

CE

4.1 Regression Models

PT

ED

[Insert Table 1 here]

AC

We have the following regression models to test our hypotheses. ROA/Ret/MB = 0 + 1 Strategic + 2 Board(Control) + 3 Firm(Control) + t tYeart + j j Industryj

(1)

CHTurn/CEOTurn = 0 + 1 Strategic + 2 Strategic*PC + 3 DemoV + 4 Board(Control) + 4 Firm(Control) + t tYeart + j j Industryj

(2)

ROA∆/Ret∆/MB∆ = 0 + 1 Strategic + 2 DemoV + 3 Board(Control) + 4 Firm(Control) + t tYeart + j j Industryj

17

(3)

ACCEPTED MANUSCRIPT Performance is measured in terms of annual stock return (Ret), market-to-book ratio (MB) and return on assets (ROA). We classify those firms in the industry sectors of finance6, utilities,

Next, if the firms have at least two directors, and either the chairperson

RI P

and special regulation.

T

energy, telecommunications, natural resources, and mass media as under government protection

or CEO are politically connected, they are labeled as connected firms. There are 611 connected

SC

chairperson observations and 285 connected CEO observations (out of which 180 strategic firms

NU

have connected chairpersons and CEOs at the same time) in our strategic sample. Using this approach, we have a total of 1,095 protected and 716 connected firm

sample (1,590 observations).

MA

observations (out of which 222 have dual identity).

Together, it constitutes our Strategic

There are 600 observations out of 5,612 SOE observations

ED

(10.7%) or a total sample of 7,959 (7.5%) are classified as connected firms. Such percentages

Mainland government.

PT

indicate that not all SOEs are strategically positioned through political connection by the Strategic is a dummy coded 1 if the firm is in the industries under

CE

government protection/regulation or is politically connected and 0 otherwise.

Protected is a

AC

dummy coded 1 if the firm is in the industries under government protection/regulation and 0 otherwise. PC is a dummy coded 1 if the firm is a politically connected firm and 0 otherwise. CHTurn is a dummy coded 1 if the chairperson is replaced and 0 otherwise.

CEOTurn is a

dummy coded 1 if the CEO is replaced and 0 otherwise. Replacement of poorly performing top management is an essential element of effective governance (Shleifer and Vishny, 1997) and also an important step for improvement.

McNeil

et al. (2004) and Huson et al. (2004) have shown that there is improved performance after forced turnover.

Barney (1991) suggests that the difference in resource endowments explains the

6

In our sample, we exclude firms in the finance sector. Therefore, although firms in the finance sector are classified as firms under government protection and special regulation, they are not included in the analysis.

18

ACCEPTED MANUSCRIPT difference in performance.

If we argue that the business advantages and protection enjoyed by

the strategic firms is a form of resources, then post-turnover performance should be related to

Ret is the change in return (year +1 minus year -1).

SC

year -1).

ROA is the change in return on assets (year +1 minus

RI P

measuring post-turnover performance.

T

political connection and government protection. We have three performance indicators for

market-to-book ratio (year +1 minus year -1).

Cheng, Chan and Leung (2010)

NU

We control for management demography in our analysis.

MB is the change in

show that management demography (e.g., age, education) and relational demographic differences

MA

between chairperson and CEO may affect financial performance (earnings per share, return on assets, market-to-book ratio and abnormal return) in Chinese firms.

If a firm would optimize

ED

the combination of the strategic position (government protection and political connections) and

PT

management capability, then the management demographic profile of strategic firms should be different from those of non-strategic firms. DemoV is the group of demographic variables of CH and CEO are the prefixes to represent the characteristics of

CE

chairperson and CEO.

Ten is the number of years staying in office.

Age is the

AC

chairpersons and CEOs, respectively. age. Edu is the number of schooling.7

We also control for the relational demography of top management team (TMT) (Chan et al., 2010).

These variables include DiffTen, which is the differential tenure of chairperson minus

that of CEO. DiffAge is the net value of the age of chairperson minus that of CEO.

DiffEdu is

the net number of schooling of chairperson minus that of CEO. In corporate finance, turnover is related to performance and governance.

Coles,

McWilliams, and Sen (2001), in the examination of the relation of governance mechanism and 7

For the number of schooling, we use three years for college education, four years for university education, six years for master degree and nine years for Ph.D degree.

19

ACCEPTED MANUSCRIPT performance, imply that the effectiveness of corporate governance mechanism is difficult to be identified and measured.

Removal of poorly performing CEO is a necessary condition of good

T

governance, and hence turnover-performance sensitivity is a measure of governance quality Volpin (2002) and DeFond and Hung (2004) suggest that the firms

RI P

(Shleifer & Vishny, 1997).

which do not dismiss the poorly performing management have weak governance. DirIndR is the proportion of independent

SC

We have four variables for Board(Control).

Borokhovich, Parrino,

NU

directors to board size, which is our measure of board independence.

and Trapani (1996) find that the likelihood of executive turnover (voluntary and forced)

MA

increases monotonically with the percentage of independent directors. percentage of shares held by directors to total shares.

DirHold is the

Jensen and Meckling (1976) argue that

Chen (2014) and Cheng et al. (2010) argue that board size is related to

PT

reduce agency cost.

ED

awarding shares to top management helps align the interests of managers and shareholders to

Therefore, we include BoardSize, which is the number

of directors on board in our models.

CEO duality is related to turnover and probability of

CE

corporate performance and governance.

Dual is a dummy coded 1 if the chairperson and

AC

financial distress (Simpson & Gleason, 1999). CEO are the same person and 0 otherwise. We have six variables for Firm(Control).

CentSOE is a dummy coded 1 if the firm is a

state-owned enterprise (SOE) controlled by central government and 0 otherwise.

LocalSOE is a

dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or county government and 0 otherwise.

Aharony et al. (2000) report that firms in protected industries are

state monopolies under direct control of State Council and Firth et al. (2006) find that turnover-performance sensitivity is higher for firms with major shareholders being legal entities. ROA is return on assets.

DA is debt to asset ratio.

20

LnAsset is log of total assets.

FirmAge is

ACCEPTED MANUSCRIPT the number of years the firm is established.

Jensen (1986) suggests that leverage is a corporate

governance measure as debtors also play a role in disciplining managers. We use firm size

In addition, large firms have more resource advantage than small firms (Barney,

RI P

larger firms.

T

(LnAsset) as a control as Dalton and Kesner (1983) find that managers are more entrenched in

SC

1991).

NU

4.2 Results for Hypothesis 1 (H1)

Aharony et al. (2000) show that the firms in protected industries are usually large

MA

monopolies, able to receive subsidies and special treatment (e.g., less stringent profitability requirement in new equity issues) from the government support and have less competition.

ED

Studies on political connections also document that politically connected firms have better

PT

performance (e.g., Cooper et al., 2010; Fisman, 2001) and enjoy various benefits (e.g., Goldman et al., 2009). Therefore, we expect the strategic firms (Strategic) with government protection

CE

and political connection should perform better.

In Table 2, we report the results for firm

AC

performance in terms of return on assets (ROA), stock price return (Ret) and market-to-book ratio (MB). Consistent with our expectation, the coefficients on Strategic are positively related to ROA (0.011), Ret (0.064) and MB (0.437), all significant at 0.01 level.

