Amsterdam Business School
Msc Business Economics
Finance Track
Master Thesis
Promotion puzzles: the influence of social connections on job promotion
Xiaoyu Liu （10604286）
Supervisor： Tomislav Ladika
Abstract
This paper examines the effect of senior managers’ social tie to CEO on their probability of promotion. Using the data of 260 representative S&P 500 firms and 3500 senior managers, the paper finds a positive correlation between manager’s social connection and probability of promotion. Besides, among three types of connections (firm connection, education connection, social club connection), social club connection plays the most important role in the promotion process. However, it is not clear that whether strong CEO power could strengthen the effect of social capital on job promotion. Besides, the senior managers with close relationship to CEO seem to have a higher education level and promotion rate, but whether the higher education level indicates higher personal ability is still to be determined. The results also show that the influence of social capital is stronger in the firms with low Tobin’s Q and stock returns. The social capital related promotions may erode the firm efficiency.
Contents
1. Introduction...3
2. Literature Review ...6
2.1 Social Capital Theory on Status Attainment ...6
2.2 Social Capital on Corporate Operations ...8
2.3 Other Determinants of Promotion ...8
3. Hypotheses...11
4. Methodology...13
4.1 Samples...13
4.2 Measures ...13
5. Empirical Results...18
6. Discussion and Conclusion...23
1. Introduction
Promotion generally refers to an upward movement of an employee to another hierarchy within the organizations. It often involves higher status, higher salaries and more responsibilities. As a core part of human resource management, job promotions have two important purposes, which are incentives and selections. For employees, job promotions are often regarded as measurement of their career success. They perceived the promotions as rewards of their work ability and performance. Compared to other incentives (e.g., money), job promotions are long-term incentives because promotions are based on the employee’ long-term performance, which encourages employees to persist their better performance. The second role of promotion is to allocate human resources. When employees are qualified for higher positions through accumulated working experience and improved personal ability, promotion assigns the employees to the proper positions. The proper match between employees and positions can finally increase the efficiency of the firms (Baker, Jensen and Murphy, 1988). Therefore, the characteristics of the promotion system in an organization have impacts on the employees’ working performance, attitude and behavior, which finally influence the organization’s performance (Ferris, Buckley and Allen, 1992). It is important to build an objective and effective promotion system. Examining the characteristics of the promotion system in the organizations can help us to further understand the different outcomes the promotion procedure generates.
Compared to the low-level employees’ promotion, the senior managers’ promotion is a much more important part in the internal labor market. Senior managers, different from other lower level employees, have overall responsibility for the firms, and the procedure of senior managers’ promotion often
indicates changes of regimes in the firms (Finkelstein & Hambrick, 1996). Therefore, selecting other senior managers may be one of the most important duties of the chief executive officer (CEO) and doing this work well can enhance the firm’s performance. However, the traditional promotion procedure often focuses on individual’s human capital (e.g., education), but ignores the social capital. Do managers who have more social capital receive more promotions? If so, what effects will it generate on the firms? Little Literature has been written about the role of social capital in the promotion systems, while related researches have shown social capital has indeed affect the corporate outcomes, such as internal capital allocation (e.g., Glaser, Lopez-de-Silanes and Sautner, 2013; Duchin and Sosyura, 2013). Therefore, this paper is developed to further understand corporate governance through examining the influence of social capital on the promotions of senior managers.
The main hypothesis in the paper is that the strength of the manager’s tie to the CEO is positively related to the probability of job promotion. Based on the first hypothesis, the second hypothesis further figures out whether CEO power can strengthen the effect of the social capital on the job promotion. These two hypotheses are tested by the OLS regressions of social capital on the promotion rate. Continuously, the third hypothesis is presented to find out whether social capital indirectly influences the firm efficiency through promoting low-ability managers rather than higher-ability managers. Two significant difference tests are constructed to test the hypothesis. One test is built between managers with connections and manager without connections and the other is built between firms with high Tobin’s Q and firms with low Tobin’s Q.
probability of promotion. Besides, among three types of connections (firm connection, education connection, social club connection), social club connection plays the most important role in the promotion process. However, it is not clear that whether strong CEO power could strengthen the effect of social capital on job promotion. Besides, the senior managers with close relationship to CEO seem to have a higher education level and promotion rate, but whether the higher education level indicates higher personal ability is still to be determined. The results also show that the influence of social capital is stronger in the firms with low Tobin’s Q and stock returns. The social capital related promotions may erode the firm efficiency.
