Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Effort allocation and governance effect of multiple-board independent directors: evidence from reputation and geographic proximity

Effort allocation and governance effect of multiple-board independent directors: evidence from... China Journal of aCC ounting StudieS , 2016 Vol. 4, no . 3, 287–307 http://dx.doi.org/10.1080/21697213.2016.1222152 Eor ff t allocation and governance effect of multiple-board independent directors: evidence from reputation and geographic proximity* a b Yi Quan and Donghua Chen d epartment of a ccounting, School of a ccountancy, Zhongnan university of economics and law, Wuhan, People’s republic of China; d epartment of a ccounting, Business School, nanjing university, nanjing, People’s republic of China ABSTRACT KEYWORDS effort allocation; geographic Using a sample of Chinese A-share listed companies from 2002 to proximity; governance effect; 2013, we investigate how multiple-board independent directors multiple-board independent allocate their effort and the consequences. The results indicate that: directors; reputation (1) Multiple-board independent directors do have a preference in their effort allocation. Specifically, more effort is allocated to companies which have a higher reputation and closer distance. (2) When directors are far from the firms they work for, their bias among companies with different reputation is easier to generate. (3) The multiple-board independent directors are more likely to leave firms with relatively lower reputation and farther distance due to poor accounting performance, even though their tenure has not expired. (4) Different allocation of effort results in different consequences. Specifically, the more directors who view the target firm as having a relatively higher reputation, the lower the CEO’s extra compensation, and the higher the sensitivity between compensation and performance; the more directors who view the target firm as being the relatively closer, the lower the CEO’s extra compensation. Providing evidence that multiple-board independent directors have a preference in their effort allocation, this paper not only contributes to the study of independent director characteristics and governance performance theoretically, but also has some policy implications on appointing director in practice. 1. Introduction With the aim of protecting the overall interests of the company, especially the legal interests of minority shareholders, the system of independent directors creates power balance and supervision by adding independent directors to the board. As an important measure to improve corporate governance, the China Securities Regulatory Commission (CSRC) prom- ulgated the “Guidance on Establishing an Independent Director System in Listed Companies” in 2001, requiring that an independent director can serve no more than five companies so CONTACT Yi Quan quanyi88888@163.com *Paper accepted by Cong Wang. © 2016 a ccounting Society of China 288 Y. QUAn AnD D. ChEn as to ensure enough time and energy are contributed to each company. There are opposite conclusions on whether multiple-board independent directors can perform their duties effec - tively. Fama and Jensen (1983) believe that appointments to more than one company could signal the independent directors’ quality to the market, and the number of board seats could be a proxy for reputation (Fama & Jensen, 1983; Kaplan & Reishus, 1990; Wang, Zhao, & Wei, 2006; Ye, Zhu, & Lu, 2011). In contrast, Fich and Shivdasani (2006), Core, holthausen and Larcker (1999), Shivdasani and Yermack (1999) believe that enough time and energy guarantees ee ff c - tive supervision, so excessive board seats will reduce the efficiency and cause poor govern- ance performance. Obviously, both views assume that independent directors assign time and effort to each company equally. however, Masulis and Mobbs (2014) find that directors with multiple directorships distribute their effort unequally according to the directorship's relative prestige. non-uniform distribution of energy input also exists in other groups. For example, Agarwal, Ma, and Mullally (2016) find that when managers hold multiple funds simultaneously, the effort they put in each fund depends on the marginal revenue and cost. In 2002, China introduced the independent director system. Based on the data from GTA CSMAR, the proportion of independent directors with multiple directorships was 15.41% at that time, and increased to 19.70% in 2013, showing a year-on-year growth trend. Due to limited time and energy, how to allocate time and energy becomes an initial issue for every independent director. Focusing on multiple-board independent directors, we try to explore the following questions: how do independent directors with multiple directorships allocate their limited time and energy among each company and does the difference in time and eo ff rt have an impact on governance performance. Using a sample of Chinese A-share listed companies from 2002 to 2013, we find that multiple-board independent directors do have a preference in their effort allocation. When directors are far from the firms they work for, their bias is more obvious. The multiple-board independent directors are more likely to leave firms with relatively lower reputation and farther distance due to poor accounting perfor - mance, even though their tenure has not expired. Different effort allocation results in differ - ent consequences. Compared with existing literature, this paper makes the following contributions. First, it deepens the understanding of the behavioural characteristics of multiple-board independ- ent directors. Most of the existing literature investigates the behavioural characteristics and governance performance under the implicit assumption that independent directors allocate equal time and effort to each company (Core, holthausen, & Larcker, 1999; Fama & Jensen, 1983; Fich & Shivdasani, 2006; Kaplan & Reishus, 1990; Shivdasani & Yermack, 1999; Vafeas, 1999; Wang et al., 2006). While Masulis and Mobbs (2014) find that directors with multiple directorships allocate their effort unequally based on the reputation incentive from the view of benefits. This paper demonstrates not only the influence of reputation incentive on allo - cation of directors’ effort from the viewpoint of benefits, but also the influence of geographic distance from the viewpoint of time costs. Second, we provide new evidence on the study of the effect of independent directors’ governance. Knyazeva, Knyazeva, and Masulis (2013) find that the greater is the range for recruiting independent director candidates, the more independent are the directors, and the more positive are the effects on corporate governance, operating performance and firm value. They also prove that to employ excellent independent directors, the largest firms (top quartile) are prone to recruit nationwide. The authors attribute good governance perfor- mance to independent directors’ independence. however, the size of the director pools does ChInA JOURnAL OF ACCOUnTInG STUDIES 289 not represent the strength of independence. Six Degrees of Separation indicate that any two people in the world can be connected by no more than six people (Milgram, 1967). Considering the issue of relatively higher reputation, independent directors will devote more time and effort to large firms, so effort allocation differences may be the underlying cause of various performance levels. Third, we provide evidence on appointing independent directors in practice. On the one hand, the experience of multiple-board independent directors helps them perform duties; on the other hand, limited time and energy restricts their performance. Therefore, whether to hire multiple directorships or not is always a puzzle for both scholars and practitioners. This study finds that the role of directors with multiple directorships is not the same in dif- ferent firms and depends on the companies’ relative reputations and distance. Therefore, reputation and distance should be considered when choosing directors with multiple directorships. Lastly, this paper enriches the literature on geographic distance. Existing studies on geo- graphic distance measure distance by spatial distance (Malloy, 2005; O'Brien & Tan, 2015) or by the differences and similarities of areas (e.g. Tan, 2003). however, due to the different levels of transportation development, sometimes equal distance in space may require quite different time to travel. In this paper, we measure the distance between independent direc - tors and the company's office from three dimensions: spatial distance, the shortest time by train and train departure frequency, forming a complement to the measurement method of the existing literature. The remainder of the paper is organised as follows. Section 2 introduces the institutional background and literature review; Section 3 develops the hypotheses based on the theo- retical analysis; Section 4 presents the research design; Section 5 reports the empirical results; Section 6 concludes the paper. 2. Institutional background and literature review 2.1. Institutional background In order to improve the corporate governance structure and standardize the operation of listed companies, on August 16, 2001, the CSRC promulgated “Guidance on Establishing Independent Director System in Listed Companies” (hereafter, Guidance), requiring listed companies to establish the independent director system. The Guidance also requires that the number of companies that an independent director can serve is no more than five to ensure enough time and energy he can contribute to each company. Data shows that from 2002 to 2013, 42,864 independent directors have been hired by Chinese A-share listed com- panies, 35,058 (81.79%) of who have served in only one company. 7,806 (18.21%) independ- ent directors have served for two or more companies. The proportion of directors with multiple directorships is not high relatively; however, the number of seats they hold is up to 19,256, accounting for 35.45%. Large relation network enables them to play a decisive role in the capital market. Therefore, the study of their behaviour and governance charac- teristics is particularly urgent and important. The greatest criticism of multiple-board independent directors by scholars and practi- tioners is that limited time and energy result in low efficiency when directors serve for several boards simultaneously. And this issue is more obvious when the companies are in different 290 Y. QUAn AnD D. ChEn locations and industries. In China, independent directors carry out duties mainly by issuing independent opinions on major issues for listed companies. If the independent directors miss board meetings three times consecutively, they can be replaced by shareholder meet- ings after the board of directors’ admission. It is difficult to observe directly the supervision behaviour of independent directors, yet the time and effort taken to carry out duties con- stitute the most important part of costs ( Tan, Zheng, & Zhou, 2006). It is hard to imagine that an independent director who never attends meetings will invest much time and energy when carrying out duties. Therefore, using meeting attendance of independent directors to measure the effort is quite reasonable. This method is also widely used by scholars both at home and abroad, such as Adams and Ferreira (2008), Masulis and Mobbs (2014), Tan et al. (2006), and so on. 2.2. Literature review Scholars who agree with directors holding multiple directorships hold the view that more than one appointment could signal the quality of the independent directors to the market (Fama & Jensen, 1983), and the number of board seats can be a proxy for reputation (Fama & Jensen, 1983; Kaplan & Reishus, 1990; Wang et al., 2006; Ye et al., 2011). For example, Ferris, Jagannathan, and Pritchard (2003) find that good performance has a positive effect on the number of appointments received by a director. They find no evidence that multiple directors shirk their responsibilities because of being too busy. They also find no evidence that multiple directors are associated with securities fraud litigation. Tang, Shen, and Du (2010) argue that when independent directos hold fewer directorships or get higher compensation from the firm, they are less likely to say no to the managers of the firm. Chen and Xie (2011) and Xie and Chen (2012) believe that the more directorships an independent director holds, the more likely they are to be in the central position of the directors’ network, and directors with higher network centrality can help alleviating under- and over-investment. Wan, Deng, and Zheng (2014) also find that a larger relationship network helps independent directors hold more social capital, which improves the supervision motivation and ability, effectively sup - pressing irregularities of listed companies. Field, Lowry, and Mkrtchyan (2013) believe that the role of directors with multiple directorships is not the same in different firms. For IPO firms lacking experience in public markets, busy boards can improve firm value. Scholars who disagree with directors holding multiple directorships hold the view that enough time and energy is an important guarantee for independent directors to play mon- itoring role effectively. Thus excessive board seats will decrease efficiency and cause poor governance performance (Core et al., 1999; Fich & Shivdasani, 2006; Shivdasani & Yermack, 1999). For example, Beasley (1996) finds that when the directorships held by an independent director decline, the probability of fraud significantly reduces. Core et al. (1999) find that there is a significant positive correlation between CEO compensation and the proportion of independent directors who hold three or more board seats. Shivdasani and Yermack (1999) believe that agency cost is positively related to the proportion of independent directors who hold more than one board seat. Fich and Shivdasani (2006) find that firms with busy boards, where a majority of outside directors hold three or more directorships, are associated with weak corporate governance, lower market-to-book ratios, weaker profitability, and lower sensitivity of CEO turnover to firm performance. They also suggest that departure of busy outside directors generates positive abnormal returns, when directors become busy ChInA JOURnAL OF ACCOUnTInG STUDIES 291 as a result of acquiring an additional directorship, previous companies in which they hold board seats experience negative abnormal returns. Research by Wei, Xiao, n ick, and Zou (2007) on independent directors and firm's operating performance also supports limitations on multiple directorships of independent directors. hunton and Rose (2011) find that the more directorships the independent directors hold, the greater reputation loss they will suffer due to ineffectively performing their duties. Cashman, Gillan, and Jun (2012) also find that there is a significant negative correlation between firm performance and directors with multiple directorships. 