We have also conducted

our performance analysis using dummies for protected firms (Protected) and connected firms (PC) separately.

Except for the coefficient on PC in the run of regression for ROA, the

coefficients on Protected and PC are all positively related to ROA, Ret and MB.

These results

indicate that strategic firms perform better than non-strategic firms and support our H1. [Insert Table 2 here]

21

ACCEPTED MANUSCRIPT We further explore if there exists a different demographic profile of top managers in strategic vs non-strategic firms.

Table 3 reports the binary logit regressions in which the

T

dependent variables are Strategic, Protected and PC dummies. Our results show that there exist

education.

RI P

some significant differences in demography in terms of chairperson age, CEO age and CEO In addition, there are also some differences in one governance variable (board size)

SC

and a few firm characteristics (SOEs, debt level, firm size and firm age).

While these findings

demographic profiles.

NU

may be exploratory, they do indicate that strategic firms employ top managers with different This finding is consistent with the management literature that when a

MA

firm owns a unique type of valuable resources, it needs a particular type of human resources to exploit them so that the valuable resources and capabilities can be developed in combination in

ED

order to attain competitive advantage (Barney, 1991).

PT

[Insert Table 3 here]

CE

4.3 Results for Hypothesis 2 (H2)

AC

Table 4 presents the results for our H2 (a and b) which examines the relationship between strategic firm and turnover. The first two columns in Table 4 show the results for all turnovers (both voluntary and forced). We also conduct the analysis using poorly performing firms because we want to examine if there exists any difference in the impact of Strategic dummy on forced turnovers.

In view of the identification problem of forced and voluntary turnovers

(Chang & Wong, 2009; Huson et al., 2004), we argue that during crisis, when the firm is in trouble, there should be a strong incentive for the board and shareholders to replace top management, leading to a forced turnover.

In this forced turnover sample, there are 313 CEO

22

ACCEPTED MANUSCRIPT turnovers (CEOTurn) and 274 chairperson turnovers (CHTurn). The results are reported in the third and fourth columns in Table 4.8

T

For both total and forced turnover samples, the Strategic dummies are all positively

RI P

significant except for CEOTurn in the forced turnover sample. Economically, this statistical finding suggests that strategic firms are more likely to have higher management turnover than This finding supports H2a.

SC

their non-strategic counterparts.

Interesting enough, the

NU

interaction dummies (Strategic*PC) are significantly negative with a bigger coefficient than Strategic (except for CEOTurn in total sample).

This result lends support to our H2b.

MA

Combining these two coefficients together; we can generate an economic interpretation that politically connected firms have a lower probability to experience chairperson turnovers under

ED

both normal and poor performing situations.

The higher overall turnovers for strategic sample

at the national level.

PT

are due to the fact that strategic firms often execute state strategy on human capital deployment It is indeed a common practice of Mainland government to rotate senior For instance, on January 16, 2012, with the approval of China

CE

executives among SOEs.

AC

Banking Regulatory Commission, extraordinary general meeting (EGM) and board meeting, the Agricultural Bank of China (ABC) and China Construction Bank (CCB) announced that ABC would appoint Jiang Chaoliang (previously as the president of China Development Bank) as chairperson and CCB would appoint Wang Hongzhang (previously as the Party member of the People’s Bank of China) as chairperson. These incidents show that in Mainland China, top managers are being moved around in SOEs which may have little to do with manager performance in their own firms, but more to do with the needs of other SOEs and state agencies.

8

Besides examining the demographic characteristics of top management on individual level, we also repeat the analyses using the relational demographic difference between chairperson and CEO. We report the results on individual level in Table 3 and Table 4. The results on relational demographic difference are qualitatively similar and available upon request.

23

ACCEPTED MANUSCRIPT Another interesting observation about the stronger negative coefficient of Strategic*PC suggests that politically connected firms tend to maintain a more stable management team even facing Yuan (2011) also finds it is less likely to replace politically connected CEOs

T

poor performance.

RI P

even they have poor performance. Our finding is consistent with the common wisdom in China that when an individual or corporation is connected to the State, the State would protect the

SC

connected entities during bad times.

NU

[Insert Table 4 here]

In Table 4, for the total turnover sample, some control variables of management Many studies have shown that management turnover is positively

MA

demography are significant.

and negatively related to age and tenure (Kang & Shivdasani, 1995).

We find that there are

ED

more top manager turnovers if they are older and have lower education level.

Among the

PT

control variables for board governance and firm characteristics, we find that the coefficients on DirHold, ROA, LnAsset and FirmAge are significant and show consistent directional relations

CE

with CHTurn and CEOTurn.

Our results suggest that top management is more likely to

AC

experience turnover when director shareholding (DirHold) is lower and the firms have lower profitability (ROA), longer establishment time (FirmAge) and are smaller in size (LnAsset)9. However, Dual, CentSOED and LocalSOED show different signs with CHTurn and CEOTurn.

Dahya, McConnell, and Travlos (2002) suggest that chairpersons with CEO duality

have more power on board and may entrench longer tenure in the firms.

The coefficients on

Dual are not related to CHTurn. However, there are negatively significant relations between Dual and CEOTurn, indicating that the CEO is less likely to experience turnover when there is CEO duality. 9

Firth et al. (2006) also find that there is a lower likelihood for large firm to replace chairperson. They explain that large firms need more skilled managers who are usually in short supply, hence deterring large firms from dismissing the chairpersons.

24

ACCEPTED MANUSCRIPT Central SOEs and local SOEs are different in top management turnover decisions. There are more chairperson turnovers and less CEO turnovers in central SOEs.

The coefficients on

T

LocalSOED are not related to CHTurn, but are negatively and significantly related to CEOTurn.

RI P

Owing to fiscal decentralization, Chang and Wong (2009) suggest that local governments should focus more on economic performance than on achievement of social and political objectives.

SC

Therefore, the firms owned by local government should be more likely than those owned by the

NU

central government to discipline poorly performing CEOs. They find that firms owned by local governments are more likely to have CEO turnover. Our result is inconsistent with that of

MA

Chang and Wong (2009) as we find CHTurn and CEOTurn to have positive and negative associations with CentSOED, respectively.10

ED

Finally, we also divide the sample into strategic vs non-strategic subsamples to examine how

PT

top manager’s turnover may relate to demographic, governance, and firm characteristics. The results are reported in Appendix 1. Overall findings seem to suggest that the top manager’s

CE

turnover (especially for CEOs’) for strategic and non-strategic firms is related to different sets of

AC

governance and firm variables (with limited common variables).

4.4 Results for Hypothesis 3 (H3) Our final analysis (H3) examines whether the strategic firms in crisis show better performance improvements after replacing the top managers for the forced turnover sample. Our rationale is based on the ‘‘grabbing hand’’ argument of Shleifer and Vishny (1998).

If the

strategic firms are under state control to fulfill certain political agenda other than maximizing firm value, these firms should also receive business and financial benefits from government in 10

However, the study of Chang and Wong (2009) is mainly on the turnover of CEO. In addition, their robustness check also shows that the activity of chairperson on firm management affects managerial turnover, the more active the chairperson, the lower the probability of CEO turnover.

25

ACCEPTED MANUSCRIPT difficult times, which are exactly the case defined in our poor-performing subsample. Thus we expect these strategic firms should rebound better financially after the management turnovers.