The paper is organized as follows: Part II is literature review, which shows the main social capital theories in the labor market and the important determinants in the promotion decisions; Part III is the hypotheses proposed in this paper. Based on the main hypothesis that social connections have positive impact on the promotion probability, the paper further discusses some situations that have moderate effects on the main hypothesis ; Part IV describes the data and methodology; Part V shows the empirical results in the regressions and models; Finally, Part VI summarizes and concludes the whole paper.
2. Literature Review
2.1 Social Capital Theory on Status Attainment
Social capital is generally considered as the investment on social relations. Different from human capital, social capital is created among people and rooted in social networks. Since the concept of social capital was first proposed in the sociology field, it has soon become popular in a wide range of other fields. An increasing number of scholars tried to invoke the concept to solve the questions confronted in their fields
and finally made the study on social capital prosperous. In the organization studies, a series of theories related to social capital is proposed and seems to be powerful explanation of some persons’ relative success in the labor market.
Granovetter (1973) proposed the weak ties theory. He argues that there is information asymmetry in the labor market and social networks offer a channel for information transfer, which matches people to the jobs. In Granovetter’s theory, strong ties typically connect people who have similar characteristics. The more similar people are, the stronger their tie is. In this sense, weak ties are more valuable than strong ties in the labor market because the weak-ties network is more wide ranging and information provided is less redundant. Granovetter (1974) then interviewed 282 men in Newton, Massachusetts and found that social activities influenced the labor markets and people who used interpersonal contacts during job-seeking periods seemed to get better jobs.
Based on these studies, Lin (1982, 1990) proposed the theory of social resources. Resources here are valued goods or materials, which can be divided into personal resources and social resources. Lin argues that weak ties can help people not only to gather efficiency information, but also to access different resources. In a hierarchy social structure, people in the same hierarchy often have higher similarity in the resources such as wealth, status and power, and the social ties among them are strong ties; people in the different hierarchies have less similarity and their social ties are weak ties. When people pursue instrumental goals (e.g., attaining status), weak ties provide a channel for people in the lower hierarchy to reach towards contacts in the higher hierarchies and to get the social resources they want. Therefore, Lin points out the one’s social resources affect the result of his or her instrumental actions and in turn,
are affected by the original position of him or her. Besides, weak ties are more useful than strong ties, considering weak ties are more likely to connect people vertically rather than horizontally.
Although Granovetter and Lin’s theories about the effect of social resources on the status attained are generally supported by the researches in United States and Europe, similar studies in Asia get different conclusion. Through empirical study in China, Bian (1997) found that people who have closer relationships with job-allocation authorities tend to get better jobs, which means people acquired jobs through strong ties rather than weak ties. Thus Bian proposed the strong ties hypothesis. Among social networks, weak ties are used to gather information and strong ties are used to access influence from authority. Individuals can use networks to influence authorities, who in turn assign jobs as favors to their contacts. The stronger the ties are, the more benefits people can get from the helper. Bian’s theory well explained the unique Asian culture in the labor market.
Burt (1992) thinks that the important factor is not the strength of the ties, but people’s position in the networks. He created an innovative theory of structure hole. The structure hole refers to a gap between two unconnected individuals with different resources. When a third person connected these two individuals, the gap is filled. The third person can thus take advantage of exchanging information between the two individuals. In this way, the individual’s social capital can be measured as the number of structure holes he/she occupies.
2.2 Social Capital on Corporate Operations
Besides the influence of social capital on status attainment, social capital has impact on the internal firm operations. Although it is agreed that CEOs are likely to allocate more capital to the connected managers (Glaser, Lopez-de-Silanes and Sautner, 2013; Duchin and Sosyura, 2013), whether the allocation is efficiency still needs discussions. The positive view presents that social capital works as an information channel that helps CEO to make better decisions. Through the better information control, CEO can allocate more capital to the department with more investment opportunities (Stein, 1997). The dark side view regards the social connections as a tool to pursue private benefits. The agency problem can lead to the investment efficiency (Stein, 2003).