3. Theoretical analysis and hypothesis development Managerial talent is a kind of scarce resource. The time invested to part-time firms will squeeze the time for full-time companies (Conyon & Read, 2006). Due to the limited time and energy, how directors with multiple directorships allocate their time and energy among each company has become the first question. Agarwal, Ma, and Mullally (2016) find that when fund managers are managing multiple funds simultaneously, the effort they put into each fund depends on the marginal management revenue and cost. For independent direc- tors, their effort invested in each company also has opportunity costs. As rational economic individuals, independent directors will consider the benefits and costs comprehensively to maximise their own interests when allocating energy. From the perspective of benefits, large-scale companies will provide independent directors with greater visibility and prestige (Adams & Ferreira, 2008; Shivdasani, 1993; Zhou, Tan, & Jian, 2008), higher compensation (Du & Zhang, 2004; Ryan & Wiggins, 2004; Sun & Zhu, 2005), and more opportunities to obtain other directorships (Fich, 2005; Yermack, 2004). hence directors will devote more time and effort to high reputation companies (Masulis & Mobbs, 2014). Prospective directors may be willing to undertake more distance costs to obtain a seat in a famous firm (Knyazeva et al., 2013). From the perspective of costs, the time and effort taken to carry out duties constitute the most part of costs, and the distance between directors and the company can reflect the cost to some extent (Tan et al., 2006). Besides attending the board meetings, independent directors also need to spend much time on financial reports and important issues (Tan et al., 2006). Far distance obviously brings inconvenience. For board meetings, more time is consumed to attend long-distance companies; for financial reports and impor - tant issues considerations, it is more convenient to transfer information when the independ- ent directors and the company are in the same area (Tan, 2003). As can be seen, higher reputation companies will increase benefits and geographic proximity will lower costs for independent directors. Based on the above analysis, we propose our first hypothesis: H1: The effort multiple-board independent directors invest in each company is significantly positively correlated with the relative prestige (h1a), and significantly negative correlated with the relative distance (h1b). Director commitment is also revealed in their willingness to remain on the board when a firm faces serious difficulties (Masulis & Mobbs, 2014). A firm in trouble requires directors to devote more time and energy, for example, more attendance at board meetings (Vafeas, 1999). From the perspective of reputation, if a troubled firm is relatively small, although the turnover will damage reputation, directors can use the time saved to serve other companies better. In contrast, if the troubled firm is a higher ranked firm, the director is more willing to work hard to help the firm out of trouble (Masulis & Mobbs, 2014). Zhou et al. (2008) also 292 Y. QUAn AnD D. ChEn find that the visibility and reputation impact of the directorship is the key consideration when independent directors resign. Because the cost of relinquishing small-scale companies is lower than large ones (e.g., Adams & Ferreira, 2008), we expect that the sensitivity of forced departure of independent directors to performance is higher in firms with relatively lower reputation. From the perspective of distance, because of the convenience of gathering infor- mation and communication, more board meetings of troubled companies nearby will not occupy independent directors with too much extra time and effort. In contrast, if the troubled company is far away, more board meetings will undoubtedly take independent directors more time and even disrupt their work plan. Tan et al. (2006) find that physical location can capture the realistic cost of independent directors, which is significantly related to inde - pendent directors’ turnover. Because the cost of relinquishing nearby companies is higher than that of long-distance ones, we expect that the sensitivity of forced departure of inde- pendent directors’ to performance is higher in firms with relatively longer distance. Based on the above analysis, we propose our second hypothesis: H2: The sensitivity of forced departure of independent directors to performance is higher in firms with relatively lower reputation and longer distance. Since the independent director system was set up in China, the validity of the system has been one of the issues scholars have been exploring. As researches go further and more detailed, scholars focus on personal characteristics of independent directors, such as edu- cation, working experience, expertise and relationship networks. however, the existing lit - erature still ignores the fact that, even the same directors can have different governance effects due to different firms they serve in. For example, from the perspective of growth stage, Field et al. (2013) find that, among IPO firms, the rich experience of busy boards con- tributes positively to firm value; however, these positive effects of busy boards extend to all but the most established firms. Benefits are lowest among Forbes 500 firms. From the per - spective of energy investment, Masulis and Mobbs (2014) find that directors with multiple directorships distribute their effort unequally according to the directorship's relative prestige, and the sensitivity of forced CEO departure to poor performance rises when a larger fraction of independent directors view the board as relatively more prestigious. In other words, more effort investment improves the governance performance. In China, issuing independent opinions on major issues in companies is an important way for independent directors to carry out their monitoring duties. The major issues include “nomination and appointment of directors”, “appointment and dismissal of senior executives”, “remuneration of directors, senior executives”. Because detailed reasons for the leaving of directors and senior executives are hard to obtain and their tenure is not regulated, it is difficult for us to distinguish whether the turnover is voluntary or forced. Taking the reality into account, it is more reasonable to examine the governance performance of independent directors from the perspective of CEO compensation. We expect that the greater proportion of independent directors who invest more effort, the lower the CEO’s extra compensation, and the higher the sensitivity between the compensation and performance. Based on the above analysis, we propose our third hypothesis: H3: The more independent directors who view the target firm as having a relatively higher reputation and closer distance, the lower the CEO’s extra compensation, and the higher the sensitivity between the compensation and performance. ChInA JOURnAL OF ACCOUnTInG STUDIES 293 4. Research design 4.1. Sample and data The sample in this study is Chinese A-share listed companies from 2002 to 2013. Since the information on board meeting attendance is disclosed from 2004, the sample period of h1 ranges from 2004 to 2013. The independent directors’ personal information and addresses are hand collected. Other data is collected from GTA CSMAR, with some supplemented manually. For the initial sample, we exclude: (1) financial and insurance firms; and (2) obser - vations with missing variables. Our final sample includes 49,188 firm-year observations. In order to avoid the effects of extreme values, all continuous variables are winsorized at both top and bottom 1% level. We use STATA software to process data in this paper. 4.2. Model settings and variable definitions We structure models (1-1) and (1-2) to test h1. If h1 is supported, b in models (1-1) and (1-2) should be significantly negative, and b should be significantly positive. Meeting12 = b + b ∗ Highest_Repu + b ∗ Lowest_Repu + b ∗ Control + e (1-1) 0 1 2 Meeting12 = b + b ∗ Nearest_Dist + b ∗ Farest_Dist + b ∗ Control + e (1-2) 0 1 2 We structure models (2-1) and (2-2) to test h2. If h2 is as expected, b4 in models (2-1) and (2-2) should be not significant or significantly positive, and b should be significantly negative. Probit(Turnover)= b + b ∗ Highest_Repu + b ∗ Lowest_Repu 0 1 2 + b ∗ ROA + b ∗ ROA ∗ Highest_Repu + b ∗ ROA ∗ Lowest_Repu 3 4 5 (2-1) + b ∗ Control + e Probit(Turnover)= b + b ∗ Nearest_Dist + b ∗ Farest_Dist 0 1 2 + b ∗ ROA + b ∗ ROA ∗ Nearest_Dist 3 4 (2-2) + b ∗ ROA ∗ Farest_Dist + b Control + e We structure models (3-1) and (3-2) to test h3. All explanatory variables are lagged for one year to alleviate the endogeneity issues. If h3 is supported, b1 in models (3-1) and (3-2) should be significantly negative, and b3 should be significantly positive. Extra_pay = b + b ∗ Highest_Rindir + b ∗ ROA t 0 1 t−1 2 t−1 (3-1) + b ∗ ROA ∗ Highest_Rindir + b ∗ Control + e 3 t−1 t−1 t−1 t−1 Extra_Pay = b + b ∗ Nearest_Rindir + b ∗ ROA t 0 1 t−1 2 t−1 (3-2) + b ∗ ROA ∗ Nearest_Rindir + b ∗ Control + e 3 t−1 t−1 t−1 t−1 294 Y. QUAn AnD D. ChEn The dependent variables are Meeting12, Turnover and Extra_Pay in models (1)-(3), respec- tively. Meeting12 represents the time and energy independent directors invest, which is defined as the proportion of board meetings that directors do not attend personally. The greater the value of Meeting12, the less time and energy they invest. We also further decom- pose Meeting12 into Meeting1 and Meeting2, which are defined as the proportion of attend- ance by agents and total absence respectively. Turnover is a dummy variable of forced turnover of independent directors, which is equal to one if independent directors leave and the term of office is shorter than three years, and 0 otherwise. Extra_Pay represents the excess compensation of CEO. In all the models, we add Year, Industry and Province dummy variables to control for year, industry and province fixed effects. Table 1 summarises our variables. Table 1. d efinition of variables. Variable Definition of variables Meeting12 Proportion of board meetings that directors do not attend personally Meeting1 Proportion of attendance by agents Meeting2 Proportion of total absence Turnover indicator variable that equals to one if independent directors leave and the term of office is shorter than three years, and 0 otherwise Extra_Pay Ceo ’s excess compensation. f ollowing Core et al. (1999) and Wu, lin, and Wang (2010), we use the regression residuals to measure that extra compensation Highest_ indicator variable that equals one if the directorship is the director's largest one measured by the firm Repu size, and 0 otherwise Lowest_Repu indicator variable that equals one if the directorship is the director's smallest one measured by the firm size, and 0 otherwise Farest_Dist indicator variable that equals one if the directorship is the director's farthest one measured by the distance between the firm and director, and 0 otherwise Nearest_Dist indicator variable that equals one if the directorship is the director's nearest one measured by the distance between the firm and director, and 0 otherwise Highest_ Proportion of independent directors who view the target firm as the highest reputation one. Highest_ Rindir Rindir=∑(Highest_Repu)/Indirector, Indirector indicates the total number of independent directors Nearest_ Proportion of independent directors who view the target firm as the nearest one, Nearest_Rindir= Rindir ∑(Nearest_Dist)/Indirector Meeting number of meetings directors required to attend in the year Indir_Pay natural logarithm of independent directors allowance Indir_Age natural logarithm of the independent directors age Indir_Tenure independent director’s tenure, calculated as the number of years between the current year and the first year the independent director assumes the duty in the firm minus the number of years she doesn’t assume the duty during this period Education independent director’s education, doctor 5, master 4, undergraduate 3, college 2, otherwise 1 Rindirector Proportion of independent directors BoardSize t otal number of board of directors Lnsize natural logarithm of the total assets at the fiscal year end ROA return on assets CEO_Edu Ceo ’s education: doctor 5, master 4, undergraduate 3, college 2, otherwise 1 Indir_Acc indicator variable that equals one if the director has accounting backgrounds, and 0 otherwise Indir_Law indicator variable that equals one if the director has legal backgrounds, and 0 otherwise Indir_Fin indicator variable that equals one if the director has financial backgrounds, and 0 otherwise Indir_PC indicator variable that equals one if the director has political backgrounds, and 0 otherwise Gender indicator variable that equals one if the director is female, and 0 otherwise CEO_Age natural logarithm of Ceo age Dual indicator variable that equals one if the chairman and the Ceo is the same person, and 0 otherwise Mshare t op management shareholding ratio Complex t he involved industry number Levevage r atio of total debt to total assets at the fiscal year end Growth growth rate of total assets SOE indicator variable that equals one if the firm is state-owned and 0 otherwise Zindex r atio of the largest shareholder to the second largest List_Age natural logarithm of listed years notes: a ccording to the definition, for independent directors who serve only in a company, the value of Highest_Repu, Lowest_Repu, Nearest_Dist, Farest_Dist are all equal to 1. ChInA JOURnAL OF ACCOUnTInG STUDIES 295 5. Empirical results and analysis 5.1. Descriptive statistics Panel A and B of Table 2 present descriptive statistics related to independent directors and firms respectively. During the sample period, directors need to attend eight board meetings on average a year, with a maximum of 72 times. Approximately 5.7% of the independent directors do not attend board meetings personally, with a maximum of 100%; the proportion of attendance by agents and absence are 4.9% and 0.8% respectively. This result indicates that directors do have a preference in their effort allocation. The turnover ratio of independ - ent directors is 15.8% on average, which is consistent with the findings of Li and Qin (2011). The average education level of independent directors is between undergraduate and master. while slightly higher than CEOs. 5.2. The basic regression analysis 5.2.1. H1a: Reputation and effort allocation Table 3 reports the results of h1a. The dependent variables in columns (1) to (3) are Meeting12, Meeting1 and Meeting2 respectively. The results in column (1) show that the coefficient on Highest_Repu is significantly negative at 1% level, the coefficient on Lowest_ Repu is not significant, and the difference between the two coefficients is significant at the 1% level (F value is 16.88, P value is 0.0000). This indicates that, for companies with relatively high (low) reputation, the proportion of board meetings that independent directors attend personally is higher (lower). Results in column (2) and (3) show that, Highest_Repu is signifi- cantly negatively correlated with Meeting1 at the 1% level, Lowest_Repu is positively but not significantly correlated with Meeting1, and the difference between the two coefficients is significant at the 1% level (F value is 16.04, P value is 0.0001). Highest_Repu is positively but not significantly correlated with Meeting2, and Lowest_Repu is significantly positively corre - lated with Meeting2 at the 10% level. however, the difference between the two coefficients is not significant (F value is 0.84, P value is 0.3603). This indicates that, for companies with relatively high reputation, the proportion of board meetings attending by agents is signifi- cantly lower; for companies with relatively low reputation, the proportion of absence is higher but not significant. The results in Table 3 overall indicate that for companies with relatively high (low) reputation, independent directors invest more (less) time and energy. h1a has been confirmed. The results of the control variables show that: Indir_Pay is significantly positively correlated with Meeting1 at 10% level, and significantly negatively correlated with Meeting2 at 1% level. This indicates that the amount of allowance can inspire independent directors to devote more time and effort to some extent, which is consistent with the findings of Adams and Ferreira (2008) and Masulis and Mobbs (2014). Meeting is significantly negatively correlated with Meeting12 and Meeting1 at 1% level, and positively correlated with Meeting2 at 1% level. This indicates that the greater the number of meetings directors require to participate, the on o ctober 19, 2013, the Central organization d epartment issued “Comments on further standardizing the part time (working) problem of party and government leaders in business”, require to clean up the illegal part time jobs (working) of Party and government leading cadres in business within a definite time. i n order to prevent the impact of this shock on the conclusions, in h2, we removed the sample in 2013. t he results are consistent when the sample in 2013 is retained. When calculating standard deviation, we cluster by independent directors, similarly hereinafter. 