T

The results are reported in Table 5. Our three performance indicators are change in return

RI P

on assets (ROA), change in stock returns (Ret) and change in market-to-book ratio (MB).

SC

The coefficients on Strategic are positively related to Ret (Rety+1 – Rety-1) and MB (MBy+1 – MBy-1), indicating that the strategic firms experience bigger performance improvements than This finding that the strategic firms perform better is

NU

those for the non-strategic firms. consistent with H3.11

MA

[Insert Table 5 here]

In Table 5, for the demographic characteristic variables, the subscript “N” indicates the We find that none of the demographic characteristics

ED

newly appointed chairperson and CEO.

PT

on the individual level is significant, except for GMEduN, which is significantly related to ROA.

CE

The insignificance of the demographic characteristics of the newly appointed top management suggests two possible interesting observations.

First, combining the finding of performance

AC

improvement for strategic firms and no relationship between performance and new top management demography, we conjecture that the replacement of poorly-performing top management as shown in Table 5 is an evidence of corporate governance (Shleifer & Vishny, 1997).

The western culture of accountability imposed on top managers often times leads to

their forced resignation even though the poor performance may not be entirely due to their fault. We argue that our result here is consistent with such a practice. Our second interesting observation is based on the result that no relationship is found between performance improvement and new top management demography but the performance 11

In the examination of managerial turnover in Chinese firms, Firth et al. (2006) report that turnover is negatively related to profitability, but find no evidence of post-turnover improvement in performance. Our results show performance improvement of strategic firms relative to non-strategic firms, which is a different benchmark.

26

ACCEPTED MANUSCRIPT improvement is positively related to Strategic.

Such a finding can be explained by the

argument that the performance improvement has little to do with the new managers’

T

demographic profile (i.e., can be viewed as a proxy of managerial ability) but more to do with

RI P

the business and financial advantages from government due to the government protection and/or political connection.

SC

In addition, we also measure the relational demography of the newly appointed top We find ROA and

NU

management team (TMT) for the characteristics of age and education level.

MB are negatively and significantly related to DiffEduN. This result suggests that the higher

MA

the education level of CEO as compared to that of the chairperson, the greater the increase in return on assets and market-to-book ratio.

Among the other control variables, we find that

ED

firms with larger board independence (DirIndR), higher shareholding percentage (DirHold) and

turnover.

PT

longer establishment time (FirmAge) have greater increase in stock return after management In addition, the local SOEs are more likely to have positive change in market-to-book

CE

ratio after turnover, a finding which is consistent with the argument of Chang and Wong (2009)

financial loss.

AC

that state shareholders have incentive to improve firm performance particularly during period of

4.5 Additional Analysis In our main analysis, we use PC which is a dummy for politically connected firm. For robustness purpose, we compile a politically connected index to measure the strength of political connectedness to repeat our analyses. This politically connected index is a weighted score index which takes into account the political connectedness of different types (former government officials, People’s Congress representatives or PPCC committee members) and at various levels

27

ACCEPTED MANUSCRIPT (from central or national down to county levels).

For the type of connectedness with

government officials, there are four levels: central, provincial, city and county.

We assign 1

T

score for deputy official and 2 scores for chief official at county level; 3 scores for deputy

RI P

official and 4 scores for chief official at city level; 5 scores for deputy official and 6 scores for chief official at provincial level; and 7 scores for deputy official and 8 scores for chief official at For the type of connectedness with the representatives of People’s Congress,

SC

the central level.

We assign political

NU

there are also four levels: national, provincial, city and county.

connectedness scores to People’s Congress membership, which are 4, 3, 2 and 1 for national,

MA

provincial, city and county levels, respectively.

For the type of connectedness with the

committee members of PPCC, there are two levels: national and regional.

We assign the score

Our new measure of political connectedness, PCScore, is the natural logarithm

PT

national PPCC.

ED

of 1 for connectedness with committee members of regional PPCC and the score of 2 for that of

of the total sum of connectedness scores for all the board members (chairperson, CEO, and

CE

directors). We repeat the regression analysis on management demography and turnover using

AC

PCScore instead of PC and report the results in Table 6. For the analysis on management demography, we show that there exist significant differences in demography for chairperson and CEO for politically and non-politically connected firms.

For the analysis on management

turnover, the overall results are weaker for the PCScore variable but still significant for chairperson, meaning that politically connected firms do not change their chairpersons as much relative to non-politically connected firms.12 [Insert Table 6 here]

12

We thank the anonymous reviewer for suggesting the politically connected index as alternative measure of political connectedness.

28

ACCEPTED MANUSCRIPT To address the concern for the potential endogeneity problem that the government decision for keeping firms under government control and political connectedness decision are not random,

We match each of the strategic firms with unique non-strategic firms

without using the same firm again for matching.

RI P

with strategic firms.13

T

we apply the propensity score matching (PSM) method to select non-strategic firms to match

We conduct the original analyses of Tables

SC

2-4 using the PSM approach by 1,590 pairs of strategic and non-strategic firms (in total 3,180

NU

observations). The results are reported in Tables 7-9. As shown in Tables 7-9, the results using 7,959 observations and 3,180 observations are qualitatively the same. Our main findings

MA

that: (1) strategic firms perform better than non-strategic firms; (2) strategic firms employ top managers with different demographic profiles; (3) strategic firms are more likely to have

ED

management turnovers; and (4) politically connected firms have a lower probability to

PT

experience chairperson turnovers under normal and poor performing situations, remain the same.

AC

CE

This suggests that our results are robust after considering the concern for endogeneity. [Insert Table 7 here] [Insert Table 8 here] [Insert Table 9 here]

5. Conclusion Nowadays, emerging market governments are able to employ a full range of sophisticated strategies to enhance the effectiveness of state-owned assets in improving their economies.

In

this paper, we use data from a transitional economy, China, which provides us an opportunity to explore two related government strategies deployed on listed firms, government protection as well as political connections. 13

Declaring a protection status is a way to protect national security

We thank the anonymous reviewer for suggesting the PSM method to address the endogeneity concern.

29

ACCEPTED MANUSCRIPT interest and these protected and regulated firms are large monopolies which enjoy the benefits of government subsidies and less competition. The ties through political connections are also

T

valuable economic interest as “guanxi (relationship)” provides important networks to reduce We show that these strategic firms are related to firm

RI P

transaction cost (Burt, 1997).

performance and management turnover, especially during firm-level crisis.

SC

Based on the conjecture that the business benefits and therefore financial performance of

NU

strategic firms do not rely heavily on top managers, the post-turnover performance change of strategic firms is independent of demographic profiles of new top managers.

MA

economy, there are many recently privatized SOEs and protected firms.

In a transitional

This study sheds light

on how the Chinese government’s strategic firms perform financially and demonstrate a different Political connection literature normally defines connected

ED

pattern of management turnover.

PT

firms by the personal connection of a single CEO.

Our paper defines a politically connected

firm by using a more restrictive standard: when the chairperson or CEO plus at least two

CE

directors have former ties in the governments at various forms and levels.

This measure may be

AC

a more comprehensive proxy of political networks. Our analysis on the effect of political connection shows that connected firms experience few turnovers in the total and forced turnover samples, indicating that top managers in connected firms may enjoy protection from the board, which can be under the influence of the government. We do not, however, condemn such protection as these strategic firms also enjoy better financial performance during the normal times and enjoy a stronger financial improvement in the post-turnover year after facing the financial distress.