2.3 Other Determinants of Promotion
Education is always considered to be related to the promotion probability in many literatures. Nelson and Phelps (1966) presented that education is an investment in human capital, which enhances people’s productivity. However, Spence (1973) and Arrow (1973) viewed education as a signal to indicate the employees’ productivity capability. They argued that individual’s ability is his/her intrinsic characteristics and isn’t affected by education level. However, because of the information asymmetric in the job market, employers have to evaluate the employees’ ability by observing their education level. Spence and Arrow also pointed out that if employers weigh the employees’ education correctly, they can get an unbiased estimation of employees’ productivity capability. Although researchers have different opinions on the function of education, empirical studies have shown that there is a positive relationship between education and promotion. Wise (1973) reported significant effects of college quality, major and grade on the probability of job promotion. It’s not surprising that students with better academic
performance can get higher wages and faster promotion rates. Stolzenberg (1978) further proposed that education has significant influence on promotion in large organizations. In the large organizations, individual’s promotion probability is more related to one’s human capital than work performance. Spilerman and Lunde (1991) also reported that the influences of schooling years and other education measurements vary in different organizations.
Although age and working age are closely related, their influences on promotion are different. Age itself has no relation with promotion-decision, but when it is connected with experience, energy and retirement, it does influence the decision-making. Rosenbaum (1981) analyzed the interaction effect of age and other factors on promotions and argued that the probability of promotion tend to decrease with the increase of age. Education is an important determinant of promotions, but its influence decreases with the increasing working age. Because most employees complete their education before work, education level is an indicator of ability for new employees. When the age increases, employees have more working experience, which is more valuable to the firms than education. Therefore, “age” has a positive effect on the promotion probability. But with the further increase of age, the employees’ energy decreased, which reduces the promotion opportunity. Sandefur (1981) regarded working experience as both general resources and organization-specific resources. These two kinds of resources have positive effects on promotions within organization boundaries. But organization-specific resources have negative effect on promotions across organization boundaries. Rosenbaum (1984) presented that early attainments and promotions have strong impacts on the later promotions. If an individual is initially at a low status job and the promotion rate is initially slow, the individual will be unable to get any significant promotion. Stein (1994) proposed that the ability of employees is determined by their experience and
position in the organizations. Therefore, elder employees have a comparative advantage in high positions. But young employees have higher probability of promotion.
The gender promotion difference is widely discussed in the world. A large number of literatures and evidences have shown that women are placed at a disadvantage place in the job promotion competition. Through the data of three large fiduciary institutions, Cabral, Marianne and Green (1981) found that with comparable qualities, women are paid less than men and have relatively lower promotion probabilities. They explained these findings as sex discriminations. Similarly, Spurr (1990) also found women were less likely than men to achieve partnership in major U.S. law firms. Ward (2001) analyzed the data of five universities in Scotland. Trough the comparisons of the promotion of male and female staff, he found women have less chance to be promoted even with the same personal characteristics (e.g., publications, experience, education) of men. Based on the data of Austrian micro census, Winter-Ebmer and Zweimuller (1997) argued that the promotion standards of women are higher than that of men, which can not be explained by the productivity difference.
3. Hypotheses
Recently, there is a strong tendency that CEOs are likely to rely on the social connections to make important corporate decisions (e.g., Glaser, Lopez-de-Silanes and Sautner, 2013; Duchin and Sosyura, 2013). In a further consideration on the internal promotion process, as the most powerful person in the organization, CEO is also likely to use the social networks information to make promotion decisions. And since the ultimate responsibility for promotion decision of senior managers rests with the CEO, the
tie to CEO may be the most useful and valuable tie in the promotion process . This paper will focus on the linkage between senior managers and CEOs. Senior manager’s social capital also mainly refers to the social relations with CEO.
Social networks can either act as a channel for information transfer (Granovetter, 1973; Lin, 1982; Burt, 1992), or act as a bridge for private benefit pursuing (Bian, 1997). In the information channel views, the CEO has imperfect information about all the candidates because of information asymmetry. However, the CEO tries to promote the candidate with high ability in an effort to enhance the firm performance. Through social networks, the CEO has relatively more information about the closer candidates, which helps the CEO to make better promotion decisions. In this way, it seems that the CEO always likes to promote the closer managers. In the private benefit pursuing views, the CEO attempts to promote the more connected managers for the purpose of self-benefitting. In both views, despite the purposes, CEOs are more likely to promote the senior managers who have closer relation to them. Besides, in the process of making promotion decisions, the CEO has to consider all the information of candidates. However, the situation that the candidates have equivalent ability or their backgrounds are ambiguous always exists. Under this scenario, it’s not strange that CEO prefer to promote the social connected manager because the social connections often build trust and comfort between each other. Thus, it is hypothesized that:
Hypothesis 1: The strength of the manager’s tie to the CEO is positively related to the probability of job promotion.
also indirectly influence the promotion decisions. However, a powerful CEO is always capable of directly controlling the promotion results and less likely to receive other’s opinions. Therefore, when there is a very powerful CEO in the organization, the other important persons’ indirect influence on the promotion decision is weakened and the social tie to CEO significantly overweighs the ties to other important person. What’s more, powerful CEOs seek to maintain their control by selecting board members who are demographically similar. Those board members can therefore be more sympathetic to them (Westphal and Zajac, 1995). Further, powerful CEOs may be likely to exert their power through promoting the social connected senior managers who are considered to be loyalty to the CEOs. Therefore, it is hypothesized that when the power of CEO increases, the effect of social ties on the probability of promotions is strengthened.