296 Y. QUAn AnD D. ChEn Table 2. Variable descriptive statistics. Panel A: director level N Mean Standard deviation Min. P25 Median P75 Max. Meeting12 41,739 0.06 0.12 0.00 0.00 0.00 0.09 1.00 Meeting1 41,739 0.05 0.11 0.00 0.00 0.00 0.06 1.00 Meeting2 41,776 0.01 0.05 0.00 0.00 0.00 0.00 1.00 Turnover 47,922 0.16 0.36 0.00 0.00 0.00 0.00 1.00 Highest_Repu 47,922 0.79 0.40 0.00 1.00 1.00 1.00 1.00 Lowest_Repu 47,922 0.79 0.40 0.00 1.00 1.00 1.00 1.00 Nearest_Dist 47,821 0.79 0.40 0.00 1.00 1.00 1.00 1.00 Farest_Dist 47,821 0.79 0.41 0.00 1.00 1.00 1.00 1.00 Meeting 41,776 8.23 4.05 1.00 6.00 8.00 10.00 72.00 Indir_Pay 41,776 10.72 0.58 8.99 10.31 10.82 11.00 12.21 Indir_Age 47,922 3.92 0.20 3.50 3.76 3.89 4.09 4.32 Indir_Tenure 47,922 0.96 0.64 0.00 0.69 1.10 1.39 2.08 Education 47,129 3.84 0.96 1.00 3.00 4.00 5.00 5.00 Indir_Acc 47,922 0.31 0.46 0.00 0.00 0.00 1.00 1.00 Indir_Law 47,922 0.15 0.35 0.00 0.00 0.00 0.00 1.00 Indir_Fin 47,922 0.11 0.31 0.00 0.00 0.00 0.00 1.00 Indir_PC 47,922 0.18 0.39 0.00 0.00 0.00 0.00 1.00 Gender 47,922 0.12 0.33 0.00 0.00 0.00 0.00 1.00 BoardSize 47,922 9.72 2.17 4.00 9.00 9.00 11.00 19.00 Rindirector 47,922 0.38 0.10 0.17 0.33 0.33 0.43 0.78 Lnsize 47,922 21.68 1.32 18.64 20.80 21.57 22.41 25.79 ROA 47,922 0.02 0.08 −0.43 0.01 0.03 0.06 0.22 Panel B: firm level Extra_Pay 10,828 0.01 0.66 −1.95 −0.37 0.03 0.43 1.47 Highest_Rindir 10,828 0.79 0.25 0.00 0.67 0.83 1.00 1.00 Nearest_Rindir 10,828 0.79 0.24 0.00 0.67 0.80 1.00 1.00 CEO_Education 10,828 3.41 0.81 1.00 3.00 3.00 4.00 5.00 CEO_Age 10,828 3.84 0.14 3.47 3.75 3.85 3.93 4.13 Duality 10,828 0.13 0.33 0.00 0.00 0.00 0.00 1.00 Mshare 10,828 0.01 0.04 0.00 0.00 0.00 0.00 0.34 Complex 10,828 3.91 1.80 1.00 2.58 3.94 5.00 9.00 Lnsize 10,828 21.69 1.28 18.54 20.84 21.60 22.43 25.51 Leverage 10,828 0.56 0.31 0.08 0.39 0.54 0.67 2.64 Growth 10,828 0.13 0.26 −0.48 −0.01 0.09 0.21 1.51 SOE 10,828 0.68 0.47 0.00 0.00 1.00 1.00 1.00 Zindex 10,828 22.29 45.68 1.02 2.03 6.00 20.88 327.50 List_Age 10,828 2.28 0.55 0.00 2.08 2.40 2.64 3.05 greater the proportion of absence, which supports the assertion about the limited energy of independent directors. Indir_Tenure is significantly positively correlated with Meeting12 and Meeting1 at 1% level, probably because the longer the tenure, the more familiar with the business, and the less necessary for independent directors to attend the meetings per- sonally. Indir_Acc, Indir_Law and Indir_Financial are significantly negatively correlated with all dependent variables at least at 5% level except Meeting2, which indicates that directors with accounting, law and finance backgrounds are more likely to attend meetings personally and less likely to attend by agents or absent. Indir_PC is significantly positively correlated with Meeting2 at 10% level, indicating that independent directors with politics background are more likely to absent from the meeting. BoardSize is significantly positively correlated with Meeting12 and Meeting1 at 1% level, which is consistent with the findings of Masulis and Mobbs (2014). Lnsize is positively correlated with Meeting12 and Meeting1 at 1% level and significantly negatively correlated with Meeting2 at 5% level, which is probably because, on one hand, the number of independent directors in large firms is larger, it is easier to attend the board meetings by agents; on the other hand, large firms can provide independent ChInA JOURnAL OF ACCOUnTInG STUDIES 297 Table 3. reputation and board meetings attendance. Variables (1) Meeting12 (2) Meeting1 (3) Meeting2 Highest_Repu −0.008*** −0.009*** 0.000 (−4.71) (−5.45) (0.44) Lowest_Repu 0.001 0.000 0.001* (0.78) (0.00) (1.81) Indir_Pay −0.001 0.002* −0.003*** (−0.86) (1.72) (−6.20) Meeting −0.001*** −0.001*** 0.000*** (−7.70) (−10.68) (6.33) Indir_Age −0.011** −0.009** 0.000 (−2.17) (−2.05) (−0.34) Indir_Tenure 0.006*** 0.005*** 0.000 (5.37) (5.21) (1.51) Education 0.009*** 0.007*** 0.002*** (9.49) (8.37) (5.81) Indir_Acc −0.023*** −0.020*** −0.002*** (−12.87) (−12.47) (−5.08) Indir_Law −0.010*** −0.008*** −0.002** (−4.08) (−3.59) (−2.40) Indir_Fin −0.005** −0.005** 0.000 (−1.99) (−2.28) (0.09) Indir_PC −0.000 −0.002 0.001* (−0.07) (−0.98) (1.80) Gender 0.004* 0.004* 0.000 (1.76) (1.76) (0.76) BoardSize 0.003*** 0.003*** 0.000* (9.14) (9.57) (1.80) Lnsize 0.002*** 0.002*** −0.000** (2.84) (3.60) (−2.05) ROA −0.042*** −0.018** −0.020*** (−4.73) (−2.35) (−6.19) Constant 0.048* −0.035 0.027*** (1.74) (−1.21) (3.57) Industry & Year & Province Control Control Control obs# 41,739 41,759 41,776 a dj-R 0.0770 0.0607 0.0418 F-value 48.05 37.46 25.60 notes: *, **, and *** to indicate 10, 5 and 1% level of significance respectively in all the regression tables. We bold all main variable of interest in order to highlight conclusions. directors with more benefits, thus the probability of absence in large firms is lower. ROA is significantly negatively correlated with Meeting12, Meeting1 and Meeting2 at least at the 5% level, indicating that directors devote more time and effort to firms in good condition. 5.2.2. H1b: Distance and effort allocation Table 4 reports the results of h1b. Panels A, B and C measure the distance between the independent directors and companies by “spatial distance”, “shortest travel time by train” and “train departure frequency” respectively. Specifically, if the spatial distance is nearest (farthest), or the travel time by train is shortest (longest), or the train departure frequency is fastest (slowest), then Nearest_Dist (Farest_Dist) equals to 1, and 0 otherwise. The depend- ent variables of Panel A, B and C in column (1) to (3) are Meeting12, Meeting1 and Meeting2 respectively. Results in Table 4 show that Nearest_Dist is significantly negatively correlated t he "spatial distance" has been calculated according to the spherical coordinates; the data of "shortest travel time by train" and "train departure frequency" come from the following website on april 30, 2015: http://www.12306.cn/mormhweb/. 298 Y. QUAn AnD D. ChEn Table 4. distance and board meetings attendance. Variables Panel A: spatial distance Panel B: shortest travel time by train Panel C: Train departure frequency (1) Meeting12 (2) Meeting1 (3) Meeting2 (1) Meeting12 (2) Meeting1 (3) Meeting2 (1) Meeting12 (2) Meeting1 (3) Meeting2 Nearest_Dist −0.008*** −0.009*** 0.000 −0.005*** −0.006*** 0.000 −0.006*** −0.006*** 0.000 (−4.73) (−5.54) (0.52) (−3.13) (−3.66) (0.37) (−3.39) (−3.98) (0.22) Farest_Dist −0.000 −0.001 0.001 −0.002 −0.002 0.001 −0.002 −0.003* 0.000 (−0.11) (−0.75) (1.38) (−0.90) (−1.59) (1.35) (−1.18) (−1.65) (0.95) Indir_Pay −0.001 0.002* −0.003*** −0.001 0.003* −0.003*** −0.001 0.003* −0.003*** (−0.821) (1.75) (−6.27) (−0.69) (1.92) (−6.18) (−0.70) (1.90) (−6.20) Meeting −0.001*** −0.001*** 0.000*** −0.001*** −0.002*** 0.000*** −0.001*** −0.002*** 0.000*** (−7.64) (−10.60) (6.37) (−7.73) (−10.70) (6.32) (−7.74) (−10.71) (6.32) Indir_Age −0.011** −0.009** −0.001 −0.011** −0.009* −0.001 −0.011** −0.009* −0.001 (−2.17) (−2.05) (−0.38) (−2.06) (−1.91) (−0.39) (−2.09) (−1.92) (−0.44) Indir_Tenure 0.006*** 0.005*** 0.000 0.006*** 0.005*** 0.000 0.006*** 0.005*** 0.000 (5.35) (5.19) (1.60) (5.35) (5.17) (1.58) (5.35) (5.18) (1.57) Education 0.009*** 0.007*** 0.002*** 0.009*** 0.007*** 0.001*** 0.009*** 0.007*** 0.001*** (9.46) (8.33) (5.85) (9.51) (8.42) (5.77) (9.43) (8.36) (5.69) Indir_Acc −0.023*** −0.020*** −0.002*** −0.023*** −0.020*** −0.002*** −0.023*** −0.020*** −0.003*** (−12.97) (−12.55) (−5.11) (−13.05) (−12.61) (−5.19) (−13.06) (−12.59) (−5.23) Indir_Law −0.010*** −0.008*** −0.001** −0.010*** −0.008*** −0.002** −0.010*** −0.008*** −0.002** (−4.07) (−3.58) (−2.38) (−4.13) (−3.62) (−2.43) (−4.14) (−3.63) (−2.45) Indir_Fin −0.005** −0.005** 0.000 −0.005** −0.005** 0.000 −0.005** −0.006** 0.000 (−2.01) (−2.30) (0.09) (−2.00) (−2.30) (0.14) (−2.01) (−2.30) (0.12) Indir_PC −0.000 −0.002 0.001* 0.000 −0.002 0.001* 0.000 −0.002 0.001* (−0.02) (−0.94) (1.85) (0.06) (−0.88) (1.91) (0.05) (−0.89) (1.90) Gender 0.004* 0.004* 0.000 0.004* 0.004* 0.000 0.004* 0.004* 0.001 (1.76) (1.76) (0.78) (1.77) (1.75) (0.80) (1.79) (1.77) (0.83) BoardSize 0.003*** 0.003*** 0.000* 0.003*** 0.003*** 0.000* 0.003*** 0.003*** 0.000* (8.94) (9.40) (1.71) (9.05) (9.47) (1.81) (9.05) (9.46) (1.83) Lnsize 0.001** 0.002*** −0.000** 0.001** 0.002*** −0.000** 0.001** 0.002*** −0.000** (2.06) (2.87) (−2.42) (2.03) (2.84) (−2.41) (2.02) (2.84) (−2.42) ROA −0.042*** −0.017** −0.020*** −0.041*** −0.017** −0.020*** −0.041*** −0.017** −0.020*** (−4.68) (−2.30) (−6.18) (−4.63) (−2.23) (−6.18) (−4.65) (−2.26) (−6.18) Constant −0.008 −0.046 0.047*** 0.229*** 0.0260 0.044*** 0.231*** 0.026 0.045*** (−0.28) (−0.00) (6.18) (3.07) (0.93) (6.28) (3.09) (0.94) (6.34) Industry & Year Control Control Control Control Control Control Control Control Control & Province obs# 41,661 41,681 41,697 41,448 41,468 41,484 41,448 41,468 41,484 a dj-R 0.0769 0.0607 0.0418 0.0771 0.0606 0.0421 0.0772 0.0607 0.0421 F-value 47.92 36.92 25.58 47.80 37.17 25.66 47.84 37.19 25.63 ChInA JOURnAL OF ACCOUnTInG STUDIES 299 with Meeting12 and Meeting1 at the 1% level under all the three measurements; Farest_Dist is negatively correlated with Meeting1 at the 10% level only under the measurement of “train departure frequency”; and the difference between the two coefficients is significant at least at the 10% level except in Panel B. This indicates that, for companies with relatively close distance, the proportion of board meetings that independent directors attend through agents is significantly lower, while the difference of absence is not significant. The results in Table 4 indicate that, for companies with relatively far distance, the proportion of independ- ent directors who are absent from board meetings is higher; for companies with relatively near distance, the proportion of board meetings independent directors attend personally (through agents) is higher (lower). h1b has been confirmed. 5.2.3. H2: Reputation (distance) and turnover performance sensitivity Table 5 reports the results of h2. All the dependent variables are Turnover. The results in column (1) and (2) show that Highest_Repu and Lowest_Repu are significantly positively correlated with Turnover, but both the coefficient and the significance level of Highest_Repu are smaller than those of Lowest_Repu, indicating that the probability of independent direc- tors leaving a relatively high reputation company is smaller than those leaving a relatively low reputation one. ROA is significantly negatively correlated with Turnover at least at the 5% level, indicating that the worse the accounting performance, the higher the probability that independent directors leave. The interaction term Highest_Repu*ROA is not significant and the interaction term Lowest_Repu*ROA is negatively significant at 10% level, indicating that the sensitivity of forced turnover of independent directors to performance is higher in firms with relatively low reputation. The results in column (3) and (4) show that Nearest_Dist and Farest_Dist are significantly positively correlated with Turnover, but both the coefficient and significance level of Farest_ Dist are larger than those of Nearest_Dist, indicating that the probability of independent directors turnover from a relatively far company is larger than from a relatively close one, which is consistent with the findings of Tan et al. (2006). ROA is significantly negatively cor - related with Turnover at the 1% level, which is consistent with column (1) and (2). however, the interaction terms in column (4) are not significant. The results in column (3) and (4) indicate that distance affects the probability of independent directors’ turnover, but does not affect the forced departure sensitivity to performance. These results are logically con- sistent with the findings in column (1) and (2) of Table 5. Knyazeva et al. (2013) believe that in order to employ excellent independent directors, the largest firms (top quartile) are prone to recruit directors nationally. The reputation of a large company is higher, while the sensi- tivity of forced departure of independent directors to performance is lower. Results in Table 5 indicate that the probability of independent directors’ turnover from a company with relatively high reputation (far distance) is smaller (larger) than from a company with a relatively low reputation (near distance). The sensitivity of forced department of inde- pendent directors to performance is higher in firms with relatively low reputation. h2 of this article has been confirmed. The results of control variables show that Indir_Tenure is signifi- cantly positively correlated with Turnover at the 1% level, probably because the Guidance however, the difference between the two coefficients is not significant. in column (3) and (4) of t able 5, we only report the regression results under the measurement of “spatial distance”, the regression results under the other two measurements are roughly unchanged, similarly hereinafter. 300 Y. QUAn AnD D. ChEn Table 5. reputation (distance) and turnover performance sensitivity. Variables (1) (2) (3) (4) Highest_Repu 0.043** 0.039* (1.95) (1.76) Lowest_Repu 0.060*** 0.074*** (2.70) (3.07) Highest_Repu*ROA 0.258 (1.05) Lowest_Repu*ROA −0.527* (−1.66) Nearest_Dist 0.042* 0.040* (1.92) (1.78) Farest_Dist 0.057*** 0.057*** (2.62) (2.55) Nearest_Dist *ROA 0.112 (0.43) Farest_Dist *ROA −0.041 (−0.15) ROA −1.060*** −0.775** −1.062*** −1.111*** (−10.52) (−2.38) (−10.54) (−3.74) Indir_Age 0.039 0.039 0.034 0.034 (0.79) (0.80) (0.70) (0.69) Indir_Tenure 0.645*** 0.646*** 0.645*** 0.645*** (34.03) (34.02) (33.94) (33.94) Rindirector 2.983*** 2.980*** 2.990*** 2.990*** (35.97) (35.94) (36.01) (36.01) Indir_Acc −0.044** −0.044** −0.042** −0.042** (−2.14) (−2.15) (−2.04) (−2.03) Indir_Law −0.013 −0.013 −0.010 −0.010 (−0.50) (−0.50) (−0.40) (−0.40) Indir_Fin 0.064** 0.064** 0.065** 0.065** (2.16) (2.16) (2.20) (2.20) Indir_PC −0.041* −0.042* −0.039 −0.039 (−1.70) (−1.72) (−1.61) (−1.60) Gender −0.003 −0.003 −0.004 −0.004 (−0.11) (−0.11) (−0.13) (−0.13) Lnsize −0.049*** −0.049*** −0.051*** −0.051*** (−6.03) (−6.02) (−6.56) (−6.56) Constant −2.220*** −2.229*** −2.163*** −2.160*** (−8.57) (−8.60) (−8.45) (−8.44) Industry & Year & Province Control Control Control Control obs# 47,922 47,922 47,821 47,821 Pseudo R 0.1239 0.1241 0.1241 0.1241 Wald chi 2642.65 2650.27 2636.24 2638.42 clearly stipulates that the term of independent directors should be no more than six years. Rindirector is significantly positively correlated with Turnover at 1% level, indicating that the larger the number of independent directors, the higher the probability of turnover. Lnsize is significantly negatively correlated with Turnover at the 1% level, which is consistent with the findings in column (1) and (2) of Table 5. 5.2.4. H3: Effort allocation and governance effect Table 6 reports the results of h3. When calculating the standard deviation of the coefficients, we cluster by industry. All dependent variables in Table 6 are Extra_Pay. The results in column (1) and (2) show that whether we are controlling ROA alone or controlling ROA and Highest_ Rindir*ROA simultaneously, Highest_Rindir is significantly negatively correlated with Extra_ Pay at the 1% level. Meanwhile, Highest_Rindir*ROA is significantly positively correlated with ChInA JOURnAL OF ACCOUnTInG STUDIES 301 Table 6. reputation (distance) and Ceo Pay performance sensitivity. Variables (1) (2) (3) (4) Highest_Rindir −0.155*** −0.174*** t−1 (−3.19) (−3.71) Highest_Rindir *ROA 0.873*** t−1 t−1 (3.39) Nearest_Rindir −0.101*** −0.097*** t−1 (−2.83) (−2.83) Nearest_Rindir *ROA −0.134 t−1 t−1 (−0.55) ROA 0.666*** 0.003 0.674*** 0.781*** t−1 (4.04) (0.01) (4.09) (4.56) CEO_Education 0.064*** 0.064*** 0.065*** 0.065*** t−1 (6.57) (6.56) (6.66) (6.67) CEO_Age 0.297*** 0.295*** 0.300*** 0.300*** t−1 (4.57) (4.55) (4.43) (4.43) Duality −0.036** −0.036** −0.037** −0.037** t−1 (−2.60) (−2.52) (−2.55) (−2.57) Mshare −0.266** −0.270** −0.263** −0.264** t−1 (−2.15) (−2.16) (−2.19) (−2.19) Complex 0.017** 0.017** 0.017** 0.017** t−1 (2.56) (2.55) (2.60) (2.60) Lnsize −0.021*** −0.021*** −0.032*** −0.032*** t−1 (−4.10) (−4.11) (−5.81) (−5.84) Leverage 0.025 0.027 0.021 0.021 t−1 (1.20) (1.24) (1.06) (1.06) Growth 0.034 0.033 0.034 0.034 t−1 (1.42) (1.38) (1.44) (1.45) SOE 0.010 0.011 0.014 0.014 t−1 (0.34) (0.38) (0.51) (0.50) Zindex −0.001*** −0.001*** −0.001*** −0.001*** t−1 (−3.01) (−3.02) (−2.99) (−2.98) List_Age 0.011 0.011 0.011 0.011 t−1 (0.57) (0.54) (0.56) (0.56) Constant −0.802* −0.764* −0.676* −0.679* t−1 (−2.02) (−1.90) (−1.89) (−1.89) Industry & Year & Province Control Control Control Control obs# 10828 10828 10828 10828 a dj-R 0.1073 0.1080 0.1055 0.1054 F-value 19.33 19.21 18.98 18.72 Extra_Pay at the 1% level. This indicates that the more independent directors who view the target firm as one with relatively higher reputation, the lower the CEO’s extra compensation, and the higher the sensitivity between the extra compensation and performance. The results in column (3) and (4) show that whether we are controlling ROA alone or controlling ROA and Nearest_Rindir*ROA simultaneously, Nearest_Rindir is significantly neg - atively correlated with Extra_Pay at 1% level, indicating that the more independent directors who view the target firm as a closer one, the lower the CEO’s extra compensation. however, Nearest_Rindir*ROA is not signic fi ant. The results in Table 6 indicate that the more independ- ent directors who view the target firm as one with relatively higher reputation, the lower the CEO’s extra compensation and the higher sensitivity between the extra compensation and performance; the more independent directors who view the target firm as a closer one, the lower the CEO’s extra compensation. h3 of this article has been confirmed. The results of the control variables show that the higher-educated and the older the CEO, the higher the extra compensation; for firms with the chairman and the CEO held by one person, 302 Y. QUAn AnD D. ChEn the higher the top management shareholding; the less complex the business, the larger the size, the higher the ownership concentration, the lower the CEO’s extra compensation. 