Such a two-way street of benefits and

responsibilities has well been documented in the political connection literature.

We add to the

literature by demonstrating consistent evidence related to management turnovers.

30

Finally,

ACCEPTED MANUSCRIPT additional analysis in future research is needed to better understand the main drivers of

AC

CE

PT

ED

MA

NU

SC

RI P

T

management turnovers for strategic and non-strategic firms.

31

ACCEPTED MANUSCRIPT Acknowledgements

T

The authors acknowledge that this research project was financially supported by the Research

Polytechnic University Project No. PolyU1471/11H).

We thank Professor Michael Firth for his

AC

CE

PT

ED

MA

NU

SC

valuable comments.

RI P

Grants Council of the Hong Kong Special Administrative Region, China (Hong Kong

32

ACCEPTED MANUSCRIPT References Agrawal, A., & Knoeber, C.K. (2001). Do some outside directors play a political role? Journal of

T

Law and Economics, 44, 179-198.

RI P

Aharony, J.C., Lee, W.J., & Wong, T.J. (2000). Financial packaging of IPO firms in China. Journal of Accounting Research, 38, 103-126.

SC

Ambler, T., Styles, C., & Wang, X. (1999). The effect of channel relationships and guanxi on the

NU

performance of inter-province export ventures in the People’s Republic of China. International Journal of Research in Marketing, 16 (1), 75-88.

MA

Barney, J.B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17, 99-120.

ED

Bass, A.E., & Chakrabarty, S. (2014). Resource security: Competition for global resources,

PT

strategic intent, and governments as owners. Journal of International Business Studies, 45, 961-979.

CE

Boddewyn, J.J., & Brewer, T.L. (1994). International-business political behavior: New

AC

theoretical directions. Academy of Management Review, 19 (1), 119-143. Borisova, G., & Megginson, W.L. (2011). Does government ownership affect the cost of debt? Evidence from privatization. Review of Financial Studies, 24, 2693-2737. Borokhovich, K.A., Parrino, R. & Trapani, T. (1996). Outside directors and CEO selection. Journal of Financial and Quantitative Analysis, 31, 377-397. Boubakri, N., Guedhami, O., Mishra, D., & Saffar, W. (2012). Political connections and the cost of equity capital. Journal of Corporate Finance, 18, 541-559. Burt, R. (1997). The contingent value of social capital. Administrative Science Quarterly, 42, 339-363.

33

ACCEPTED MANUSCRIPT Calomiris, C.W., Fisman, R., & Wang, Y. (2010). Profiting from government stakes in a command economy: Evidence from Chinese asset sales. Journal of Financial Economics, 96,

T

399-412.

RI P

Chan, R.Y.K., Cheng, L.T.W., & Leung, T.Y. (2010). Corporate performance implications of relational demographic differences: On age and titles of chairpersons vs. general managers of

SC

listed Chinese companies. British Journal of Management, 22, 96-113.

NU

Chang, E.C., & Wong, S.M.L. (2009). Governance with multiple objectives: Evidence from top executive turnover in China. Journal of Corporate Finance, 15, 230-244.

MA

Chen, I. (2014). Financial crisis and the dynamics of corporate governance: Evidence from Taiwan’s listed firms. International Review of Economics and Finance, 32, 3-28.

ED

Cheng, L.T.W., Chan, R.Y.K., & Leung, T.Y. (2010). Management demography and corporate

PT

performance: Evidence from China. International Business Review, 19, 261-275. Cheng, L.T.W., Leung, T.Y., & Yu, W. (2014). Information arrival, changes in R-square and

AC

67-81.

CE

pricing asymmetry of corporate news. International Review of Economics and Finance, 33,

Claessens, S., Feijen, E., & Laeven, L. (2008). Political connections and preferential access to finance: The role of campaign contributions. Journal of Financial Economics, 88, 554-580. Coles, J.W., McWilliams, V.B., & Sen, N. (2001). An examination of the relationship of governance mechanisms to performance. Journal of Management, 27, 23-50. Cooper, M.J., Gulen, H., & Ovtchinnikov, A.V. (2010). Corporate political contributions and stock returns. Journal of Finance, 65 (2), 687-724. Cull, R., & Xu, L.C. (2005). Institutions, ownership, and finance: The determinants of profit reinvestment among Chinese firms. Journal of Financial Economics, 77 (1), 117-146.

34

ACCEPTED MANUSCRIPT Dalton, D.R., & Kesner, I.F. (1983). Inside/outside succession and organizational size: The pragmatics of executive replacement. Academy of Management Journal, 26, 736-742.

RI P

and top management turnover. Journal of Finance, 57, 461-483.

T

Dahya, J., McConnell, J., & Travlos, N. (2002). The Cadbury committee, corporate performance,

DeFond, M.L., & Hung, M. (2004). Investor protection and corporate governance: Evidence

SC

from worldwide CEO turnover. Journal of Accounting Research, 42, 269-312.

NU

Economist (2012, Jan, Special Issue). How State Enterprise is Spreading. Faccio, M. (2006). Politically connected firms. American Economic Journal, 96, 369-386.

MA

Fama, E. (1980). Agency problems and the theory of the firm. Journal of Political Economy, 88, 103-288.

ED

Fan, J.P.H., Wong, T.J. & Zhang, T. (2007). Politically connected CEOs, corporate governance,

Economics, 84, 330-357.

PT

and post-IPO performance of China’s newly partially privatized firms. Journal of Financial

CE

Farrell, K.A., & Whidbee, D.A. (2003). Impact of firm performance expectations on CEO turnover

AC

and replacement decisions. Journal of Accounting and Economics, 36 (1-3), 165-196. Firth, M., Fung, P.M.Y., & Rui, O.M. (2006). Firm performance, governance structure and top management turnover in a transitional economy. Journal of Management Studies, 43, 1289-1330. Fisman, R. (2001). Estimating the value of political connections. American Economic Review, 91, 1095-1102. Goldman, E., Rocholl, J., & So, J. (2009). Do political connected boards affect firm value? Review of Financial Studies, 22, 2331-2360.

35

ACCEPTED MANUSCRIPT Groves, T., Hong, Y., McMillan, J., & Naughton, B. (1995). China’s evolving managerial labor market. Journal of Political Economy, 103, 873-892.

T

Hillman, A.L. (1982). Declining industries and political-support protectionist motives. The

RI P

American Economic Review, 72 (5), 1180-1187.

Holburn, G.L.F., & Zelner, B.A. (2010). Political capabilities, policy risk and international

SC

investment strategy: Evidence from the global electric power generation industry. Strategic

NU

Management Journal, 31, 1290-1315.

Huson, M., Parrino, R., & Starks, L. (2001). Internal monitoring mechanisms and CEO turnover:

MA

A long-term perspective. Journal of Finance, 56, 2265-2298. Huson, M., Malatesta, P., & Parrino, R., (2004). Managerial succession and firm performance.

ED

Journal of Financial Economics, 74, 237-275.

PT

Jensen, M.C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review, 76, 323-329.