Hypothesis 2: The CEO’s power can strengthen the effect of social ties on the probability of promotions.
The effect of social networks on the firm performance is ambiguous. The reason that CEOs tend to promote the more connected managers could either be pursuing private interests or be in an effort to maximize firm value. In the former reason, there is no doubt that the promoted manager with stronger ties to CEO would have relatively lower ability than other promoted manager. And in the latter reason, those promoted managers with stronger ties are considered to have relatively higher ability.
Hypothesis 3: The promoted managers with closer ties to CEO have higher ability than other promoted managers.
4. Methodology
4.1 Samples
The sample used in this paper is drawn from BoardEx and Compustat, including approximately 260 representative S&P 500 firms and 3500 senior managers. The term senior manager is defined as vice president, division president, regional president, division vice president, regional vice president and so on. The term social capital is defined as the linkage between senior managers and CEOs. In the sample, 82.1% senior managers are male. The average age of senior managers is 53.8 years. 95.2% CEOs are male. The CEOs have approximately 7.7 years tenure, with an average age of 57.7 years.
4.2 Measures
As the dependent variable, the probability of job promotion for senior managers cannot be observed. However, it is easily understood that if an individual has a higher probability of getting promoted, he/she will have a relatively higher promotion speed (more promotion times during fixed years) than others. The higher the promotion rate is, the higher probability of job promotions is. This suggests that the promotion rate could be a good measure of the probability of promotion. Because some senior managers maybe stay longer than the current CEO in the primary firm, the senior managers’ former promotions cannot be attributed to the current CEO and the overall promotion rate is not applicable in this case. In order to solve the problem, this paper uses the promotion rate during the current CEO’s tenure as a measure of the probability of job promotion. Therefore, the promotion rate is calculated as:
The most important independent variable is the senior manager’s social capital, which is the linkage between the senior manager and the CEO. One consideration arisen in the measurement of social capital is the potential reverse causality problem. That is, senior managers who get promotions can further develop connections with the CEO. Duchin and Sosyura (2013) address the similar problem by excluding the connections formed during senior managers’ tenure at the current firm. However, if the paper excludes the connections formed during the current firm, few managers will have connections with the CEOs, which can not support the data analysis. Therefore, the paper uses another less restrict solution, which is to only exclude the connection built during the current CEO tenure because the paper only examine the senior managers’ promotion rate during the CEO tenure. As for the detailed measurement of social capital, Bourdieu (1986) presented that social capital can be represented by aggregating the size of networks and volume of capital obtained by network members. However, this representation neglects the tie strength among members and assumes each member has the equivalent tie strength and relationship. In this paper, because the only tie needed to be measured is the tie to CEO and CEOs are considered to have equivalent amount of capital, the senior manager’s social capital is measured by the numbers of tie to CEO and each tie’s strength. The tie strength is quantified by the overlap years that CEO and senior managers build a relationship through the tie. And the linkage between senior managers and CEOs are defined as several types: education, social clubs, previous employments and so on. Therefore, the senior manager’s social capital is measured as:
Social Capital = ∑ tie × overlap years
the paper also constructs each type of social connections as dummy variables. The social connections are mainly defined as three categories: firm connection, education connection, and social club connection. Each type of connection is encoded to 1 if managers have such kind of connection with CEO.
The moderate variable is CEO power. The power of CEO comes from many sources, such as structure power, ownership power, expert power, and prestige power (Finkelstein, 1992). Thus it is difficult to measure the CEO power from just one aspect. However, it is generally agreed that the CEO power is closely related to CEO tenure (Hill and Phan, 1991), CEO compensation (Boyd, 1994) and CEO ownership (Finkelstein, 1992). Therefore, this paper tries to construct a principle component analysis on those factors (CEO tenure, CEO ownership, CEO compensation) and extract the representative CEO power measurement. Table 1 shows the result of factor analysis. The eigenvalue of main factor is less than 1 and the value KMO (Kaiser-Mayer-Olkin) test is only 0.51, which indicate factor analysis is not an ideal method to extract the CEO power measurement from CEO tenure, CEO ownership and CEO compensation. Because the principal component analysis is not applicable, the paper is going to use the CEO tenure, CEO ownership and CEO compensation as proxies for CEO power.