5.3. Additional tests 5.3.1. Reputation and effort allocation grouped by distance In h1, we find that independent directors allocate more time and effort to companies with higher reputation and closer distance. however, we did not control the distance between the independent directors and the companies when exploring the relative reputation. Therefore, we are not sure whether the relationship between the reputation and the effort allocation is ae ff cted by distance. By distinguishing whether the distance between the inde - pendent directors and the companies is farther than the median in that year, Table 7 divides the full sample into “long distance group” and “short distance group” and re-examine h1a. The results in long distance group show that Highest_Repu is negatively significantly corre - lated with Meeting12 and Meeting1 at 1% level; Lowest_Repu is significantly positively Table 7. reputation and board meetings attendance grouped by distance. Long distance group Short distance group Variables (1) Meeting12 (2) Meeting1 (3) Meeting2 (4) Meeting12 (5) Meeting1 (6) Meeting2 Highest_Repu −0.012*** −0.013*** 0.000 −0.003 −0.003* 0.000 (−4.623) (−5.25) (0.37) (−1.37) (−1.85) (0.51) Lowest_Repu 0.002 −0.001 0.002** 0.001 0.001 0.000 (0.58) (−0.32) (2.26) (0.63) (0.50) (0.18) Indir_Pay −0.004* −0.001 −0.003*** −0.000 0.004** −0.003*** (−1.90) (−0.36) (−4.37) (−0.06) (2.24) (−4.60) Meeting −0.001*** −0.002*** 0.000*** −0.001*** −0.001*** 0.000*** (−6.22) (−9.04) (4.79) (−4.13) (−5.58) (4.10) Indir_Age −0.010 −0.008 −0.000 −0.005 −0.004 −0.000 (−1.31) (−1.19) (−0.20) (−0.78) (−0.71) (−0.09) Indir_Tenure 0.006*** 0.005*** 0.000 0.006*** 0.006*** 0.000 (3.77) (3.49) (1.06) (4.32) (4.46) (1.07) Education 0.009*** 0.007*** 0.002*** 0.009*** 0.007*** 0.001*** (6.21) (5.36) (4.08) (7.74) (6.92) (4.05) Indir_Accounting −0.024*** −0.021*** −0.002*** −0.018*** −0.015*** −0.002*** (−8.88) (−8.60) (−3.21) (−8.46) (−8.07) (−3.55) Indir_Law −0.010*** −0.008*** −0.001* −0.007** −0.005* −0.001 (−2.79) (−2.52) (−1.68) (−2.23) (−1.81) (−1.55) Indir_Financial −0.006 −0.006* 0.001 −0.007** −0.006** −0.001 (−1.48) (−1.87) (0.52) (−2.20) (−2.17) (−0.58) Indir_PC 0.004 0.001 0.002* −0.003 −0.004* 0.001 (1.14) (0.34) (1.90) (−1.25) (−1.76) (0.71) Gender 0.003 0.004 −0.000 0.003 0.002 0.001 (1.02) (1.38) (−0.48) (1.30) (0.92) (1.29) BoardSize 0.004*** 0.004*** −0.000 0.003*** 0.002*** 0.000*** (7.23) (8.14) (−0.37) (5.48) (5.07) (2.87) Lnsize 0.003*** 0.003*** 0.000 0.001 0.002** −0.001*** (2.51) (2.67) (0.10) (1.20) (2.35) (−3.48) ROA −0.060*** −0.025** −0.029*** −0.021* −0.011 −0.008* (−4.73) (−2.32) (−6.25) (−1.79) (−1.08) (−1.88) Constant 0.062 0.033 0.026*** 0.173** −0.088*** 0.039*** (1.39) (0.84) (2.60) (2.12) (−2.67) (3.82) Industry & Year & Control Control Control Control Control Control Province obs# 20,667 20,680 20,683 21,072 21,079 21,093 a dj-R 0.0945 0.0761 0.0510 0.0754 0.0595 0.0464 F-value 15.17 12.55 5.67 10.57 8.75 5.40 ChInA JOURnAL OF ACCOUnTInG STUDIES 303 correlated with Meeting2 at 5% level. The results in the short distance group show that only Highest_Repu is significantly negatively correlated with Meeting1 at 10% level. The difference between the two coefficients is significant at least at the 10% level. The results grouped by distance show that the bias is more obvious when the distance is far. For example, directors attend board meetings personally more often for companies with relatively high reputation compared with companies with relatively low reputation. While when the distance becomes close, the bias is relatively weak. The results in Table 7 indicate that directors with multiple directorships allocate their effort unequally based on not only the relative prestige, but also the relative distance. 5.3.2. Multiple-board independent directors and their opinion The Guidance stipulates that independent directors should voice one of the following opin- ions on major issues of listed companies: consent; reservations and the reasons; objections and the reasons; unable to express an opinion and its obstacle. Tang et al. (2010) compare the effects of directorships that independent directors hold on the opinions they issued. They find that the fewer directorships an independent director holds or the higher remu- neration he obtains, the lower the possibility that independent directors say "no". A more detailed problem is whether the possibility of multiple-board independent directors say "no" to different firms is different. Is the possibility also affected by the relative reputation and distance? Referring to the literature of Chen, Wang, and Duan (2015) and Ye et al. (2011), we report the regression results of independent directors’ opinions in Table 8. When calcu- lating the standard deviation of the coec ffi ients, we cluster by year to control the correlation among years. All dependent variables in Table 8 are Opinion, an indicator variable that equals one if the opinion is negative, and 0 otherwise. Results in column (1) show that coefficient on Highest_Repu is not significant, the coefficient on Lowest_Repu is positively significant at 5% level, and the difference between the two coefficients is significant at 10% level (F value is 3.24, P value is 0.0719). Results in column (2) show that coefficient on Farest_Dist is posi- tively significant at 10% level, the coefficient on Nearest_Dist is not significant, and the dif- ference between the two coefficients is significant at 10% level (F value is 2.74, P value is 0.0981). Results in Table 8 indicate that the independent directors voice significantly less (more) negative opinions to companies with relatively high reputation (far distance). It is possible because the corporate governance of companies with relatively high reputation is better; and the problems have been communicated before voting due to the low commu- nication costs of near-by companies. 5.4. Robustness test To further improve the reliability of our conclusions, we perform the following robustness test. First, we exclude the single-board independent directors sample (independent directors who serve only one company) and re-examine the hypotheses. Second, when defining Highest_Repu, Lowest_Repu, Nearest_Dist and Farest_Dist, we assigned 1 when the o f course, another possible explanation is that due to the higher cost of board seats lost, the independent directors would issue fewer negative opinions to companies with relatively high reputation (near distance). f or this problem, we will make a deeper analysis in the future. f or space considerations, the results are not reported in the text, but are available on application to the corresponding author. 304 Y. QUAn AnD D. ChEn Table 8. Multiple-board independent directors and their opinion. Variables (1) (2) Highest_Repu 0.134 (0.71) Lowest_Repu 0.643** (2.34) Farest_Dist 0.420* (1.80) Nearest_Dist 0.232 (0.90) Meeting 0.032 0.034 (0.89) (0.96) Boardlock 0.263* 0.253 (1.78) (1.36) Indir_Pay 0.016 0.027 (0.07) (0.11) Indir_Age −0.742 −0.710 (−1.32) (−1.29) Indir_Tenure 0.134 0.128 (0.75) (0.72) Education 0.106*** 0.108*** (2.60) (2.64) Indir_Acc 0.064 0.068 (0.62) (0.68) Indir_Law 0.141 0.151 (0.46) (0.50) Indir_Fin 0.063 0.067 (0.37) (0.39) Indir_PC 0.260 0.264 (1.30) (1.31) Gender −0.069 −0.072 (−0.47) (−0.49) Audit −1.003*** −0.986*** (−3.16) (−3.12) Duality 0.010 0.016 (0.04) (0.07) Mshare 2.324* 2.381* (1.76) (1.81) Instihold 0.956*** 0.964*** (2.95) (3.034) Board 0.700 0.693 (1.50) (1.47) Lnsize −0.110 −0.153 (−0.99) (−1.45) ROA −2.754** −2.762** (−2.45) (−2.44) Growth −0.312 −0.330 (−1.19) (−1.234) SOE −0.027 −0.018 (−0.11) (−0.07) Zindex −0.008** −0.008** (−2.45) (−2.38) List_Age −0.255 −0.265 (−0.89) (−0.91) Constant −3.786 −2.950 (−0.94) (−0.71) Industry & Year Control Control obs# 172,691 172,330 Pseudo R 0.1492 0.1480 Wald chi 1093.54 1083.02 ChInA JOURnAL OF ACCOUnTInG STUDIES 305 directorship is the director's largest/smallest (nearest/farest) one measured by the size (dis- tance) of the firm respectively. Referring to the method of Masulis and Mobbs (2014), we re-examine our primary findings using the following alternative measurements: If the direc - torship is 10% larger (smaller) than the director's smallest (largest) one measured by firm size, Highest_Repu (Lowest_Repu) equals 1, and 0 otherwise. If the directorship is 10% nearer (farther) than the director's farthest (nearest) one measured by firm distance, Nearest_Dist (Farest_Dist) equals 1, and 0 otherwise. Third, the Guidance clearly stipulates that the term of independent directors should not exceed six years. Considering the institutional back- ground, we re-define Turnover in accordance with whether the term of office is shorter than six years and re-examine h2. Fourth, we exclude the retirement sample of independent directors. Fifth, we use market value, and number of employees as our reputation measure- ments. Sixth, to avoid the endogeneity problem that corporate governance of large com- panies is better, we match each firm in which all the independent directors view the target firm as relatively higher reputation with a control firm in the same year and industry that is the most similar in firm size and has at least an independent director who does not view the target firm as the relatively higher reputation, and re-examine h3. The results of the above tests remain unchanged, so our conclusions still hold. 6. Conclusion Using a sample of Chinese A-share listed companies from 2002 to 2013, we investigate how multiple-board independent directors allocate their effort among different firms and the consequences. The results indicate that: (1) Multiple-board independent directors do have preference in their effort allocation. Specifically, more effort will be allocated to companies which have higher reputation, closer geographic distance and more convenient transportation. (2) When directors are far from the firms they work for, their bias among companies with different reputation is easier to generate. (3) The multiple-board independent directors are more likely to leave firms with relatively lower reputation and farther distance due to poor accounting performance, even though their tenure has not expired. (4) Different effort allocation results in different consequences. Specifically, the more independent directors who view the target firm as having a relatively higher reputation, the lower the CEO’s extra compensation, and the higher the sensitivity between compensation and performance; the more independent directors who view the target firm as being relatively closer, the lower the CEO’s extra compensation. Providing evidence that multiple-board independent directors have preference in their effort allocation, this paper not only contributes to the study of independent director characteristics and governance performance theoretically, but also has some policy implications on appointing independent director in practice. however, some deficiencies exist in this study. For example, besides attending board meetings, independent directors also need to spend time on financial reports and important issues (Tan et al., 2006). We measure the effort independent directors invest to companies only from the board meeting attendance, which is too simple. Also, different full-time work of independent directors will cause different schemes of time allocation, and different business matching degree between full-time and part-time work also will affect the supervision efficiency. A deeper analysis by refining identities, industries and professional background of independent directors will be explored in the future. 306 Y. QUAn AnD D. ChEn Acknowledgements and Funding This work was supported by national natural Science Foundation of China [71602191] and [71372032]. We appreciate the helpful comments from Professor Yu Xin and the anonymous reviewers. Disclosure statement no potential conflict of interest was reported by the authors. References Adams, R., & Ferreira, D. (2008). Do directors perform for pay? Journal of Financial Economics, 46, 154–171. Agarwal, V., Ma, L., & Mullally, K. (2016). Managerial multitasking in the mutual fund industry. Available at SSRn: http://ssrn.com/abstract=1910367 or http://dx.doi.org/10.2139/ssrn.1910367 Beasley, M. S. (1996). An empirical analysis of the relation between the board of director composition and financial statement fraud. The Accounting Review, 71, 443–465. Cashman, G. D., Gillan, S. L., & Jun, C. (2012). Going overboard? On busy directors and firm value. Journal of Banking & Finance, 36, 3248–3259. Chen, R., Wang, Z., & Duan, C. Q. (2015). Study on “reverse elimination” of independent directors: An empirical evidence from independent advices. China Industrial Economy, 8, 145–160 (In Chinese). Chen, Y. S., & Xie, D. R. (2011). network location, independent directors governance and investment efficiency. Management World, 7, 113–127 (In Chinese). Conyon, M. J., & Read, L. E. (2006). A model of the supply of executives for outside directorships. Journal of Corporate Finance, 12, 645–659. Core, J. E., holthausen, R. W., & Larcker, D. F. (1999). Corporate governance, chief executive officer compensation, and firm performance. Journal of Financial Economics, 51, 371–406. Du, S. L., & Zhang, J. (2004). An empirical study of independent directors’ compensation. Accounting Research, 9, 82–88 (In Chinese). Fama, E., & Jensen, M. (1983). The separation of ownership and control. Journal of Law and Economics, 26, 301–325. Ferris, S. P., Jagannathan, M., & Pritchard, A. C. (2003). Too busy to mind the business? Monitoring by directors with multiple board appointments. Journal of Finance, 58, 1087–1112. Fich, E. M. (2005). Are some outside directors better than others? Evidence from director appointments by Fortune 1000 firms. The Journal of Business, 78, 1943–1972. Fich, E. M., & Shivdasani, A. (2006). Are busy boards effective monitors? Journal of Finance, 61, 689–724. Field, L., Lowry, M., & Mkrtchyan, A. (2013). Are busy boards detrimental? Journal of Financial Economics., 109, 63–82. hunton, J. E., & Rose, J. M. (2011). Retracted: Effects of anonymous whistle-blowing and perceived reputation threats on investigations of whistle-blowing allegations by audit committee members. Journal of Management Studies, 48, 75–98. Kaplan, S. n., & Reishus, D. (1990). Outside directorships and corporate performance. Journal of Financial Economics, 27, 389–410. Knyazeva, A., Knyazeva, D., & Masulis, R. (2013). The supply of corporate directors and board independence. Review of Financial Studies, 26, 1561–1605. Li, Y., & Qin, Y. h. (2011). Media coverage, reputation mechanism and resignation of independent directors. Finance & Trade Economics, 3, 36–41 (In Chinese). Malloy, C. J. (2005). The geography of equity analysis. Journal of Finance, 60, 719–755. Masulis, R. W., & Mobbs, S. (2014). Independent director incentives: Where do talented directors spend their limited time and energy? Journal of Financial Economics, 111, 406–429. Milgram, S. (1967). The small world problem. Psychology Today, 2, 185–195. O'Brien, P. C., & Tan, h. (2015). Geographic proximity and analyst coverage decisions: Evidence from IPOs. Journal of Accounting and Economics, 59, 41–59. ChInA JOURnAL OF ACCOUnTInG STUDIES 307 Ryan, h., & Wiggins, R. (2004). Who is in whose pocket? Director compensation, board independence, and barriers to effective monitoring. Journal of Financial Economics, 73, 497–524. Shivdasani, A. (1993). Board composition, ownership structure, and hostile takeovers. Journal of Accounting and Economics, 16, 167–198. Shivdasani, A., & Yermack, D. (1999). CEO involvement in the selection of new board members: An empirical analysis. Journal of Finance, 54, 1829–1853. Sun, Z. R., & Zhu, X. M. (2005). The theory study and actuality analysis on the independent director’s compensation system in Chinese public corporations. Nankai Business Review, 8, 21–29 (In Chinese). Tan, J. S. (2003). Independent director and corporate governance: Based on Chinese listed companies. Beijing: China Financial and Economic Publishing house (In Chinese). Tan, J. S. (2003). Independent director and corporate governance: Based on Chinese listed companies. Beijing: China Financial and Economic Publishing house (In Chinese). Tang, X. S., Shen, h., & Du, J. (2010). The supervision motivation of independent directors: Empirical evidence based on the independent advice. Management World, 22, 138–149 (In Chinese). Vafeas, n. (1999). Board meeting frequency and firm performance. Journal of Financial Economics, 53, 113–142. Wan, L. Y., Deng, L., & Zheng, X. L. (2012). network location, independent directors governance and violations: Based on partially observable Bivariate Probit model. Systems Engineering-Theory & Practice, 34, 3091–3102 (In Chinese). Wang, Y. T., Zhao, Z. Y., & Wei, X. Y. (2006). Does independence of the board affect firm performance? Economic Research Journal, 5, 62–73 (In Chinese). Wei, G., Xiao, Z. Z., n ick, T., & Zou, h. (2007). Background of independent directors and corporate performance. Economic Research Journal, 3, 92–105 (In Chinese). Wu, L. S., Lin, J. Y., & Wang, Y. P. (2010). Salary external fairness, equity nature and firm performance. Management World, 3, 117–126. Xie, D. R., & Chen, Y. S. (2012). Board network: Definition, characteristics and measurement. Accounting Research, 3, 44–51 (In Chinese). Ye, K. T., Zhu, J. G., & Lu, Z. F. (2011). The independence of independent directors: Evidence from board voting behavior. Economic Research Journal, 1, 126–139 (In Chinese). Yermack, D. (2004). Remuneration, retention, and reputation incentives for outside directors. The Journal of Finance, 59, 2281–2308. Zhou, F., Tan, J. S., & Jian, Y. Y. (2008). Reputation incentive or economic incentive: An empirical study on job-hopping behaviors of independent directors. China Accounting Review, 2, 177–192 (In Chinese). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png China Journal of Accounting Studies Taylor & Francis

Effort allocation and governance effect of multiple-board independent directors: evidence from reputation and geographic proximity

China Journal of Accounting Studies , Volume 4 (3): 21 – Jul 2, 2016

Loading next page...