CE

Jensen, M., & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency costs and

AC

ownership structure. Journal of Financial Economics, 3, 305-360. Jenter, D., & Kanaan, F. (2015). CEO turnover and relative performance evaluation. Journal of Finance, forthcoming, DOI: 10.1111/jofi.12282. Jones, S.L., Megginson, W.L., Nash, R.C., & Netter, J.M. (1999). Share issue privatizations as financial means to political and economic ends. Journal of Financial Economics, 53, 217-253. Kang, J., & Shivdasani, A. (1995). Firm performance, corporate governance, and top executive turnover in Japan. Journal of Financial Economics, 38, 29-58. Lajili, K., & Zéghal, D. (2006). Market performance impacts of human capital disclosure. Journal of Accounting and Public Policy, 25, 171-194.

36

ACCEPTED MANUSCRIPT Li, H., & Atuahene-Gima, K. (2001). Product innovation strategy and the performance of new technology ventures in China. Academy of Management Journal, 44 (6), 1123-1134.

T

Li, J.J., Poppo, L., & Zhou, K.Z. (2008). Do managerial ties in China always produce value?

RI P

Competition, uncertainty, and domestic vs. foreign firms. Strategic Management Journal, 29 (4), 383-400.

SC

McNeil, C., Niehaus, G., & Powers, E. (2004). Management turnover in subsidiaries of

NU

conglomerates versus stand-alone firms. Journal of Financial Economics, 72, 63-69. Megginson, W.L., Nash, R.C., Netter, J.M., & Poulsen, A.B. (2004). The choice of private

MA

versus public capital markets: Evidence from privatizations. Journal of Finance, 59 (6), 2835-2870.

ED

Melitz, M. (2005). When and how should infant industries be protected? Journal of International

PT

Economics, 66, 177-196.

Murphy, K.J. (1999). Executive compensation. In O. Ashenfelter and D. Card (Eds.), Handbook of

CE

Labor Economics (pp. 2485-2563). Amsterdam: Elsevier.

AC

Nee, V., Opper, S., & Wong, S.M.L. (2007). Developmental state and corporate governance in China. Management and Organization Review, 3 (1), 19-53. Park, S.H., & Luo, Y. (2001). Guanxi and organizational dynamics: Organizational networking in Chinese firms. Strategic Management Journal, 22, 455-477. Parrino, R., Sias, R.W., & Starks, L.T. (2003). Voting with their feet: Institutional ownership changes around forced CEO turnover. Journal of Financial Economics, 68 (1), 3-46. Peng, M.W., & Luo, Y. (2000). Managerial ties and firm performance in a transition economy: The nature of a micro-macro link. Academy of Management Journal, 43 (3), 486-501.

37

ACCEPTED MANUSCRIPT Shen, C.H., Lin, C.Y., & Wang, Y.C. (2015). Do strong corporate governance firms still require political connection and vice versa? International Review of Economics and Finance, 39,

T

107-120.

RI P

Shleifer, A., & Vishny, R.W. (1997). A survey of corporate governance. Journal of Finance, 52, 737-783.

NU

cures. Cambridge, MA: Harvard University Press.

SC

Shleifer, A., & Vishny, R.W. (1998). The grabbing hand: Government pathologies and their

Simpson, W.G., & Gleason, A.E. (1999). Board structure, ownership, and financial distress in

MA

banking firms. International Review of Economics and Finance, 8, 281-292. Su, Z.Q., Fung, H.G., Huang, D.S., & Shen, C.H. (2014). Cash dividends, expropriation, and

ED

political connections: Evidence from China. International Review of Economics and Finance,

PT

29, 260-272.

Sun, Q., & Tong, W.H.S. (2003). China share issue privatization: The extent of its success.

CE

Journal of Financial Economics, 70 (2), 183-222.

AC

Teece, D.J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18 (7), 509–533. Volpin, P. (2002). Governance with poor investor protection: Evidence from top executive turnover in Italy. Journal of Financial Economics, 64 (1), 61-90. Yuan, Q. (2011). Public governance, political connectedness, and CEO turnover: Evidence from Chinese state-owned enterprises. Working paper, Melbourne University.

38

ACCEPTED MANUSCRIPT Table 1 Descriptive Statistics

6878 6380 3926

19.0000 71.0000 9.0000 22.0000 75.0000 9.0000 0.6250 0.7481 19.0000

1.0000 26.0000 0.0000 1.0000 26.0000 0.0000 0.0000 0.0000 5.0000

2.8517 7.4514 2.5006 2.4503 6.7121 2.4484 0.1064 0.0789 2.2147

0.0320 2.5689 0.4843 20.2467 21.2052 27.8091 9.0000 26.0000

-3.9919 0.0081 14.9375 1.0000

0.1917 0.4944 1.0553 4.0892

-0.1085 -0.0246 4.5715 -0.1098 -0.0581 2.1648 -0.9305 -0.5983 37.8300

-4.0158 -2.8945 -38.2488

0.7311 0.8636 11.9270

NU

1081 1579 4033

ED

6593 6168

MA

0.0177 0.5108 21.3129 9.3851

1366 1791

3.0000 50.0000 4.0000 3.0000 45.0000 4.0000 0.3333 0.0000 9.0000

RI P

3.8803 49.8269 3.6404 3.2695 45.9354 3.5927 0.3038 0.0158 9.8404

T

Mean Median Maximum Minimum Standard Deviation

SC

Dummy Code = 1 Dummy Code = 0 1590 6369

PT

Strategic CHTen CHAge CHEdu CEOTen CEOAge CEOEdu DirIndR DirHold BoardSize Dual CentSOE LocalSOE ROA DA LnAsset FirmAge CHTurn CEOTurn ROA Ret MB

AC

CE

Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. CH is the prefix to represent the characteristic of chairperson. CEO is the prefix to represent the characteristic of CEO. Ten is the number of years staying in office. Age is the age. Edu is the number of schooling. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. ROA is return on assets. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. CHTurn is a dummy coded 1 if the chairperson is replaced and 0 otherwise. CEOTurn is a dummy coded 1 if the CEO is replaced and 0 otherwise. ROA is the change in return on assets. Ret is the change in return. MB is the change in market-to-book ratio.