Table 1 Results of Factor Analysis
Variable CEO Power
Tenure 0.48 Ownership 0.48 Compensation 0.12 Eigenvalue 0.47 % of Variance 197.45 Cum. % of Variance 197.45
male and 0 if the senior manager is female. Rosenbaum (1981) reported the effect of age on probability of job promotion is like an inverted U shape, which means the probability of promotion tend to increase in the early periods and decrease in the late periods. However, since the senior managers all have a relatively high age. The age effect is more likely to be a linear relationship. Thus, the paper only use age as control variable, excluding age squared. The education level is measured by the number of universities or colleges that senior managers have attended, based on the concept that if the manager have attended one university, he maybe have a bachelor degree and if he have attended two or more universities, he maybe have a master or higher degree.
Table 2 Means, Standard Deviations and Correlations
Club Variable Mean Std. Dev. Promotion rate Social Capital Firm Connection Education Connection Connection Gend er Age Educati on CEO tenure CEO ownership Promotion rate 0.30 0.35 Social Capital 4.77 7.63 0.06 Firm Connection 0.66 0.47 0.11 0.29 Education Connection 0.00 0.05 0.01 0.14 0.06 Club Connection 0.00 0.07 0.12 0.11 0.07 -0.01 Gender 0.82 0.38 0.07 0.04 -0.02 0.04 0.00 Age 53.78 6.35 -0.08 0.10 0.07 -0.03 0.00 0.03 Education 1.50 0.80 0.03 0.03 0.10 -0.03 0.07 -0.07 0.04 CEO tenure 7.80 4.44 -0.14 -0.03 0.01 -0.01 -0.02 -0.02 0.01 -0.03 CEO ownership 0.80 3.06 -0.06 0.06 0.06 -0.02 -0.02 -0.01 -0.01 0.04 0.45 CEO pay 9549.55 8726.98 0.01 0.06 0.06 0.03 0.12 -0.04 -0.04 0.12 0.10 0.02
Table 3 OLS Regressions of Social Capital on Promotion Rate (×100)
5. Empirical Results
Table 3 presents the results of the OLS regressions, which identifies the role of social capital in the promotion process. Model 1 is a simple regression of social capital on the promotion rate. Then control variables and moderate variables are added by turns in model 2-8.
28.81*** 27.63*** 51.18*** 52.60*** 63.55*** 53.50*** 55.42*** 65.07*** Intercept (0.70) (1.45) (7.62) (12.59) (12.80) (12.60) (12.81) (13.00) 0.27*** 0.27** 0.17✝ 0.25✝ -0.08 0.26 0.03 -0.28 Social Capital (0.08) (0.08) (0.1) (0.14) (0.30) (0.15) (0.19) (0.32) 1.45 2.95 8.11* 7.77* 8.05* 8.00* 7.71* Gender (1.56) (2.5) (3.83) (3.79) (3.82) (3.82) (3.80) -0.38** -0.54* -0.52* -0.54* -0.55* -0.52* Age (0.14) (0.22) (0.22) (0.22) (0.22) (0.22) 1.69 1.48 1.80 1.75 1.47 Education (1.59) (1.58) (1.59) (1.60) (1.59) -1.46*** -1.51*** CEO Tenure (0.38) (0.43) 0.04 0.05
Tenure × Social Capital
(0.03) (0.04) -1.010 0.28 CEO ownership (0.68) (0.77) 0.01 -0.03 Ownership × Social Capital (0.04) (0.05) -0.25 -0. 14 CEO Compensation (/1000) (0.22) (0.22) 0. 02✝ 0. 02 Compensation (/1000) × Social Capital (0.01) (0.01) R-squared 0.0036 0.0038 0.0058 0.0163 0.0363 0.0209 0.0199 0.0394 Ajusted R-squared 0.0033 0.0032 0.0042 0.0114 0.0291 0.0136 0.0126 0.0274 Note: significant results are indicated by *** p<0.001, ** p<0.01, * p<0.05, ✝p<0.1
Models 1-4 indicate a positive relationship between social capital and the promotion rate, which strongly support hypothesis 1. Managers with strong social ties to CEO are at an advantage place in the promotion competitions.