 
/lp/taylor-francis/effort-allocation-and-governance-effect-of-multiple-board-independent-zY1yHbtPcv

References (41)

Publisher
Taylor & Francis
Copyright
© 2016 Accounting Society of China
ISSN
2169-7221
eISSN
2169-7213
DOI
10.1080/21697213.2016.1222152
Publisher site
See Article on Publisher Site

Abstract

China Journal of aCC ounting StudieS , 2016 Vol. 4, no . 3, 287–307 http://dx.doi.org/10.1080/21697213.2016.1222152 Eor ff t allocation and governance effect of multiple-board independent directors: evidence from reputation and geographic proximity* a b Yi Quan and Donghua Chen d epartment of a ccounting, School of a ccountancy, Zhongnan university of economics and law, Wuhan, People’s republic of China; d epartment of a ccounting, Business School, nanjing university, nanjing, People’s republic of China ABSTRACT KEYWORDS effort allocation; geographic Using a sample of Chinese A-share listed companies from 2002 to proximity; governance effect; 2013, we investigate how multiple-board independent directors multiple-board independent allocate their effort and the consequences. The results indicate that: directors; reputation (1) Multiple-board independent directors do have a preference in their effort allocation. Specifically, more effort is allocated to companies which have a higher reputation and closer distance. (2) When directors are far from the firms they work for, their bias among companies with different reputation is easier to generate. (3) The multiple-board independent directors are more likely to leave firms with relatively lower reputation and farther distance due to poor accounting performance, even though their tenure has not expired. (4) Different allocation of effort results in different consequences. Specifically, the more directors who view the target firm as having a relatively higher reputation, the lower the CEO’s extra compensation, and the higher the sensitivity between compensation and performance; the more directors who view the target firm as being the relatively closer, the lower the CEO’s extra compensation. Providing evidence that multiple-board independent directors have a preference in their effort allocation, this paper not only contributes to the study of independent director characteristics and governance performance theoretically, but also has some policy implications on appointing director in practice. 1. Introduction With the aim of protecting the overall interests of the company, especially the legal interests of minority shareholders, the system of independent directors creates power balance and supervision by adding independent directors to the board. As an important measure to improve corporate governance, the China Securities Regulatory Commission (CSRC) prom- ulgated the “Guidance on Establishing an Independent Director System in Listed Companies” in 2001, requiring that an independent director can serve no more than five companies so CONTACT Yi Quan quanyi88888@163.com *Paper accepted by Cong Wang. © 2016 a ccounting Society of China 288 Y. QUAn AnD D. ChEn as to ensure enough time and energy are contributed to each company. There are opposite conclusions on whether multiple-board independent directors can perform their duties effec - tively. Fama and Jensen (1983) believe that appointments to more than one company could signal the independent directors’ quality to the market, and the number of board seats could be a proxy for reputation (Fama & Jensen, 1983; Kaplan & Reishus, 1990; Wang, Zhao, & Wei, 2006; Ye, Zhu, & Lu, 2011). In contrast, Fich and Shivdasani (2006), Core, holthausen and Larcker (1999), Shivdasani and Yermack (1999) believe that enough time and energy guarantees ee ff c - tive supervision, so excessive board seats will reduce the efficiency and cause poor govern- ance performance. Obviously, both views assume that independent directors assign time and effort to each company equally. however, Masulis and Mobbs (2014) find that directors with multiple directorships distribute their effort unequally according to the directorship's relative prestige. non-uniform distribution of energy input also exists in other groups. For example, Agarwal, Ma, and Mullally (2016) find that when managers hold multiple funds simultaneously, the effort they put in each fund depends on the marginal revenue and cost. In 2002, China introduced the independent director system. Based on the data from GTA CSMAR, the proportion of independent directors with multiple directorships was 15.41% at that time, and increased to 19.70% in 2013, showing a year-on-year growth trend. Due to limited time and energy, how to allocate time and energy becomes an initial issue for every independent director. Focusing on multiple-board independent directors, we try to explore the following questions: how do independent directors with multiple directorships allocate their limited time and energy among each company and does the difference in time and eo ff rt have an impact on governance performance. Using a sample of Chinese A-share listed companies from 2002 to 2013, we find that multiple-board independent directors do have a preference in their effort allocation. When directors are far from the firms they work for, their bias is more obvious. The multiple-board independent directors are more likely to leave firms with relatively lower reputation and farther distance due to poor accounting perfor - mance, even though their tenure has not expired. Different effort allocation results in differ - ent consequences. Compared with existing literature, this paper makes the following contributions. First, it deepens the understanding of the behavioural characteristics of multiple-board independ- ent directors. Most of the existing literature investigates the behavioural characteristics and governance performance under the implicit assumption that independent directors allocate equal time and effort to each company (Core, holthausen, & Larcker, 1999; Fama & Jensen, 1983; Fich & Shivdasani, 2006; Kaplan & Reishus, 1990; Shivdasani & Yermack, 1999; Vafeas, 1999; Wang et al., 2006). While Masulis and Mobbs (2014) find that directors with multiple directorships allocate their effort unequally based on the reputation incentive from the view of benefits. This paper demonstrates not only the influence of reputation incentive on allo - cation of directors’ effort from the viewpoint of benefits, but also the influence of geographic distance from the viewpoint of time costs. Second, we provide new evidence on the study of the effect of independent directors’ governance. Knyazeva, Knyazeva, and Masulis (2013) find that the greater is the range for recruiting independent director candidates, the more independent are the directors, and the more positive are the effects on corporate governance, operating performance and firm value. They also prove that to employ excellent independent directors, the largest firms (top quartile) are prone to recruit nationwide. The authors attribute good governance perfor- mance to independent directors’ independence. however, the size of the director pools does ChInA JOURnAL OF ACCOUnTInG STUDIES 289 not represent the strength of independence. Six Degrees of Separation indicate that any two people in the world can be connected by no more than six people (Milgram, 1967). Considering the issue of relatively higher reputation, independent directors will devote more time and effort to large firms, so effort allocation differences may be the underlying cause of various performance levels. Third, we provide evidence on appointing independent directors in practice. On the one hand, the experience of multiple-board independent directors helps them perform duties; on the other hand, limited time and energy restricts their performance. Therefore, whether to hire multiple directorships or not is always a puzzle for both scholars and practitioners. This study finds that the role of directors with multiple directorships is not the same in dif- ferent firms and depends on the companies’ relative reputations and distance. Therefore, reputation and distance should be considered when choosing directors with multiple directorships. Lastly, this paper enriches the literature on geographic distance. Existing studies on geo- graphic distance measure distance by spatial distance (Malloy, 2005; O'Brien & Tan, 2015) or by the differences and similarities of areas (e.g. Tan, 2003). however, due to the different levels of transportation development, sometimes equal distance in space may require quite different time to travel. In this paper, we measure the distance between independent direc - tors and the company's office from three dimensions: spatial distance, the shortest time by train and train departure frequency, forming a complement to the measurement method of the existing literature. The remainder of the paper is organised as follows. Section 2 introduces the institutional background and literature review; Section 3 develops the hypotheses based on the theo- retical analysis; Section 4 presents the research design; Section 5 reports the empirical results; Section 6 concludes the paper. 2. Institutional background and literature review 2.1. Institutional background In order to improve the corporate governance structure and standardize the operation of listed companies, on August 16, 2001, the CSRC promulgated “Guidance on Establishing Independent Director System in Listed Companies” (hereafter, Guidance), requiring listed companies to establish the independent director system. The Guidance also requires that the number of companies that an independent director can serve is no more than five to ensure enough time and energy he can contribute to each company. Data shows that from 2002 to 2013, 42,864 independent directors have been hired by Chinese A-share listed com- panies, 35,058 (81.79%) of who have served in only one company. 7,806 (18.21%) independ- ent directors have served for two or more companies. The proportion of directors with multiple directorships is not high relatively; however, the number of seats they hold is up to 19,256, accounting for 35.45%. Large relation network enables them to play a decisive role in the capital market. Therefore, the study of their behaviour and governance charac- teristics is particularly urgent and important. The greatest criticism of multiple-board independent directors by scholars and practi- tioners is that limited time and energy result in low efficiency when directors serve for several boards simultaneously. And this issue is more obvious when the companies are in different 290 Y. QUAn AnD D. ChEn locations and industries. In China, independent directors carry out duties mainly by issuing independent opinions on major issues for listed companies. If the independent directors miss board meetings three times consecutively, they can be replaced by shareholder meet- ings after the board of directors’ admission. It is difficult to observe directly the supervision behaviour of independent directors, yet the time and effort taken to carry out duties con- stitute the most important part of costs ( Tan, Zheng, & Zhou, 2006). It is hard to imagine that an independent director who never attends meetings will invest much time and energy when carrying out duties. Therefore, using meeting attendance of independent directors to measure the effort is quite reasonable. This method is also widely used by scholars both at home and abroad, such as Adams and Ferreira (2008), Masulis and Mobbs (2014), Tan et al. (2006), and so on. 2.2. Literature review Scholars who agree with directors holding multiple directorships hold the view that more than one appointment could signal the quality of the independent directors to the market (Fama & Jensen, 1983), and the number of board seats can be a proxy for reputation (Fama & Jensen, 1983; Kaplan & Reishus, 1990; Wang et al., 2006; Ye et al., 2011). For example, Ferris, Jagannathan, and Pritchard (2003) find that good performance has a positive effect on the number of appointments received by a director. They find no evidence that multiple directors shirk their responsibilities because of being too busy. They also find no evidence that multiple directors are associated with securities fraud litigation. Tang, Shen, and Du (2010) argue that when independent directos hold fewer directorships or get higher compensation from the firm, they are less likely to say no to the managers of the firm. Chen and Xie (2011) and Xie and Chen (2012) believe that the more directorships an independent director holds, the more likely they are to be in the central position of the directors’ network, and directors with higher network centrality can help alleviating under- and over-investment. Wan, Deng, and Zheng (2014) also find that a larger relationship network helps independent directors hold more social capital, which improves the supervision motivation and ability, effectively sup - pressing irregularities of listed companies. Field, Lowry, and Mkrtchyan (2013) believe that the role of directors with multiple directorships is not the same in different firms. For IPO firms lacking experience in public markets, busy boards can improve firm value. Scholars who disagree with directors holding multiple directorships hold the view that enough time and energy is an important guarantee for independent directors to play mon- itoring role effectively. Thus excessive board seats will decrease efficiency and cause poor governance performance (Core et al., 1999; Fich & Shivdasani, 2006; Shivdasani & Yermack, 1999). For example, Beasley (1996) finds that when the directorships held by an independent director decline, the probability of fraud significantly reduces. Core et al. (1999) find that there is a significant positive correlation between CEO compensation and the proportion of independent directors who hold three or more board seats. Shivdasani and Yermack (1999) believe that agency cost is positively related to the proportion of independent directors who hold more than one board seat. Fich and Shivdasani (2006) find that firms with busy boards, where a majority of outside directors hold three or more directorships, are associated with weak corporate governance, lower market-to-book ratios, weaker profitability, and lower sensitivity of CEO turnover to firm performance. They also suggest that departure of busy outside directors generates positive abnormal returns, when directors become busy ChInA JOURnAL OF ACCOUnTInG STUDIES 291 as a result of acquiring an additional directorship, previous companies in which they hold board seats experience negative abnormal returns. Research by Wei, Xiao, n ick, and Zou (2007) on independent directors and firm's operating performance also supports limitations on multiple directorships of independent directors. hunton and Rose (2011) find that the more directorships the independent directors hold, the greater reputation loss they will suffer due to ineffectively performing their duties. Cashman, Gillan, and Jun (2012) also find that there is a significant negative correlation between firm performance and directors with multiple directorships. 