39

ACCEPTED MANUSCRIPT

Table 2

IP

T

Analysis on Financial Performance

0.224 93.1 0.00 7959

0.226 86.8 0.00 7959

CE

0.224 93.1 0.00 7959

PT

ED

MA NU S

-0.489 0.00 -0.491 0.00 -0.471 0.00 -1.564 0.00 -1.814 0.00 -1.537 0.00 13.284 0.00 11.967 0.00 13.041 0.00 0.011 0.00** 0.064 0.00** 0.437 0.00** 0.013 0.02* 0.104 0.00** 1.115 0.00** 0.001 0.56 0.055 0.01** 0.287 0.02* 0.048 0.32 0.049 0.30 0.021 0.68 2.477 0.00** 2.281 0.00** 2.478 0.00** 1.832 0.07 2.171 0.03* 1.927 0.05* 0.110 0.00** 0.109 0.00** 0.093 0.00** 1.658 0.00** 1.802 0.00** 1.651 0.00** 1.411 0.00** 1.859 0.00** 1.414 0.00** -0.002 0.06 -0.002 0.08 -0.002 0.06 0.008 0.01** 0.006 0.05* 0.008 0.01** -0.013 0.50 -0.006 0.77 -0.012 0.54 -0.001 0.91 0.001 0.91 0.002 0.81 0.034 0.07 -0.044 0.03* -0.034 0.08 0.418 0.01** -0.451 0.00** -0.429 0.00** -0.001 0.90 -0.001 0.89 -0.001 0.85 0.011 0.56 0.022 0.30 0.010 0.60 -0.155 0.30 -0.149 0.32 -0.185 0.22 0.004 0.41 0.004 0.40 0.005 0.34 -0.031 0.05* -0.030 0.07 -0.030 0.05* -0.534 0.00** -0.553 0.00** -0.559 0.00** -0.117 0.00** -0.117 0.00** -0.118 0.00** -0.040 0.00** -0.042 0.00** -0.041 0.00** -0.351 0.01** -0.359 0.01** -0.358 0.01** 0.027 0.00** 0.027 0.00** 0.027 0.00** 0.045 0.00** 0.057 0.00** 0.045 0.00** -0.406 0.00** -0.390 0.00** -0.415 0.00** -0.001 0.04* -0.001 0.04* -0.002 0.01** 0.017 0.00** 0.022 0.00** 0.017 0.00** 0.047 0.00** 0.064 0.00** 0.053 0.00**

AC

Intercept Strategic Protected PC DirIndR DirHold BoardSize Dual CentSOE LocalSOE DA LnAsset FirmAge Industry Dummies included Year Dummies Included Adj R2 F-statistics p-value N

CR

ROA Ret MB Coeff p-value Coeff p-value Coeff p-value Coeff p-value Coeff p-value Coeff p-value Coeff p-value Coeff p-value Coeff p-value

0.423 177.9 0.00 7959

0.358 165.3 0.00 7959

0.423 177.8 0.00 7959

0.120 33.7 0.00 7959

0.104 29.8 0.00 7959

0.120 35.0 0.00 7959

ROA is return on assets. Ret is the annual stock return. MB is the market-to-book ratio. Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

40

ACCEPTED MANUSCRIPT Table 3 Analysis on Management Demography

T

0.00 0.00** 0.00** 0.04* 0.98 0.00** 0.77 0.03* 0.21 0.00** 0.33 0.00** 0.00** 0.36 0.00** 0.00** 0.00**

PC Coeff -11.066 0.020 0.020 0.011 -0.029 0.019 0.060 1.079 -0.503 0.141 -0.352 0.142 0.369 0.559 -0.267 0.253 -0.058 554.8 0.00 7959

RI P

-11.391 -0.057 -0.019 -0.030 0.000 0.022 0.004 -1.354 -0.977 0.059 -0.108 0.785 0.851 -0.666 -1.726 0.484 -0.055 607.6 0.00 7959

SC

0.00 0.47 0.04* 0.41 0.60 0.00** 0.02* 0.26 0.07 0.00** 0.17 0.00** 0.00** 0.87 0.00** 0.00** 0.00**

MA

-9.449 -0.009 -0.009 -0.011 -0.007 0.019 0.031 -0.597 -1.141 0.099 -0.134 0.508 0.755 0.112 -0.960 0.352 -0.045 1026.0 0.00 7959

ED

Intercept CHTen CHAge CHEdu CEOTen CEOAge CEOEdu DirIndR DirHold BoardSize Dual CentSOE LocalSOE ROA DA LnAsset FirmAge LR Statistic p-value N

Protected Coeff p-value

NU

Strategic Coeff p-value

p-value 0.00 0.24 0.00** 0.55 0.14 0.01** 0.00** 0.15 0.53 0.00** 0.02* 0.36 0.00** 0.18 0.15 0.00** 0.00**

AC

CE

PT

Protected is a dummy coded 1 if the firm is in the industries under government protection/regulation and 0 otherwise. PC is a dummy coded 1 if the firm is a politically connected firm and 0 otherwise. Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. CH is the prefix to represent the characteristic of chairperson. CEO is the prefix to represent the characteristic of CEO. Ten is the number of years staying in office. Age is the age. Edu is the number of schooling. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. ROA is return on assets. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

41

ACCEPTED MANUSCRIPT Table 4 Analysis on Management Turnover

SC

NU

ED

-0.115 -3.692 0.004 0.068 0.391 -0.060 -0.549 0.071 -0.157 0.032 373.0 0.000 7959

0.59 -0.196 0.77 1.528 0.40 0.00** 0.225 0.04* 0.902 0.02* 0.00** -0.117 0.38 -1.321 0.01** 0.09 -0.003 0.17 0.00** 0.014 0.24 0.00** -0.097 0.00** -0.011 0.35 0.019 0.00** -0.048 0.00** 0.85 -0.927 0.08 -0.880 0.46 0.00** -1.154 0.03* -2.685 0.45 0.77 0.021 0.12 0.034 0.39 0.46 -0.269 0.00** 0.089 0.67 0.00** -0.194 0.02* 0.227 0.33 0.44 -0.319 0.00** 0.048 0.80 0.01** -0.557 0.00** -0.104 0.45 0.23 0.036 0.52 0.026 0.57 0.00** -0.099 0.00** -0.154 0.07 0.00** 0.025 0.00** 0.014 0.53 228.2 80.6 0.000 0.00 7959 896

MA

0.393 0.519 -0.626 -0.020 0.025 -0.123

PT

Intercept Strategic Strategic*PC CHTen CHAge CHEdu CEOTen CEOAge CEOEdu DirIndR DirHold BoardSize Dual CentSOE LocalSOE ROA DA LnAsset FirmAge LR Statistic p-value N

RI P

T

Total Turnover Sample Forced Turnover Sample CHTurn CEOTurn CHTurn CEOTurn Coeff p-value Coeff p-value Coeff p-value Coeff p-value 2.558 0.14 -0.491 0.24 0.930 0.05*

-0.005 0.012 -0.053 -1.500 -1.018 0.028 -0.048 -0.270 -0.092 -0.211 -0.002 -0.134 0.010 65.6 0.00 896

0.08 0.34 0.08 0.20 0.68 0.48 0.81 0.23 0.60 0.13 0.97 0.09 0.65

AC

CE

CHTurn is a dummy coded 1 if the chairperson is replaced and 0 otherwise. CEOTurn is a dummy coded 1 if the CEO is replaced and 0 otherwise. Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. CH is the prefix to represent the characteristic of chairperson. CEO is the prefix to represent the characteristic of CEO. Ten is the number of years staying in office. Age is the age. Edu is the number of schooling. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. ROA is return on assets. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

42

ACCEPTED MANUSCRIPT Table 5 Analysis on Post-turnover Performance for Forced Turnover Sample

T

ROA Ret MB Coeff p-value Coeff p-value Coeff p-value Coeff p-value Coeff p-value Coeff p-value