However, the moderating effect of CEO power is insignificant, which doesn’t support hypothesis 2. The possible conclusion is that CEO power does not have impact on the strength of the effect of senior managers’ social capital on the job promotion. However, CEO tenure, CEO ownership and CEO compensation are possibly not perfect proxies of CEO power since the power of CEO comes from many sources, such as structure power, ownership power, expert power, and prestige power (Finkelstein, 1992). One combined measurement of CEO power maybe work much better. The above drawbacks of model-setting can result in misleading conclusions. Therefore, further study needs to be done to the measurement of CEO power in order to get a more convincing conclusion of the moderating effect of CEO power.
The models present that gender and education are positively associated with the probability of job promotion, while age has a negative effect on the likelihood of promotion. The results are consistent with previous literatures. However, the education variable is not significant as expected. One possible explanation is that education is a less important determinant on the high-level promotions than on the low-level promotions. The influence of education on promotion probability decreases with the increasing working age (Rosenbaum, 1981). In addition, because education is used as a proxy of ability by employers when the information is asymmetric (Spence, 1973; Arrow, 1973), when it comes to high-level promotion, CEOs usually have already had some basic understanding of the candidates and do not
need to rely on the education to estimate their abilities. Another possible explanation of the
insignificance of education is the data inaccuracy. The education data is extracted from the senior managers’ network, based on the concept that if the senior manager is connected with someone via a university, he should have attended the university. Such extraction leads to a large amount of missing data and can not guarantee the managers’ real education level. Therefore, the inaccuracy of data may
result in the insignificance of the education variable.
Table 4 OLS Regressions of Social Connections on Promotion Rate (×100)
Variable Model 9 Model 10 Model 11 Model 12 Model 13
25.61*** 24.22*** 46.14*** 48.81*** 59.63*** Intercept (1.03) (1.63) (7.63) (12.51) (12.85) 6.51*** 6.49*** 7.50*** 8.74** 8.20** Firm Connection (1.26) (1.26) (1.83) (2.87) (2.93) 3.30 3.17 1.98 0.30 -2.22 Education Connection (11.06) (11.06) (14.19) (14.20) (14.20) 36.23*** 36.36*** 46.24*** 41.01** 37.48** Club Connection (8.75) (8.75) (10.86) (13.30) (13.77) 1.71 3.16 8.41* 7.94* Gender (1.55) (2.48) (3.79) (3.77) -0.37** -0.54* -0.55* Age (0.14) (0.22) (0.22) 0.95 0.80 Education (1.58) (1.59) -1.29*** CEO Tenure (0.38) 0.02 Tenure × Social Capital
(0.02) 0.05 CEO ownership
(0.75) -0.02 Ownership × Social Capital
(0.05) -57.32 CEO Compensation (/1000) (21.26) 0.00 Compensation (/1000) × Social Capital (0.01) R-squared 0.0132 0.0136 0.0236 0.0365 0.0571 Adjusted R-squared 0.0124 0.0124 0.0210 0.0293 0.0429 Note: significant results are indicated by *** p<0.001, ** p<0.01, * p<0.05, ✝p<0.1
In order to further identify the influence of each type of social connections on the promotion probability, the paper then replace the social capital variable by three dummy variables, each represents one kind of social connections. Table 4 shows the results of the regressions.
The coefficients of firm connection and social club connection are significant while the coefficient of education connection is insignificant. What’s more, the coefficient of social club connection is much higher than that of firm connection. It shows that among all types of connections, social club connection is the most important connections that help to increase the managers’ promotion probability. It is easily understood that CEO can know managers’ ability better through firm connections and education connections. But whether CEO can get managers’ ability information from social club connections is an unknown question. From this side, the fact that social club connection plays the most important role in the promotion process seems to be a bad signal that CEO is more likely to promote closely related managers at the motivation of pursuing private benefit.
The coefficient and significance of other variables in model 9-13 do not have significant difference from those in model 1-8. The results from table 4 still support hypothesis 1 and do not support hypothesis 2.
Table 5 t-Test for the Significant Difference of Managers
Group Gender Age Education Promotion Probability
Managers without connections(=0) 0.81 54.05 1.42 0.28
Managers with connections(>5) 0.84 54.01 1.58 0.32
Difference -0.03* 0.04 -0.16** -0.04**
Table 5 shows the differences between managers without connections to CEO and managers with close relation to CEO. It turns out that senior managers who are closer to CEOs are more tend to have a high education level and promotion probability.