3. Theoretical analysis and hypothesis development Managerial talent is a kind of scarce resource. The time invested to part-time firms will squeeze the time for full-time companies (Conyon & Read, 2006). Due to the limited time and energy, how directors with multiple directorships allocate their time and energy among each company has become the first question. Agarwal, Ma, and Mullally (2016) find that when fund managers are managing multiple funds simultaneously, the effort they put into each fund depends on the marginal management revenue and cost. For independent direc- tors, their effort invested in each company also has opportunity costs. As rational economic individuals, independent directors will consider the benefits and costs comprehensively to maximise their own interests when allocating energy. From the perspective of benefits, large-scale companies will provide independent directors with greater visibility and prestige (Adams & Ferreira, 2008; Shivdasani, 1993; Zhou, Tan, & Jian, 2008), higher compensation (Du & Zhang, 2004; Ryan & Wiggins, 2004; Sun & Zhu, 2005), and more opportunities to obtain other directorships (Fich, 2005; Yermack, 2004). hence directors will devote more time and effort to high reputation companies (Masulis & Mobbs, 2014). Prospective directors may be willing to undertake more distance costs to obtain a seat in a famous firm (Knyazeva et al., 2013). From the perspective of costs, the time and effort taken to carry out duties constitute the most part of costs, and the distance between directors and the company can reflect the cost to some extent (Tan et al., 2006). Besides attending the board meetings, independent directors also need to spend much time on financial reports and important issues (Tan et al., 2006). Far distance obviously brings inconvenience. For board meetings, more time is consumed to attend long-distance companies; for financial reports and impor - tant issues considerations, it is more convenient to transfer information when the independ- ent directors and the company are in the same area (Tan, 2003). As can be seen, higher reputation companies will increase benefits and geographic proximity will lower costs for independent directors. Based on the above analysis, we propose our first hypothesis: H1: The effort multiple-board independent directors invest in each company is significantly positively correlated with the relative prestige (h1a), and significantly negative correlated with the relative distance (h1b). Director commitment is also revealed in their willingness to remain on the board when a firm faces serious difficulties (Masulis & Mobbs, 2014). A firm in trouble requires directors to devote more time and energy, for example, more attendance at board meetings (Vafeas, 1999). From the perspective of reputation, if a troubled firm is relatively small, although the turnover will damage reputation, directors can use the time saved to serve other companies better. In contrast, if the troubled firm is a higher ranked firm, the director is more willing to work hard to help the firm out of trouble (Masulis & Mobbs, 2014). Zhou et al. (2008) also 292 Y. QUAn AnD D. ChEn find that the visibility and reputation impact of the directorship is the key consideration when independent directors resign. Because the cost of relinquishing small-scale companies is lower than large ones (e.g., Adams & Ferreira, 2008), we expect that the sensitivity of forced departure of independent directors to performance is higher in firms with relatively lower reputation. From the perspective of distance, because of the convenience of gathering infor- mation and communication, more board meetings of troubled companies nearby will not occupy independent directors with too much extra time and effort. In contrast, if the troubled company is far away, more board meetings will undoubtedly take independent directors more time and even disrupt their work plan. Tan et al. (2006) find that physical location can capture the realistic cost of independent directors, which is significantly related to inde - pendent directors’ turnover. Because the cost of relinquishing nearby companies is higher than that of long-distance ones, we expect that the sensitivity of forced departure of inde- pendent directors’ to performance is higher in firms with relatively longer distance. Based on the above analysis, we propose our second hypothesis: H2: The sensitivity of forced departure of independent directors to performance is higher in firms with relatively lower reputation and longer distance. Since the independent director system was set up in China, the validity of the system has been one of the issues scholars have been exploring. As researches go further and more detailed, scholars focus on personal characteristics of independent directors, such as edu- cation, working experience, expertise and relationship networks. however, the existing lit - erature still ignores the fact that, even the same directors can have different governance effects due to different firms they serve in. For example, from the perspective of growth stage, Field et al. (2013) find that, among IPO firms, the rich experience of busy boards con- tributes positively to firm value; however, these positive effects of busy boards extend to all but the most established firms. Benefits are lowest among Forbes 500 firms. From the per - spective of energy investment, Masulis and Mobbs (2014) find that directors with multiple directorships distribute their effort unequally according to the directorship's relative prestige, and the sensitivity of forced CEO departure to poor performance rises when a larger fraction of independent directors view the board as relatively more prestigious. In other words, more effort investment improves the governance performance. In China, issuing independent opinions on major issues in companies is an important way for independent directors to carry out their monitoring duties. The major issues include “nomination and appointment of directors”, “appointment and dismissal of senior executives”, “remuneration of directors, senior executives”. Because detailed reasons for the leaving of directors and senior executives are hard to obtain and their tenure is not regulated, it is difficult for us to distinguish whether the turnover is voluntary or forced. Taking the reality into account, it is more reasonable to examine the governance performance of independent directors from the perspective of CEO compensation. We expect that the greater proportion of independent directors who invest more effort, the lower the CEO’s extra compensation, and the higher the sensitivity between the compensation and performance. Based on the above analysis, we propose our third hypothesis: H3: The more independent directors who view the target firm as having a relatively higher reputation and closer distance, the lower the CEO’s extra compensation, and the higher the sensitivity between the compensation and performance. ChInA JOURnAL OF ACCOUnTInG STUDIES 293 4. Research design 4.1. Sample and data The sample in this study is Chinese A-share listed companies from 2002 to 2013. Since the information on board meeting attendance is disclosed from 2004, the sample period of h1 ranges from 2004 to 2013. The independent directors’ personal information and addresses are hand collected. Other data is collected from GTA CSMAR, with some supplemented manually. For the initial sample, we exclude: (1) financial and insurance firms; and (2) obser - vations with missing variables. Our final sample includes 49,188 firm-year observations. In order to avoid the effects of extreme values, all continuous variables are winsorized at both top and bottom 1% level. We use STATA software to process data in this paper. 4.2. Model settings and variable definitions We structure models (1-1) and (1-2) to test h1. If h1 is supported, b in models (1-1) and (1-2) should be significantly negative, and b should be significantly positive. Meeting12 = b + b ∗ Highest_Repu + b ∗ Lowest_Repu + b ∗ Control + e (1-1) 0 1 2 Meeting12 = b + b ∗ Nearest_Dist + b ∗ Farest_Dist + b ∗ Control + e (1-2) 0 1 2 We structure models (2-1) and (2-2) to test h2. If h2 is as expected, b4 in models (2-1) and (2-2) should be not significant or significantly positive, and b should be significantly negative. Probit(Turnover)= b + b ∗ Highest_Repu + b ∗ Lowest_Repu 0 1 2 + b ∗ ROA + b ∗ ROA ∗ Highest_Repu + b ∗ ROA ∗ Lowest_Repu 3 4 5 (2-1) + b ∗ Control + e Probit(Turnover)= b + b ∗ Nearest_Dist + b ∗ Farest_Dist 0 1 2 + b ∗ ROA + b ∗ ROA ∗ Nearest_Dist 3 4 (2-2) + b ∗ ROA ∗ Farest_Dist + b Control + e We structure models (3-1) and (3-2) to test h3. All explanatory variables are lagged for one year to alleviate the endogeneity issues. If h3 is supported, b1 in models (3-1) and (3-2) should be significantly negative, and b3 should be significantly positive. Extra_pay = b + b ∗ Highest_Rindir + b ∗ ROA t 0 1 t−1 2 t−1 (3-1) + b ∗ ROA ∗ Highest_Rindir + b ∗ Control + e 3 t−1 t−1 t−1 t−1 Extra_Pay = b + b ∗ Nearest_Rindir + b ∗ ROA t 0 1 t−1 2 t−1 (3-2) + b ∗ ROA ∗ Nearest_Rindir + b ∗ Control + e 3 t−1 t−1 t−1 t−1 294 Y. QUAn AnD D. ChEn The dependent variables are Meeting12, Turnover and Extra_Pay in models (1)-(3), respec- tively. Meeting12 represents the time and energy independent directors invest, which is defined as the proportion of board meetings that directors do not attend personally. The greater the value of Meeting12, the less time and energy they invest. We also further decom- pose Meeting12 into Meeting1 and Meeting2, which are defined as the proportion of attend- ance by agents and total absence respectively. Turnover is a dummy variable of forced turnover of independent directors, which is equal to one if independent directors leave and the term of office is shorter than three years, and 0 otherwise. Extra_Pay represents the excess compensation of CEO. In all the models, we add Year, Industry and Province dummy variables to control for year, industry and province fixed effects. Table 1 summarises our variables. Table 1. d efinition of variables. Variable Definition of variables Meeting12 Proportion of board meetings that directors do not attend personally Meeting1 Proportion of attendance by agents Meeting2 Proportion of total absence Turnover indicator variable that equals to one if independent directors leave and the term of office is shorter than three years, and 0 otherwise Extra_Pay Ceo ’s excess compensation. f ollowing Core et al. (1999) and Wu, lin, and Wang (2010), we use the regression residuals to measure that extra compensation Highest_ indicator variable that equals one if the directorship is the director's largest one measured by the firm Repu size, and 0 otherwise Lowest_Repu indicator variable that equals one if the directorship is the director's smallest one measured by the firm size, and 0 otherwise Farest_Dist indicator variable that equals one if the directorship is the director's farthest one measured by the distance between the firm and director, and 0 otherwise Nearest_Dist indicator variable that equals one if the directorship is the director's nearest one measured by the distance between the firm and director, and 0 otherwise Highest_ Proportion of independent directors who view the target firm as the highest reputation one. Highest_ Rindir Rindir=∑(Highest_Repu)/Indirector, Indirector indicates the total number of independent directors Nearest_ Proportion of independent directors who view the target firm as the nearest one, Nearest_Rindir= Rindir ∑(Nearest_Dist)/Indirector Meeting number of meetings directors required to attend in the year Indir_Pay natural logarithm of independent directors allowance Indir_Age natural logarithm of the independent directors age Indir_Tenure independent director’s tenure, calculated as the number of years between the current year and the first year the independent director assumes the duty in the firm minus the number of years she doesn’t assume the duty during this period Education independent director’s education, doctor 5, master 4, undergraduate 3, college 2, otherwise 1 Rindirector Proportion of independent directors BoardSize t otal number of board of directors Lnsize natural logarithm of the total assets at the fiscal year end ROA return on assets CEO_Edu Ceo ’s education: doctor 5, master 4, undergraduate 3, college 2, otherwise 1 Indir_Acc indicator variable that equals one if the director has accounting backgrounds, and 0 otherwise Indir_Law indicator variable that equals one if the director has legal backgrounds, and 0 otherwise Indir_Fin indicator variable that equals one if the director has financial backgrounds, and 0 otherwise Indir_PC indicator variable that equals one if the director has political backgrounds, and 0 otherwise Gender indicator variable that equals one if the director is female, and 0 otherwise CEO_Age natural logarithm of Ceo age Dual indicator variable that equals one if the chairman and the Ceo is the same person, and 0 otherwise Mshare t op management shareholding ratio Complex t he involved industry number Levevage r atio of total debt to total assets at the fiscal year end Growth growth rate of total assets SOE indicator variable that equals one if the firm is state-owned and 0 otherwise Zindex r atio of the largest shareholder to the second largest List_Age natural logarithm of listed years notes: a ccording to the definition, for independent directors who serve only in a company, the value of Highest_Repu, Lowest_Repu, Nearest_Dist, Farest_Dist are all equal to 1. ChInA JOURnAL OF ACCOUnTInG STUDIES 295 5. Empirical results and analysis 5.1. Descriptive statistics Panel A and B of Table 2 present descriptive statistics related to independent directors and firms respectively. During the sample period, directors need to attend eight board meetings on average a year, with a maximum of 72 times. Approximately 5.7% of the independent directors do not attend board meetings personally, with a maximum of 100%; the proportion of attendance by agents and absence are 4.9% and 0.8% respectively. This result indicates that directors do have a preference in their effort allocation. The turnover ratio of independ - ent directors is 15.8% on average, which is consistent with the findings of Li and Qin (2011). The average education level of independent directors is between undergraduate and master. while slightly higher than CEOs. 5.2. The basic regression analysis 5.2.1. H1a: Reputation and effort allocation Table 3 reports the results of h1a. The dependent variables in columns (1) to (3) are Meeting12, Meeting1 and Meeting2 respectively. The results in column (1) show that the coefficient on Highest_Repu is significantly negative at 1% level, the coefficient on Lowest_ Repu is not significant, and the difference between the two coefficients is significant at the 1% level (F value is 16.88, P value is 0.0000). This indicates that, for companies with relatively high (low) reputation, the proportion of board meetings that independent directors attend personally is higher (lower). Results in column (2) and (3) show that, Highest_Repu is signifi- cantly negatively correlated with Meeting1 at the 1% level, Lowest_Repu is positively but not significantly correlated with Meeting1, and the difference between the two coefficients is significant at the 1% level (F value is 16.04, P value is 0.0001). Highest_Repu is positively but not significantly correlated with Meeting2, and Lowest_Repu is significantly positively corre - lated with Meeting2 at the 10% level. however, the difference between the two coefficients is not significant (F value is 0.84, P value is 0.3603). This indicates that, for companies with relatively high reputation, the proportion of board meetings attending by agents is signifi- cantly lower; for companies with relatively low reputation, the proportion of absence is higher but not significant. The results in Table 3 overall indicate that for companies with relatively high (low) reputation, independent directors invest more (less) time and energy. h1a has been confirmed. The results of the control variables show that: Indir_Pay is significantly positively correlated with Meeting1 at 10% level, and significantly negatively correlated with Meeting2 at 1% level. This indicates that the amount of allowance can inspire independent directors to devote more time and effort to some extent, which is consistent with the findings of Adams and Ferreira (2008) and Masulis and Mobbs (2014). Meeting is significantly negatively correlated with Meeting12 and Meeting1 at 1% level, and positively correlated with Meeting2 at 1% level. This indicates that the greater the number of meetings directors require to participate, the on o ctober 19, 2013, the Central organization d epartment issued “Comments on further standardizing the part time (working) problem of party and government leaders in business”, require to clean up the illegal part time jobs (working) of Party and government leading cadres in business within a definite time. i n order to prevent the impact of this shock on the conclusions, in h2, we removed the sample in 2013. t he results are consistent when the sample in 2013 is retained. When calculating standard deviation, we cluster by independent directors, similarly hereinafter. 