PT

ED

MA

NU

SC

RI P

Intercept -0.168 0.84 -0.189 0.83 -1.527 0.04 -1.175 0.12 -2.455 0.88 -8.172 0.65 Strategic 0.021 0.78 0.019 0.79 0.314 0.00** 0.335 0.00** 3.204 0.05* 4.345 0.02* CHAgeN 0.001 0.84 0.004 0.48 0.038 0.70 CHEduN -0.020 0.10 -0.016 0.24 -0.490 0.06 CEOAgeN 0.001 0.85 0.006 0.25 -0.140 0.21 CEOEduN 0.034 0.04* 0.013 0.38 0.315 0.24 DiffAgeN 0.001 0.85 -0.001 0.80 0.081 0.34 DiffEduN -0.023 0.05* -0.013 0.26 -0.439 0.04* DirIndR 0.293 0.65 -0.060 0.93 2.268 0.00** 2.350 0.00** 1.544 0.86 -2.240 0.83 DirHold 0.267 0.65 -0.047 0.94 5.210 0.00** 4.965 0.00** 12.657 0.16 8.871 0.34 BoardSize 0.009 0.63 0.013 0.50 -0.023 0.20 -0.024 0.18 -0.435 0.16 -0.412 0.20 Dual -0.036 0.78 -0.002 0.99 -0.046 0.65 -0.034 0.74 0.110 0.95 -0.021 0.99 CentSOE -0.089 0.41 -0.081 0.46 -0.020 0.86 -0.025 0.82 1.378 0.48 1.264 0.48 LocalSOE 0.030 0.68 0.050 0.50 0.078 0.37 0.084 0.33 2.609 0.05* 2.364 0.06 DA -0.048 0.46 -0.045 0.49 0.011 0.69 0.013 0.63 -0.680 0.11 -0.651 0.10 LnAsset -0.014 0.75 -0.004 0.92 -0.011 0.77 -0.007 0.86 0.403 0.63 0.424 0.64 FirmAge 0.005 0.63 -0.003 0.72 0.051 0.00** 0.051 0.00** 0.106 0.48 0.139 0.32 Industry Dummies included Year Dummies Included Adjusted R-square 0.004 0.036 0.369 0.368 0.004 0.003 F Statistic 1.063 1.585 8.465 8.932 1.057 1.037 p-value 0.383 0..036 0.000 0.000 0.390 0.416

AC

CE

ROA is the change in return on assets (year +1 minus year -1). Ret is the change in return (year +1 minus year -1). MB is the change in market-to-book ratio (year +1 minus year -1). Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. CH is the prefix to represent the characteristic of chairperson. CEO is the prefix to represent the characteristic of CEO. Age is the age. Edu is the number of schooling. DiffAge is the age difference with the age of chairperson minus that of CEO. DiffEdu is the difference in number of schooling with the number of schooling of chairperson minus that of CEO. The subscript “N” indicates the newly appointed chairperson and CEO. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

43

ACCEPTED MANUSCRIPT Table 6 Analysis on Management Demography and Turnover using PCScore

0.89 0.00** 0.71 0.48 0.00** 0.49 0.00** 0.22 0.00** 0.00**

PT

0.00 7955

-0.082 -3.642 0.006 -0.064 0.404 -0.054 -0.551 0.071 -0.150 0.034

-0.244 0.279 -0.072

-0.012 0.019 -0.048 -0.913 -1.143 0.022 0.270 -0.191 -0.318 -0.558 0.036 -0.098 0.025

363.63 0.00 7959

228.62 0.00 7959

T

0.07 0.00** 0.00** 0.01** 0.01** 0.00** 0.00** 0.23 0.00** 0.77 0.00** 0.00** 0.27 0.67 0.00** 0.89

0.76 0.00** 0.00** 0.07 0.00** 0.00**

Forced Turnover Sample CHTurn CEOTurn Coeff p-value Coeff p-value

RI P

-0.007 0.010 0.014 -0.013 0.004 0.016 0.682 0.165 0.054 0.008 0.085 0.072 0.046 -0.007 0.108 0.000 0.119 27.754

0.223 0.530 -0.207 -0.022 0.025 -0.123

0.44 0.30 0.09

0.35 0.00** 0.00** 0.09 0.03* 0.11 0.00** 0.02* 0.00** 0.00** 0.53 0.00** 0.00**

0.10 0.30 0.16 0.78 0.75 0.43 0.69 0.12 0.65 0.13 0.94 0.10 0.44

0.72 0.04* 0.28

SC

0.00**

NU

-2.729

ED

p-value

CE

Intercept Strategic Strategic*PCScore CHTen CHAge CHEdu CEOTen CEOAge CEOEdu DirIndR DirHold BoardSize Dual CentSOE LocalSOE ROA DA LnAsset FirmAge Adjusted R2 F-statistic LR Statistic p-value N

Coeff

Total Turnover Sample CHTurn CEOTurn Coeff p-value Coeff p-value

MA

PCScore

2.918 0.09 1.378 0.753 0.03* -0.488 -0.351 0.06 0.377 -0.003 0.15 0.010 0.39 -0.097 0.00** -0.005 0.013 -0.042 -1.124 0.41 -0.383 -3.729 0.32 -0.782 0.039 0.34 0.031 -0.106 0.60 0.080 0.218 0.34 -0.346 0.083 0.65 -0.079 -0.100 0.46 -0.206 0.032 0.47 0.003 -0.201 0.01** -0.128 -0.005 0.81 0.017

58.28 0.00 896

72.90 0.00 896

AC

PCScore is the natural logarithm of the total score of the different rankings in government hierarchy, People’s Congress and People Political Consultative Conference held by the board members. Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. CHTurn is a dummy coded 1 if the chairperson is replaced and 0 otherwise. CEOTurn is a dummy coded 1 if the CEO is replaced and 0 otherwise. CH is the prefix to represent the characteristic of chairperson. CEO is the prefix to represent the characteristic of CEO. Ten is the number of years staying in office. Age is the age. Edu is the number of schooling. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. ROA is return on assets. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

44

ACCEPTED MANUSCRIPT Table 7 Analysis on Financial Performance using Propensity Score Matching Method ROA Ret MB Coeff p-value Coeff p-value Coeff p-value

0.178 28.5 0.00 3180

6.595 0.474 4.339 0.644 -0.002 0.118 -0.456 -0.758 -0.428 -0.127 -0.001

T

0.00** 0.01** 0.00** 0.00** 0.79 0.01** 0.39 0.51 0.23 0.00** 0.24

RI P

-0.680 0.052 1.060 1.657 0.001 0.079 0.029 -0.019 -0.033 0.043 0.004

SC

0.00** 0.00** 0.91 0.00** 0.16 0.47 0.44 0.28 0.10 0.00** 0.00**

NU

-0.286 0.012 0.003 0.083 -0.001 0.004 0.005 0.006 -0.056 0.018 -0.003

MA

Intercept Strategic DirIndR DirHold BoardSize Dual CentSOE LocalSOE DA LnAsset FirmAge Industry Dummies included Year Dummies Included Adj R2 F-statistics p-value N

0.455 92.5 0.00 3180

0.00** 0.01** 0.00** 0.41 0.93 0.47 0.09 0.00** 0.17 0.18 0.95

0.114 13.4 0.00 3180

AC

CE

PT

ED

ROA is return on assets. Ret is the annual stock return. MB is the market-to-book ratio. Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

45

ACCEPTED MANUSCRIPT Table 8 Analysis on Management Demography using Propensity Score Matching Method Coeff

p-value

-6.813 -0.021 0.010 -0.003 -0.045 0.025 0.019 -0.291 -0.515 0.107 -0.281 0.038 0.154 -0.043 -0.717 0.217 -0.005 292.7 0.00 3180