It is very interesting to see that managers with more connection are more likely to be male. This maybe resulted from the high percentage of male CEO (95.2%), which makes male managers at an advantage place for getting more connections to CEO. The research of Singh, Kumra and Vinnicombe (2002) also gives a possible explanation of the phenomenon. They argued that although women managers have recognized that social network is a useful tool, they reject to use it and prefer to rely on high performance to get a promotion.
Managers with stronger ties to CEO have comparatively higher education level and higher promotion rate. Spence (1973) and Arrow (1973) presented that if employers weigh the employees’ education correctly, they can get an unbiased estimation of employees’ productivity capability. Therefore, if the education is an unbiased estimation of senior managers’ ability, hypothesis 3 is supported. However, there is no evidence proving that education is a perfect proxy for ability and thus hypothesis 3 is not very convincing. One reason for the positive relation between social ties and education level is that CEOs have relatively higher education level and they are more likely to build connections with people who have similarities with them. Then CEOs with higher education level should have more connections with managers with higher education level.
Table 6 t-Test for the Significant Difference of Firms
Firms with low Tobin's Q (<=1.5) -0.02 4.98 Firms with high Tobin's Q (>1.5) 0.13 4.53
Difference -0.15*** -0.45**
Note: significant difference results are indicated by *** p<0.001, ** p<0.01, * p<0.05
Table 6 shows the significant difference test between firms with low Tobin’s Q and firms with high Tobin’s Q. The social capital here presents the average social capital of senior manager in each firm. The results in table 6 indicate that firms with low Tobin’s Q have relatively low stock returns and strong social relations. It seems that the fact that CEOs are likely to promote closely related managers has bad effects on the firm performance. Hypothesis 3 is rejected. The promoted managers with closer ties to CEO have lower abilities than other managers. Besides, the motivation of CEOs promoting closer managers is more likely to be private benefit pursuing.
6. Discussion and Conclusion
This paper analyzed the effect of senior managers’ social tie to CEO on their probability of job promotion. The results indicate that managers having a close relationship with the CEO tend to get a high promotion probability. Besides, among all three types of connections, the social club connection plays the most important role in increasing managers’ promotion probability, which release a bad signal that CEOs are more likely to promote closely related managers at the motivation of pursuing private benefit.
There are no strong supports for hypothesis 2, predicting that CEO power didn’t have moderating effect on the influence of social networks on the job promotion. Because CEO tenure, CEO ownership and
CEO compensation may be poor proxies for CEO power, further studies can be done for this aspect.
The paper also did a significant difference test between managers with connections and managers without connections. Senior managers who are closer to CEOs are more tend to have a high education level and promotion probability. Although the managers’ education level can represent their ability to some extent (Spence, 1973; Arrow, 1973), further studies need to be done to examine hypothesis 3. Because of the limited data, this paper can’t do the difference significant test of managers’ ability. However, if the ability can be measured by the managers’ past working performance, the difference significant test between managers with connections and managers without connections can directly show that whether the close related managers have a relative high ability level. In this way, the paper can find out the influence of social connections on the firm efficiency and finally build measures towards social capital in the corporate governance.
Another significant difference test shows that firms with low Tobin’s Q have relatively low stock returns and strong social relations. It seems that the promoted managers with closer ties to CEO have lower ability than other promoted managers. Such promotions are more likely to derive from private benefit pursuing motivation.
Finally, there are still some problems in this paper. One of the most important problems should be the omitted variables, which lead to a quite low R-squared in the models presented in table 3 and 4. For example, the managers’ ability may be the most important determinant in the promotion process. However, the paper can not observe the senior managers’ ability directly due to the data availability. If it
is possible, the senior managers’ productivity can be measured by their former job performance or their division sales, which needs further data collecting.
Another problem is the measurement of CEO power. The power of CEO comes from many sources, such as structure power, ownership power, expert power, and prestige power (Finkelstein, 1992). But this paper doesn’t use a combined measurement of CEO power, which can result in misleading results.
The significant difference tests in the paper only give a quick look at the correlations. Further causality demonstration can be done, which needs to control a bunch of other variables.
References
Arrow, Kenneth J. 1973. "Higher Education as a Filter." Journal of Public Economics 2: 193-216. Baker, George P., Michael C. Jensen, and Kevin J. Murphy. 1988. "Compensation and incentives:
Practice vs. theory." The journal of Finance 43.3: 593-616.
Bian, Yanjie. 1997. "Bringing Strong Ties Back in: Indirect Ties, Network Bridges, and Job Searches in China." American Sociological Review 62: 366-85.
Bourdieu, Pierre. 1986. “The forms of Capital.” Handbook of Theory and Research for the Sociology of
Education, New York: Greenwood.