296 Y. QUAn AnD D. ChEn Table 2. Variable descriptive statistics. Panel A: director level N Mean Standard deviation Min. P25 Median P75 Max. Meeting12 41,739 0.06 0.12 0.00 0.00 0.00 0.09 1.00 Meeting1 41,739 0.05 0.11 0.00 0.00 0.00 0.06 1.00 Meeting2 41,776 0.01 0.05 0.00 0.00 0.00 0.00 1.00 Turnover 47,922 0.16 0.36 0.00 0.00 0.00 0.00 1.00 Highest_Repu 47,922 0.79 0.40 0.00 1.00 1.00 1.00 1.00 Lowest_Repu 47,922 0.79 0.40 0.00 1.00 1.00 1.00 1.00 Nearest_Dist 47,821 0.79 0.40 0.00 1.00 1.00 1.00 1.00 Farest_Dist 47,821 0.79 0.41 0.00 1.00 1.00 1.00 1.00 Meeting 41,776 8.23 4.05 1.00 6.00 8.00 10.00 72.00 Indir_Pay 41,776 10.72 0.58 8.99 10.31 10.82 11.00 12.21 Indir_Age 47,922 3.92 0.20 3.50 3.76 3.89 4.09 4.32 Indir_Tenure 47,922 0.96 0.64 0.00 0.69 1.10 1.39 2.08 Education 47,129 3.84 0.96 1.00 3.00 4.00 5.00 5.00 Indir_Acc 47,922 0.31 0.46 0.00 0.00 0.00 1.00 1.00 Indir_Law 47,922 0.15 0.35 0.00 0.00 0.00 0.00 1.00 Indir_Fin 47,922 0.11 0.31 0.00 0.00 0.00 0.00 1.00 Indir_PC 47,922 0.18 0.39 0.00 0.00 0.00 0.00 1.00 Gender 47,922 0.12 0.33 0.00 0.00 0.00 0.00 1.00 BoardSize 47,922 9.72 2.17 4.00 9.00 9.00 11.00 19.00 Rindirector 47,922 0.38 0.10 0.17 0.33 0.33 0.43 0.78 Lnsize 47,922 21.68 1.32 18.64 20.80 21.57 22.41 25.79 ROA 47,922 0.02 0.08 −0.43 0.01 0.03 0.06 0.22 Panel B: firm level Extra_Pay 10,828 0.01 0.66 −1.95 −0.37 0.03 0.43 1.47 Highest_Rindir 10,828 0.79 0.25 0.00 0.67 0.83 1.00 1.00 Nearest_Rindir 10,828 0.79 0.24 0.00 0.67 0.80 1.00 1.00 CEO_Education 10,828 3.41 0.81 1.00 3.00 3.00 4.00 5.00 CEO_Age 10,828 3.84 0.14 3.47 3.75 3.85 3.93 4.13 Duality 10,828 0.13 0.33 0.00 0.00 0.00 0.00 1.00 Mshare 10,828 0.01 0.04 0.00 0.00 0.00 0.00 0.34 Complex 10,828 3.91 1.80 1.00 2.58 3.94 5.00 9.00 Lnsize 10,828 21.69 1.28 18.54 20.84 21.60 22.43 25.51 Leverage 10,828 0.56 0.31 0.08 0.39 0.54 0.67 2.64 Growth 10,828 0.13 0.26 −0.48 −0.01 0.09 0.21 1.51 SOE 10,828 0.68 0.47 0.00 0.00 1.00 1.00 1.00 Zindex 10,828 22.29 45.68 1.02 2.03 6.00 20.88 327.50 List_Age 10,828 2.28 0.55 0.00 2.08 2.40 2.64 3.05 greater the proportion of absence, which supports the assertion about the limited energy of independent directors. Indir_Tenure is significantly positively correlated with Meeting12 and Meeting1 at 1% level, probably because the longer the tenure, the more familiar with the business, and the less necessary for independent directors to attend the meetings per- sonally. Indir_Acc, Indir_Law and Indir_Financial are significantly negatively correlated with all dependent variables at least at 5% level except Meeting2, which indicates that directors with accounting, law and finance backgrounds are more likely to attend meetings personally and less likely to attend by agents or absent. Indir_PC is significantly positively correlated with Meeting2 at 10% level, indicating that independent directors with politics background are more likely to absent from the meeting. BoardSize is significantly positively correlated with Meeting12 and Meeting1 at 1% level, which is consistent with the findings of Masulis and Mobbs (2014). Lnsize is positively correlated with Meeting12 and Meeting1 at 1% level and significantly negatively correlated with Meeting2 at 5% level, which is probably because, on one hand, the number of independent directors in large firms is larger, it is easier to attend the board meetings by agents; on the other hand, large firms can provide independent ChInA JOURnAL OF ACCOUnTInG STUDIES 297 Table 3. reputation and board meetings attendance. Variables (1) Meeting12 (2) Meeting1 (3) Meeting2 Highest_Repu −0.008*** −0.009*** 0.000 (−4.71) (−5.45) (0.44) Lowest_Repu 0.001 0.000 0.001* (0.78) (0.00) (1.81) Indir_Pay −0.001 0.002* −0.003*** (−0.86) (1.72) (−6.20) Meeting −0.001*** −0.001*** 0.000*** (−7.70) (−10.68) (6.33) Indir_Age −0.011** −0.009** 0.000 (−2.17) (−2.05) (−0.34) Indir_Tenure 0.006*** 0.005*** 0.000 (5.37) (5.21) (1.51) Education 0.009*** 0.007*** 0.002*** (9.49) (8.37) (5.81) Indir_Acc −0.023*** −0.020*** −0.002*** (−12.87) (−12.47) (−5.08) Indir_Law −0.010*** −0.008*** −0.002** (−4.08) (−3.59) (−2.40) Indir_Fin −0.005** −0.005** 0.000 (−1.99) (−2.28) (0.09) Indir_PC −0.000 −0.002 0.001* (−0.07) (−0.98) (1.80) Gender 0.004* 0.004* 0.000 (1.76) (1.76) (0.76) BoardSize 0.003*** 0.003*** 0.000* (9.14) (9.57) (1.80) Lnsize 0.002*** 0.002*** −0.000** (2.84) (3.60) (−2.05) ROA −0.042*** −0.018** −0.020*** (−4.73) (−2.35) (−6.19) Constant 0.048* −0.035 0.027*** (1.74) (−1.21) (3.57) Industry & Year & Province Control Control Control obs# 41,739 41,759 41,776 a dj-R 0.0770 0.0607 0.0418 F-value 48.05 37.46 25.60 notes: *, **, and *** to indicate 10, 5 and 1% level of significance respectively in all the regression tables. We bold all main variable of interest in order to highlight conclusions. directors with more benefits, thus the probability of absence in large firms is lower. ROA is significantly negatively correlated with Meeting12, Meeting1 and Meeting2 at least at the 5% level, indicating that directors devote more time and effort to firms in good condition. 5.2.2. H1b: Distance and effort allocation Table 4 reports the results of h1b. Panels A, B and C measure the distance between the independent directors and companies by “spatial distance”, “shortest travel time by train” and “train departure frequency” respectively. Specifically, if the spatial distance is nearest (farthest), or the travel time by train is shortest (longest), or the train departure frequency is fastest (slowest), then Nearest_Dist (Farest_Dist) equals to 1, and 0 otherwise. The depend- ent variables of Panel A, B and C in column (1) to (3) are Meeting12, Meeting1 and Meeting2 respectively. Results in Table 4 show that Nearest_Dist is significantly negatively correlated t he "spatial distance" has been calculated according to the spherical coordinates; the data of "shortest travel time by train" and "train departure frequency" come from the following website on april 30, 2015: http://www.12306.cn/mormhweb/. 298 Y. QUAn AnD D. ChEn Table 4. distance and board meetings attendance. Variables Panel A: spatial distance Panel B: shortest travel time by train Panel C: Train departure frequency (1) Meeting12 (2) Meeting1 (3) Meeting2 (1) Meeting12 (2) Meeting1 (3) Meeting2 (1) Meeting12 (2) Meeting1 (3) Meeting2 Nearest_Dist −0.008*** −0.009*** 0.000 −0.005*** −0.006*** 0.000 −0.006*** −0.006*** 0.000 (−4.73) (−5.54) (0.52) (−3.13) (−3.66) (0.37) (−3.39) (−3.98) (0.22) Farest_Dist −0.000 −0.001 0.001 −0.002 −0.002 0.001 −0.002 −0.003* 0.000 (−0.11) (−0.75) (1.38) (−0.90) (−1.59) (1.35) (−1.18) (−1.65) (0.95) Indir_Pay −0.001 0.002* −0.003*** −0.001 0.003* −0.003*** −0.001 0.003* −0.003*** (−0.821) (1.75) (−6.27) (−0.69) (1.92) (−6.18) (−0.70) (1.90) (−6.20) Meeting −0.001*** −0.001*** 0.000*** −0.001*** −0.002*** 0.000*** −0.001*** −0.002*** 0.000*** (−7.64) (−10.60) (6.37) (−7.73) (−10.70) (6.32) (−7.74) (−10.71) (6.32) Indir_Age −0.011** −0.009** −0.001 −0.011** −0.009* −0.001 −0.011** −0.009* −0.001 (−2.17) (−2.05) (−0.38) (−2.06) (−1.91) (−0.39) (−2.09) (−1.92) (−0.44) Indir_Tenure 0.006*** 0.005*** 0.000 0.006*** 0.005*** 0.000 0.006*** 0.005*** 0.000 (5.35) (5.19) (1.60) (5.35) (5.17) (1.58) (5.35) (5.18) (1.57) Education 0.009*** 0.007*** 0.002*** 0.009*** 0.007*** 0.001*** 0.009*** 0.007*** 0.001*** (9.46) (8.33) (5.85) (9.51) (8.42) (5.77) (9.43) (8.36) (5.69) Indir_Acc −0.023*** −0.020*** −0.002*** −0.023*** −0.020*** −0.002*** −0.023*** −0.020*** −0.003*** (−12.97) (−12.55) (−5.11) (−13.05) (−12.61) (−5.19) (−13.06) (−12.59) (−5.23) Indir_Law −0.010*** −0.008*** −0.001** −0.010*** −0.008*** −0.002** −0.010*** −0.008*** −0.002** (−4.07) (−3.58) (−2.38) (−4.13) (−3.62) (−2.43) (−4.14) (−3.63) (−2.45) Indir_Fin −0.005** −0.005** 0.000 −0.005** −0.005** 0.000 −0.005** −0.006** 0.000 (−2.01) (−2.30) (0.09) (−2.00) (−2.30) (0.14) (−2.01) (−2.30) (0.12) Indir_PC −0.000 −0.002 0.001* 0.000 −0.002 0.001* 0.000 −0.002 0.001* (−0.02) (−0.94) (1.85) (0.06) (−0.88) (1.91) (0.05) (−0.89) (1.90) Gender 0.004* 0.004* 0.000 0.004* 0.004* 0.000 0.004* 0.004* 0.001 (1.76) (1.76) (0.78) (1.77) (1.75) (0.80) (1.79) (1.77) (0.83) BoardSize 0.003*** 0.003*** 0.000* 0.003*** 0.003*** 0.000* 0.003*** 0.003*** 0.000* (8.94) (9.40) (1.71) (9.05) (9.47) (1.81) (9.05) (9.46) (1.83) Lnsize 0.001** 0.002*** −0.000** 0.001** 0.002*** −0.000** 0.001** 0.002*** −0.000** (2.06) (2.87) (−2.42) (2.03) (2.84) (−2.41) (2.02) (2.84) (−2.42) ROA −0.042*** −0.017** −0.020*** −0.041*** −0.017** −0.020*** −0.041*** −0.017** −0.020*** (−4.68) (−2.30) (−6.18) (−4.63) (−2.23) (−6.18) (−4.65) (−2.26) (−6.18) Constant −0.008 −0.046 0.047*** 0.229*** 0.0260 0.044*** 0.231*** 0.026 0.045*** (−0.28) (−0.00) (6.18) (3.07) (0.93) (6.28) (3.09) (0.94) (6.34) Industry & Year Control Control Control Control Control Control Control Control Control & Province obs# 41,661 41,681 41,697 41,448 41,468 41,484 41,448 41,468 41,484 a dj-R 0.0769 0.0607 0.0418 0.0771 0.0606 0.0421 0.0772 0.0607 0.0421 F-value 47.92 36.92 25.58 47.80 37.17 25.66 47.84 37.19 25.63 ChInA JOURnAL OF ACCOUnTInG STUDIES 299 with Meeting12 and Meeting1 at the 1% level under all the three measurements; Farest_Dist is negatively correlated with Meeting1 at the 10% level only under the measurement of “train departure frequency”; and the difference between the two coefficients is significant at least at the 10% level except in Panel B. This indicates that, for companies with relatively close distance, the proportion of board meetings that independent directors attend through agents is significantly lower, while the difference of absence is not significant. The results in Table 4 indicate that, for companies with relatively far distance, the proportion of independ- ent directors who are absent from board meetings is higher; for companies with relatively near distance, the proportion of board meetings independent directors attend personally (through agents) is higher (lower). h1b has been confirmed. 5.2.3. H2: Reputation (distance) and turnover performance sensitivity Table 5 reports the results of h2. All the dependent variables are Turnover. The results in column (1) and (2) show that Highest_Repu and Lowest_Repu are significantly positively correlated with Turnover, but both the coefficient and the significance level of Highest_Repu are smaller than those of Lowest_Repu, indicating that the probability of independent direc- tors leaving a relatively high reputation company is smaller than those leaving a relatively low reputation one. ROA is significantly negatively correlated with Turnover at least at the 5% level, indicating that the worse the accounting performance, the higher the probability that independent directors leave. The interaction term Highest_Repu*ROA is not significant and the interaction term Lowest_Repu*ROA is negatively significant at 10% level, indicating that the sensitivity of forced turnover of independent directors to performance is higher in firms with relatively low reputation. The results in column (3) and (4) show that Nearest_Dist and Farest_Dist are significantly positively correlated with Turnover, but both the coefficient and significance level of Farest_ Dist are larger than those of Nearest_Dist, indicating that the probability of independent directors turnover from a relatively far company is larger than from a relatively close one, which is consistent with the findings of Tan et al. (2006). ROA is significantly negatively cor - related with Turnover at the 1% level, which is consistent with column (1) and (2). however, the interaction terms in column (4) are not significant. The results in column (3) and (4) indicate that distance affects the probability of independent directors’ turnover, but does not affect the forced departure sensitivity to performance. These results are logically con- sistent with the findings in column (1) and (2) of Table 5. Knyazeva et al. (2013) believe that in order to employ excellent independent directors, the largest firms (top quartile) are prone to recruit directors nationally. The reputation of a large company is higher, while the sensi- tivity of forced departure of independent directors to performance is lower. Results in Table 5 indicate that the probability of independent directors’ turnover from a company with relatively high reputation (far distance) is smaller (larger) than from a company with a relatively low reputation (near distance). The sensitivity of forced department of inde- pendent directors to performance is higher in firms with relatively low reputation. h2 of this article has been confirmed. The results of control variables show that Indir_Tenure is signifi- cantly positively correlated with Turnover at the 1% level, probably because the Guidance however, the difference between the two coefficients is not significant. in column (3) and (4) of t able 5, we only report the regression results under the measurement of “spatial distance”, the regression results under the other two measurements are roughly unchanged, similarly hereinafter. 300 Y. QUAn AnD D. ChEn Table 5. reputation (distance) and turnover performance sensitivity. Variables (1) (2) (3) (4) Highest_Repu 0.043** 0.039* (1.95) (1.76) Lowest_Repu 0.060*** 0.074*** (2.70) (3.07) Highest_Repu*ROA 0.258 (1.05) Lowest_Repu*ROA −0.527* (−1.66) Nearest_Dist 0.042* 0.040* (1.92) (1.78) Farest_Dist 0.057*** 0.057*** (2.62) (2.55) Nearest_Dist *ROA 0.112 (0.43) Farest_Dist *ROA −0.041 (−0.15) ROA −1.060*** −0.775** −1.062*** −1.111*** (−10.52) (−2.38) (−10.54) (−3.74) Indir_Age 0.039 0.039 0.034 0.034 (0.79) (0.80) (0.70) (0.69) Indir_Tenure 0.645*** 0.646*** 0.645*** 0.645*** (34.03) (34.02) (33.94) (33.94) Rindirector 2.983*** 2.980*** 2.990*** 2.990*** (35.97) (35.94) (36.01) (36.01) Indir_Acc −0.044** −0.044** −0.042** −0.042** (−2.14) (−2.15) (−2.04) (−2.03) Indir_Law −0.013 −0.013 −0.010 −0.010 (−0.50) (−0.50) (−0.40) (−0.40) Indir_Fin 0.064** 0.064** 0.065** 0.065** (2.16) (2.16) (2.20) (2.20) Indir_PC −0.041* −0.042* −0.039 −0.039 (−1.70) (−1.72) (−1.61) (−1.60) Gender −0.003 −0.003 −0.004 −0.004 (−0.11) (−0.11) (−0.13) (−0.13) Lnsize −0.049*** −0.049*** −0.051*** −0.051*** (−6.03) (−6.02) (−6.56) (−6.56) Constant −2.220*** −2.229*** −2.163*** −2.160*** (−8.57) (−8.60) (−8.45) (−8.44) Industry & Year & Province Control Control Control Control obs# 47,922 47,922 47,821 47,821 Pseudo R 0.1239 0.1241 0.1241 0.1241 Wald chi 2642.65 2650.27 2636.24 2638.42 clearly stipulates that the term of independent directors should be no more than six years. Rindirector is significantly positively correlated with Turnover at 1% level, indicating that the larger the number of independent directors, the higher the probability of turnover. Lnsize is significantly negatively correlated with Turnover at the 1% level, which is consistent with the findings in column (1) and (2) of Table 5. 5.2.4. H3: Effort allocation and governance effect Table 6 reports the results of h3. When calculating the standard deviation of the coefficients, we cluster by industry. All dependent variables in Table 6 are Extra_Pay. The results in column (1) and (2) show that whether we are controlling ROA alone or controlling ROA and Highest_ Rindir*ROA simultaneously, Highest_Rindir is significantly negatively correlated with Extra_ Pay at the 1% level. Meanwhile, Highest_Rindir*ROA is significantly positively correlated with ChInA JOURnAL OF ACCOUnTInG STUDIES 301 Table 6. reputation (distance) and Ceo Pay performance sensitivity. Variables (1) (2) (3) (4) Highest_Rindir −0.155*** −0.174*** t−1 (−3.19) (−3.71) Highest_Rindir *ROA 0.873*** t−1 t−1 (3.39) Nearest_Rindir −0.101*** −0.097*** t−1 (−2.83) (−2.83) Nearest_Rindir *ROA −0.134 t−1 t−1 (−0.55) ROA 0.666*** 0.003 0.674*** 0.781*** t−1 (4.04) (0.01) (4.09) (4.56) CEO_Education 0.064*** 0.064*** 0.065*** 0.065*** t−1 (6.57) (6.56) (6.66) (6.67) CEO_Age 0.297*** 0.295*** 0.300*** 0.300*** t−1 (4.57) (4.55) (4.43) (4.43) Duality −0.036** −0.036** −0.037** −0.037** t−1 (−2.60) (−2.52) (−2.55) (−2.57) Mshare −0.266** −0.270** −0.263** −0.264** t−1 (−2.15) (−2.16) (−2.19) (−2.19) Complex 0.017** 0.017** 0.017** 0.017** t−1 (2.56) (2.55) (2.60) (2.60) Lnsize −0.021*** −0.021*** −0.032*** −0.032*** t−1 (−4.10) (−4.11) (−5.81) (−5.84) Leverage 0.025 0.027 0.021 0.021 t−1 (1.20) (1.24) (1.06) (1.06) Growth 0.034 0.033 0.034 0.034 t−1 (1.42) (1.38) (1.44) (1.45) SOE 0.010 0.011 0.014 0.014 t−1 (0.34) (0.38) (0.51) (0.50) Zindex −0.001*** −0.001*** −0.001*** −0.001*** t−1 (−3.01) (−3.02) (−2.99) (−2.98) List_Age 0.011 0.011 0.011 0.011 t−1 (0.57) (0.54) (0.56) (0.56) Constant −0.802* −0.764* −0.676* −0.679* t−1 (−2.02) (−1.90) (−1.89) (−1.89) Industry & Year & Province Control Control Control Control obs# 10828 10828 10828 10828 a dj-R 0.1073 0.1080 0.1055 0.1054 F-value 19.33 19.21 18.98 18.72 Extra_Pay at the 1% level. This indicates that the more independent directors who view the target firm as one with relatively higher reputation, the lower the CEO’s extra compensation, and the higher the sensitivity between the extra compensation and performance. The results in column (3) and (4) show that whether we are controlling ROA alone or controlling ROA and Nearest_Rindir*ROA simultaneously, Nearest_Rindir is significantly neg - atively correlated with Extra_Pay at 1% level, indicating that the more independent directors who view the target firm as a closer one, the lower the CEO’s extra compensation. however, Nearest_Rindir*ROA is not signic fi ant. The results in Table 6 indicate that the more independ- ent directors who view the target firm as one with relatively higher reputation, the lower the CEO’s extra compensation and the higher sensitivity between the extra compensation and performance; the more independent directors who view the target firm as a closer one, the lower the CEO’s extra compensation. h3 of this article has been confirmed. The results of the control variables show that the higher-educated and the older the CEO, the higher the extra compensation; for firms with the chairman and the CEO held by one person, 302 Y. QUAn AnD D. ChEn the higher the top management shareholding; the less complex the business, the larger the size, the higher the ownership concentration, the lower the CEO’s extra compensation. 