-5.069 -0.034 -0.012 -0.014 -0.031 0.030 0.009 -1.398 -0.444 0.085 -0.156 0.402 0.394 -0.570 -1.078 0.211 -0.036 196.0 0.00 2192

-7.292 0.002 0.020 -0.022 -0.053 0.033 0.059 -0.003 -0.429 0.169 -0.456 -0.031 0.059 0.932 -0.992 0.145 0.014 147.3 0.00 1432

0.00** 0.95 0.03* 0.39 0.03* 0.00** 0.03* 1.00 0.61 0.00** 0.02* 0.88 0.72 0.49 0.01** 0.01** 0.44

SC

RI P

0.00** 0.06 0.12 0.50 0.11 0.00** 0.66 0.09 0.62 0.00** 0.27 0.01** 0.00** 0.62 0.08 0.00** 0.01**

NU

MA

0.00** 0.16 0.13 0.87 0.01** 0.00** 0.27 0.67 0.42 0.00** 0.02* 0.77 0.16 0.97 0.09 0.00** 0.66

PC

T

Protected Coeff p-value

ED

Intercept CHTen CHAge CHEdu CEOTen CEOAge CEOEdu DirIndR DirHold BoardSize Dual CentSOE LocalSOE ROA DA LnAsset FirmAge LR Statistic p-value N

Strategic Coeff p-value

AC

CE

PT

Protected is a dummy coded 1 if the firm is in the industries under government protection/regulation and 0 otherwise. PC is a dummy coded 1 if the firm is a politically connected firm and 0 otherwise. Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. CH is the prefix to represent the characteristic of chairperson. CEO is the prefix to represent the characteristic of CEO. Ten is the number of years staying in office. Age is the age. Edu is the number of schooling. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. ROA is return on assets. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

46

ACCEPTED MANUSCRIPT Table 9 Analysis on Management Turnover using Propensity Score Matching Method

0.174 0.95 -0.266 0.57 0.963 0.05*

RI P

SC

NU

ED

-1.324 -4.018 -0.030 -0.148 0.052 -0.327 -1.586 -0.184 -0.042 0.011 148.6 0.000 3180

0.14 -1.478 0.15 1.806 0.60 0.00** 0.361 0.01** 1.269 0.01** 0.00** -0.220 0.10 -1.393 0.02* 0.20 -0.009 0.03* 0.00** 0.049 0.07 0.00** -0.047 0.46 -0.013 0.47 0.025 0.00** -0.004 0.84 0.17 -2.044 0.01 -3.296 0.26 0.01** -0.032 0.97 -13.464 0.57 0.20 -0.022 0.28 -0.125 0.13 0.37 -0.329 0.03* -0.171 0.71 0.75 -0.082 0.57 0.554 0.21 0.02* -0.373 0.00** -0.169 0.67 0.04* -1.124 0.02* -0.483 0.18 0.36 -0.149 0.29 -0.140 0.36 0.41 -0.021 0.65 -0.171 0.26 0.49 0.022 0.10 0.014 0.76 72.7 39.9 0.000 0.07 3180 292

MA

-1.639 0.453 -0.544 -0.024 0.042 -0.109

PT

Intercept Strategic Strategic*PC CHTen CHAge CHEdu CEOTen CEOAge CEOEdu DirIndR DirHold BoardSize Dual CentSOE LocalSOE ROA DA LnAsset FirmAge LR Statistic p-value N

T

Total Turnover Sample Forced Turnover Sample CHTurn CEOTurn CHTurn CEOTurn Coeff p-value Coeff p-value Coeff p-value Coeff p-value

0.004 0.45 0.032 0.26 0.039 0.50 -6.746 0.00** 10.644 0.68 -0.140 0.08 -1.096 0.04* 0.302 0.50 -0.169 0.65 -0.667 0.10 -0.241 0.14 0.002 0.99 0.033 0.48 51.6 0.01 292

AC

CE

CHTurn is a dummy coded 1 if the chairperson is replaced and 0 otherwise. CEOTurn is a dummy coded 1 if the CEO is replaced and 0 otherwise. Strategic is a dummy coded 1 if the firm is in the industries under government protection/regulation or is politically connected and 0 otherwise. PC is a dummy coded 1 if the firm is a politically connected firm and 0 otherwise. CH is the prefix to represent the characteristic of chairperson. CEO is the prefix to represent the characteristic of CEO. Ten is the number of years staying in office. Age is the age. Edu is the number of schooling. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. ROA is return on assets. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

47

ACCEPTED MANUSCRIPT Appendix 1 Analysis between Strategic Firms and Management Turnover using Subsamples

109.339 0.000 1590

299.984 0.000 6369

-2.074

0.12

1.827

0.01

0.022 0.034 0.020 -1.691 -0.203 -0.052 -0.133 -0.020 -0.385 -1.384 0.137 0.001 0.027

0.36 0.00** 0.48 0.16 0.87 0.09 0.53 0.93 0.05* 0.05* 0.44 0.99 0.19

-0.023 0.014 -0.066 -1.813 -1.789 0.040 -0.280 -0.259 -0.262 -0.502 0.025 -0.152 0.001

0.10 0.00** 0.00** 0.00** 0.00** 0.01** 0.00** 0.01** 0.00** 0.00** 0.64 0.00** 0.87

SC

-0.103 -3.384 0.010 0.056 0.445 -0.012 -0.438 0.061 -0.162 0.033

0.33 0.19 0.00** 0.00**

0.88 0.00** 0.59 0.57 0.00** 0.89 0.02* 0.25 0.00** 0.00**

NU

0.35 0.06 0.52 0.60 0.99 0.05* 0.12 0.05* 0.20 0.03*

0.821 -0.018 0.018 -0.131

MA

1.296 -4.682 -0.021 0.122 -0.003 -0.451 -1.475 0.336 -0.084 0.053

0.02 0.61 0.00** 0.00**

ED

-3.648 -0.014 0.061 -0.094

PT

Intercept CHTen CHAge CHEdu CEOTen CEOAge CEOEdu DirIndR DirHold BoardSize Dual CentSOE LocalSOE ROA DA LnAsset FirmAge Industry Dummies included Year Dummies Included LR Statistic p-value N

RI P

T

CHTurn CEOTurn Strategic Sample Non-Strateg ic Sample Strategic Sample Non-Strateg ic Sample Coeff p-value Coeff p-value Coeff p-value Coeff p-value

52.667 0.005 1590

185.018 0.000 6369

AC

CE

CHTurn is a dummy coded 1 if the chairperson is replaced and 0 otherwise. CEOTurn is a dummy coded 1 if the CEO is replaced and 0 otherwise. CH is the prefix to represent the characteristic of chairperson. CEO is the prefix to represent the characteristic of CEO. Ten is the number of years staying in office. Age is the age. Edu is the number of schooling. DirIndR is the proportion of independent directors to board size. DirHold is the percentage of shares held by directors to total shares. BoardSize is the number of directors on board. Dual is a dummy coded 1 if the chairperson and CEO are the same person and 0 otherwise. CentSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by central government and 0 otherwise. LocalSOE is a dummy coded 1 if the firm is a state-owned enterprise controlled by province, city or country government and 0 otherwise. ROA is return on assets. DA is debt to asset ratio. LnAsset is log of total assets. FirmAge is the number of years the firm is established. * 0.05 significance level ** 0.01 significance level

48