Boyd, Brian K. 1994. "Board control and CEO compensation." Strategic Management Journal 15.5: 335-344.
Burt, Ronald E. 1992. Structural Holes: The Social Structure of Competition. Cambridge, MA: Harvard University Press.
Cabral, Robert, F. Marianne, and C. Green. 1981. “Men and Women in Fiduciary Institutions: A Study of Sex Differences in Career Development.” Review of Economics and Statistics 63: 573–80.
Duchin, Ran, and Denis Sosyura. 2013. “Divisional Managers and Internal Capital Markets.” Journal of
Finance 68: 387-429.
Ferris, G. R., Buckley, M. R., and Allen, G. M. 1992. “Promotion systems in organizations.” Human
Resource Planning 15: 47-68.
Finkelstein, Sydney. 1992. "Power in top management teams: Dimensions, measurement, and validation." Academy of Management Journal 35.3: 505-538.
Finkelstein, S., and D.C. Hambrick. 1996. Strategic Leadership: Top Executives and Their Effects on
Organizations. West Publishing Company.
Glaser, Markus, Florencio Lopez-de-Silanes, and Zacharias Sautner. 2013. “Opening the black box: internal capital markets and managerial power in conglomerates.” Journal of Finance 68(4): 1577-1631
Gompers, Paul A., Joy L. Ishii, and AndrewMetrick, 2003. “Corporate governance and equity prices.”
Quarterly Journal of Economics 118: 107–155.
Granovetter, Mark. 1973. "The Strength of Weak Ties." American Journal of Sodology 78:1360-80. -. 1974. Getting a Job: A Study of Contacts and Careers. Harvard University Press.
Hill, Charles WL, and Phillip Phan. 1991. "CEO tenure as a determinant of CEO pay." Academy of
Management Journal 34.3: 707-717.
Lin, Nan. 1982. “Social Resources and Instrumental Action." In Social Structure and Network Analysis, edited by P. Marsden and N. Un. Pp. 131-47. Sage Publications.
-. 1990. "Social Resources and Social Mobility- A Structural Theory of Status Attainment.” Pp. 247-71 in Social Mobility and Social Structure, edited by Ronald Breiger. Cambridge University Press.
Rosenbaum, James E. 1981. “Careers in a corporate hierarchy: a longitudinal analysis of earnings and level attainments.” Research in Social Stratification and Mobility 1: 95-124.
-. 1984. Career Mobility in a Corporate Hierarchy. Orlando, Florida: Academic Press, Inc.
Sandefur, Gray D. 1981. “Organizational boundaries and upward job shifts.” Social Science Research 10: 67-82.
Singh, Val, Savita Kumra, and Susan Vinnicombe. 2002. "Gender and impression management: Playing the promotion game." Journal of Business Ethics 37.1: 77-89.
Spence, Michael. 1973. "Job Market Signalling." Quarterly Journal of Economics 87: 355-74.
Spilerman, Seymour, and Tormod Lunde. 1991. “Features of Education Attainment and Job Promotion Prospects.” American Journal of Sociology 97(3): 689-720.
Spurr, Stephen J. 1990. “Sex Discrimination in the Legal Profession: A Study of Promotion.” Industrial
and Labor Relations Review 43(4):406–17.
Stein, Jeremy C. 1997. “ Internal capital markets and the competition for corporate resources.” Journal
of Finance 52: 111–133.
-. 2003. “Agency, information and corporate investment.” in George M. Constantinides, Milton Harris, and Ren´e M. Stulz, eds.: Handbook of the Economics of Finance (Elsevier, Amsterdam).
Stein, S. 1994. “Ability, Promotion and Optimal Retirement.” Journal of Labor Economics 12: 119-137. Stolzenberg, R. M. 1978. “Bringing the boss back in: Employer size, employee schooling, and
socioeconomic achievement.” American Sociological Review: 813-828.
Ward, Melanie E. 2001. “Gender and Promotion in the Academic Profession.” Scottish Journal of
Political Economy 48: 283-302.
Westphal, James D., and Edward J. Zajac. 1995. "Who shall govern? CEO/board power, demographic similarity, and new director selection." Administrative Science Quarterly: 60-83.
Winter-Ebmer, Rudolf, and Josef Zweimuller. 1997. “Unequal Assignment and Unequal Promotion in Job Ladders.” Journal of Labor Economics 15(1):43–71.
Wise, David A. 1975. “Personal Attributes, Job Performance and Probability of Promotion.”