5.3. Additional tests 5.3.1. Reputation and effort allocation grouped by distance In h1, we find that independent directors allocate more time and effort to companies with higher reputation and closer distance. however, we did not control the distance between the independent directors and the companies when exploring the relative reputation. Therefore, we are not sure whether the relationship between the reputation and the effort allocation is ae ff cted by distance. By distinguishing whether the distance between the inde - pendent directors and the companies is farther than the median in that year, Table 7 divides the full sample into “long distance group” and “short distance group” and re-examine h1a. The results in long distance group show that Highest_Repu is negatively significantly corre - lated with Meeting12 and Meeting1 at 1% level; Lowest_Repu is significantly positively Table 7. reputation and board meetings attendance grouped by distance. Long distance group Short distance group Variables (1) Meeting12 (2) Meeting1 (3) Meeting2 (4) Meeting12 (5) Meeting1 (6) Meeting2 Highest_Repu −0.012*** −0.013*** 0.000 −0.003 −0.003* 0.000 (−4.623) (−5.25) (0.37) (−1.37) (−1.85) (0.51) Lowest_Repu 0.002 −0.001 0.002** 0.001 0.001 0.000 (0.58) (−0.32) (2.26) (0.63) (0.50) (0.18) Indir_Pay −0.004* −0.001 −0.003*** −0.000 0.004** −0.003*** (−1.90) (−0.36) (−4.37) (−0.06) (2.24) (−4.60) Meeting −0.001*** −0.002*** 0.000*** −0.001*** −0.001*** 0.000*** (−6.22) (−9.04) (4.79) (−4.13) (−5.58) (4.10) Indir_Age −0.010 −0.008 −0.000 −0.005 −0.004 −0.000 (−1.31) (−1.19) (−0.20) (−0.78) (−0.71) (−0.09) Indir_Tenure 0.006*** 0.005*** 0.000 0.006*** 0.006*** 0.000 (3.77) (3.49) (1.06) (4.32) (4.46) (1.07) Education 0.009*** 0.007*** 0.002*** 0.009*** 0.007*** 0.001*** (6.21) (5.36) (4.08) (7.74) (6.92) (4.05) Indir_Accounting −0.024*** −0.021*** −0.002*** −0.018*** −0.015*** −0.002*** (−8.88) (−8.60) (−3.21) (−8.46) (−8.07) (−3.55) Indir_Law −0.010*** −0.008*** −0.001* −0.007** −0.005* −0.001 (−2.79) (−2.52) (−1.68) (−2.23) (−1.81) (−1.55) Indir_Financial −0.006 −0.006* 0.001 −0.007** −0.006** −0.001 (−1.48) (−1.87) (0.52) (−2.20) (−2.17) (−0.58) Indir_PC 0.004 0.001 0.002* −0.003 −0.004* 0.001 (1.14) (0.34) (1.90) (−1.25) (−1.76) (0.71) Gender 0.003 0.004 −0.000 0.003 0.002 0.001 (1.02) (1.38) (−0.48) (1.30) (0.92) (1.29) BoardSize 0.004*** 0.004*** −0.000 0.003*** 0.002*** 0.000*** (7.23) (8.14) (−0.37) (5.48) (5.07) (2.87) Lnsize 0.003*** 0.003*** 0.000 0.001 0.002** −0.001*** (2.51) (2.67) (0.10) (1.20) (2.35) (−3.48) ROA −0.060*** −0.025** −0.029*** −0.021* −0.011 −0.008* (−4.73) (−2.32) (−6.25) (−1.79) (−1.08) (−1.88) Constant 0.062 0.033 0.026*** 0.173** −0.088*** 0.039*** (1.39) (0.84) (2.60) (2.12) (−2.67) (3.82) Industry & Year & Control Control Control Control Control Control Province obs# 20,667 20,680 20,683 21,072 21,079 21,093 a dj-R 0.0945 0.0761 0.0510 0.0754 0.0595 0.0464 F-value 15.17 12.55 5.67 10.57 8.75 5.40 ChInA JOURnAL OF ACCOUnTInG STUDIES 303 correlated with Meeting2 at 5% level. The results in the short distance group show that only Highest_Repu is significantly negatively correlated with Meeting1 at 10% level. The difference between the two coefficients is significant at least at the 10% level. The results grouped by distance show that the bias is more obvious when the distance is far. For example, directors attend board meetings personally more often for companies with relatively high reputation compared with companies with relatively low reputation. While when the distance becomes close, the bias is relatively weak. The results in Table 7 indicate that directors with multiple directorships allocate their effort unequally based on not only the relative prestige, but also the relative distance. 5.3.2. Multiple-board independent directors and their opinion The Guidance stipulates that independent directors should voice one of the following opin- ions on major issues of listed companies: consent; reservations and the reasons; objections and the reasons; unable to express an opinion and its obstacle. Tang et al. (2010) compare the effects of directorships that independent directors hold on the opinions they issued. They find that the fewer directorships an independent director holds or the higher remu- neration he obtains, the lower the possibility that independent directors say "no". A more detailed problem is whether the possibility of multiple-board independent directors say "no" to different firms is different. Is the possibility also affected by the relative reputation and distance? Referring to the literature of Chen, Wang, and Duan (2015) and Ye et al. (2011), we report the regression results of independent directors’ opinions in Table 8. When calcu- lating the standard deviation of the coec ffi ients, we cluster by year to control the correlation among years. All dependent variables in Table 8 are Opinion, an indicator variable that equals one if the opinion is negative, and 0 otherwise. Results in column (1) show that coefficient on Highest_Repu is not significant, the coefficient on Lowest_Repu is positively significant at 5% level, and the difference between the two coefficients is significant at 10% level (F value is 3.24, P value is 0.0719). Results in column (2) show that coefficient on Farest_Dist is posi- tively significant at 10% level, the coefficient on Nearest_Dist is not significant, and the dif- ference between the two coefficients is significant at 10% level (F value is 2.74, P value is 0.0981). Results in Table 8 indicate that the independent directors voice significantly less (more) negative opinions to companies with relatively high reputation (far distance). It is possible because the corporate governance of companies with relatively high reputation is better; and the problems have been communicated before voting due to the low commu- nication costs of near-by companies. 5.4. Robustness test To further improve the reliability of our conclusions, we perform the following robustness test. First, we exclude the single-board independent directors sample (independent directors who serve only one company) and re-examine the hypotheses. Second, when defining Highest_Repu, Lowest_Repu, Nearest_Dist and Farest_Dist, we assigned 1 when the o f course, another possible explanation is that due to the higher cost of board seats lost, the independent directors would issue fewer negative opinions to companies with relatively high reputation (near distance). f or this problem, we will make a deeper analysis in the future. f or space considerations, the results are not reported in the text, but are available on application to the corresponding author. 304 Y. QUAn AnD D. ChEn Table 8. Multiple-board independent directors and their opinion. Variables (1) (2) Highest_Repu 0.134 (0.71) Lowest_Repu 0.643** (2.34) Farest_Dist 0.420* (1.80) Nearest_Dist 0.232 (0.90) Meeting 0.032 0.034 (0.89) (0.96) Boardlock 0.263* 0.253 (1.78) (1.36) Indir_Pay 0.016 0.027 (0.07) (0.11) Indir_Age −0.742 −0.710 (−1.32) (−1.29) Indir_Tenure 0.134 0.128 (0.75) (0.72) Education 0.106*** 0.108*** (2.60) (2.64) Indir_Acc 0.064 0.068 (0.62) (0.68) Indir_Law 0.141 0.151 (0.46) (0.50) Indir_Fin 0.063 0.067 (0.37) (0.39) Indir_PC 0.260 0.264 (1.30) (1.31) Gender −0.069 −0.072 (−0.47) (−0.49) Audit −1.003*** −0.986*** (−3.16) (−3.12) Duality 0.010 0.016 (0.04) (0.07) Mshare 2.324* 2.381* (1.76) (1.81) Instihold 0.956*** 0.964*** (2.95) (3.034) Board 0.700 0.693 (1.50) (1.47) Lnsize −0.110 −0.153 (−0.99) (−1.45) ROA −2.754** −2.762** (−2.45) (−2.44) Growth −0.312 −0.330 (−1.19) (−1.234) SOE −0.027 −0.018 (−0.11) (−0.07) Zindex −0.008** −0.008** (−2.45) (−2.38) List_Age −0.255 −0.265 (−0.89) (−0.91) Constant −3.786 −2.950 (−0.94) (−0.71) Industry & Year Control Control obs# 172,691 172,330 Pseudo R 0.1492 0.1480 Wald chi 1093.54 1083.02 ChInA JOURnAL OF ACCOUnTInG STUDIES 305 directorship is the director's largest/smallest (nearest/farest) one measured by the size (dis- tance) of the firm respectively. Referring to the method of Masulis and Mobbs (2014), we re-examine our primary findings using the following alternative measurements: If the direc - torship is 10% larger (smaller) than the director's smallest (largest) one measured by firm size, Highest_Repu (Lowest_Repu) equals 1, and 0 otherwise. If the directorship is 10% nearer (farther) than the director's farthest (nearest) one measured by firm distance, Nearest_Dist (Farest_Dist) equals 1, and 0 otherwise. Third, the Guidance clearly stipulates that the term of independent directors should not exceed six years. Considering the institutional back- ground, we re-define Turnover in accordance with whether the term of office is shorter than six years and re-examine h2. Fourth, we exclude the retirement sample of independent directors. Fifth, we use market value, and number of employees as our reputation measure- ments. Sixth, to avoid the endogeneity problem that corporate governance of large com- panies is better, we match each firm in which all the independent directors view the target firm as relatively higher reputation with a control firm in the same year and industry that is the most similar in firm size and has at least an independent director who does not view the target firm as the relatively higher reputation, and re-examine h3. The results of the above tests remain unchanged, so our conclusions still hold. 6. Conclusion Using a sample of Chinese A-share listed companies from 2002 to 2013, we investigate how multiple-board independent directors allocate their effort among different firms and the consequences. The results indicate that: (1) Multiple-board independent directors do have preference in their effort allocation. Specifically, more effort will be allocated to companies which have higher reputation, closer geographic distance and more convenient transportation. (2) When directors are far from the firms they work for, their bias among companies with different reputation is easier to generate. (3) The multiple-board independent directors are more likely to leave firms with relatively lower reputation and farther distance due to poor accounting performance, even though their tenure has not expired. (4) Different effort allocation results in different consequences. Specifically, the more independent directors who view the target firm as having a relatively higher reputation, the lower the CEO’s extra compensation, and the higher the sensitivity between compensation and performance; the more independent directors who view the target firm as being relatively closer, the lower the CEO’s extra compensation. Providing evidence that multiple-board independent directors have preference in their effort allocation, this paper not only contributes to the study of independent director characteristics and governance performance theoretically, but also has some policy implications on appointing independent director in practice. however, some deficiencies exist in this study. For example, besides attending board meetings, independent directors also need to spend time on financial reports and important issues (Tan et al., 2006). We measure the effort independent directors invest to companies only from the board meeting attendance, which is too simple. Also, different full-time work of independent directors will cause different schemes of time allocation, and different business matching degree between full-time and part-time work also will affect the supervision efficiency. A deeper analysis by refining identities, industries and professional background of independent directors will be explored in the future. 306 Y. QUAn AnD D. ChEn Acknowledgements and Funding This work was supported by national natural Science Foundation of China [71602191] and [71372032]. We appreciate the helpful comments from Professor Yu Xin and the anonymous reviewers. Disclosure statement no potential conflict of interest was reported by the authors. References Adams, R., & Ferreira, D. (2008). Do directors perform for pay? Journal of Financial Economics, 46, 154–171. Agarwal, V., Ma, L., & Mullally, K. (2016). Managerial multitasking in the mutual fund industry. Available at SSRn: http://ssrn.com/abstract=1910367 or http://dx.doi.org/10.2139/ssrn.1910367 Beasley, M. S. (1996). An empirical analysis of the relation between the board of director composition and financial statement fraud. The Accounting Review, 71, 443–465. Cashman, G. D., Gillan, S. L., & Jun, C. (2012). Going overboard? On busy directors and firm value. Journal of Banking & Finance, 36, 3248–3259. Chen, R., Wang, Z., & Duan, C. Q. (2015). Study on “reverse elimination” of independent directors: An empirical evidence from independent advices. China Industrial Economy, 8, 145–160 (In Chinese). Chen, Y. S., & Xie, D. R. (2011). network location, independent directors governance and investment efficiency. Management World, 7, 113–127 (In Chinese). Conyon, M. J., & Read, L. E. (2006). A model of the supply of executives for outside directorships. Journal of Corporate Finance, 12, 645–659. Core, J. E., holthausen, R. W., & Larcker, D. F. (1999). Corporate governance, chief executive officer compensation, and firm performance. Journal of Financial Economics, 51, 371–406. Du, S. L., & Zhang, J. (2004). An empirical study of independent directors’ compensation. Accounting Research, 9, 82–88 (In Chinese). Fama, E., & Jensen, M. (1983). The separation of ownership and control. Journal of Law and Economics, 26, 301–325. Ferris, S. P., Jagannathan, M., & Pritchard, A. C. (2003). Too busy to mind the business? Monitoring by directors with multiple board appointments. Journal of Finance, 58, 1087–1112. Fich, E. M. (2005). Are some outside directors better than others? Evidence from director appointments by Fortune 1000 firms. The Journal of Business, 78, 1943–1972. Fich, E. M., & Shivdasani, A. (2006). Are busy boards effective monitors? Journal of Finance, 61, 689–724. Field, L., Lowry, M., & Mkrtchyan, A. (2013). Are busy boards detrimental? Journal of Financial Economics., 109, 63–82. hunton, J. E., & Rose, J. M. (2011). Retracted: Effects of anonymous whistle-blowing and perceived reputation threats on investigations of whistle-blowing allegations by audit committee members. Journal of Management Studies, 48, 75–98. Kaplan, S. n., & Reishus, D. (1990). Outside directorships and corporate performance. Journal of Financial Economics, 27, 389–410. Knyazeva, A., Knyazeva, D., & Masulis, R. (2013). The supply of corporate directors and board independence. Review of Financial Studies, 26, 1561–1605. Li, Y., & Qin, Y. h. (2011). Media coverage, reputation mechanism and resignation of independent directors. Finance & Trade Economics, 3, 36–41 (In Chinese). Malloy, C. J. (2005). The geography of equity analysis. Journal of Finance, 60, 719–755. Masulis, R. W., & Mobbs, S. (2014). Independent director incentives: Where do talented directors spend their limited time and energy? Journal of Financial Economics, 111, 406–429. Milgram, S. (1967). The small world problem. Psychology Today, 2, 185–195. O'Brien, P. C., & Tan, h. (2015). Geographic proximity and analyst coverage decisions: Evidence from IPOs. Journal of Accounting and Economics, 59, 41–59. ChInA JOURnAL OF ACCOUnTInG STUDIES 307 Ryan, h., & Wiggins, R. (2004). Who is in whose pocket? Director compensation, board independence, and barriers to effective monitoring. Journal of Financial Economics, 73, 497–524. Shivdasani, A. (1993). Board composition, ownership structure, and hostile takeovers. Journal of Accounting and Economics, 16, 167–198. Shivdasani, A., & Yermack, D. (1999). CEO involvement in the selection of new board members: An empirical analysis. Journal of Finance, 54, 1829–1853. Sun, Z. R., & Zhu, X. M. (2005). The theory study and actuality analysis on the independent director’s compensation system in Chinese public corporations. Nankai Business Review, 8, 21–29 (In Chinese). Tan, J. S. (2003). Independent director and corporate governance: Based on Chinese listed companies. Beijing: China Financial and Economic Publishing house (In Chinese). Tan, J. S. (2003). Independent director and corporate governance: Based on Chinese listed companies. Beijing: China Financial and Economic Publishing house (In Chinese). Tang, X. S., Shen, h., & Du, J. (2010). The supervision motivation of independent directors: Empirical evidence based on the independent advice. Management World, 22, 138–149 (In Chinese). Vafeas, n. (1999). Board meeting frequency and firm performance. Journal of Financial Economics, 53, 113–142. Wan, L. Y., Deng, L., & Zheng, X. L. (2012). network location, independent directors governance and violations: Based on partially observable Bivariate Probit model. Systems Engineering-Theory & Practice, 34, 3091–3102 (In Chinese). Wang, Y. T., Zhao, Z. Y., & Wei, X. Y. (2006). Does independence of the board affect firm performance? Economic Research Journal, 5, 62–73 (In Chinese). Wei, G., Xiao, Z. Z., n ick, T., & Zou, h. (2007). Background of independent directors and corporate performance. Economic Research Journal, 3, 92–105 (In Chinese). Wu, L. S., Lin, J. Y., & Wang, Y. P. (2010). Salary external fairness, equity nature and firm performance. Management World, 3, 117–126. Xie, D. R., & Chen, Y. S. (2012). Board network: Definition, characteristics and measurement. Accounting Research, 3, 44–51 (In Chinese). Ye, K. T., Zhu, J. G., & Lu, Z. F. (2011). The independence of independent directors: Evidence from board voting behavior. Economic Research Journal, 1, 126–139 (In Chinese). Yermack, D. (2004). Remuneration, retention, and reputation incentives for outside directors. The Journal of Finance, 59, 2281–2308. Zhou, F., Tan, J. S., & Jian, Y. Y. (2008). Reputation incentive or economic incentive: An empirical study on job-hopping behaviors of independent directors. China Accounting Review, 2, 177–192 (In Chinese).

Journal

China Journal of Accounting StudiesTaylor & Francis

Published: Jul 2, 2016

Keywords: effort allocation; geographic proximity; governance effect; multiple-board independent directors; reputation

There are no references for this article.