Ownership Concentration and
Corporate Social Performance: An
Empirical Analysis
Onno Huyghe
Master Business Administration, Finance, University of Groningen, Groningen
Summary
This paper analyses the effect of ownership concentration on corporate social performance based on firm level data of 691 European Multinational Enterprises (MNE's). It tests whether firms with different types of blockholders have different corporate social performance and whether ownership concentration is related to different types of corporate social performance. It finds strong empirical support for the hypothesis that shareholder concentration leads to worse corporate social performance.
By: Onno Huyghe
Thesis Coordinator: Prof. dr. B.W. Lensink Institution: Rijksuniversiteit Groningen
Faculty of Economics and Business
Table of Content
Summary ... 1
Table of Content ... 2
List of Abbreviations ... 2
1. Introduction... 3
2. Corporate social responsibility ... 4
3. Ownership concentration ... 7
Ownership concentration - agency theory ... 8
Financial performance - ownership concentration... 9
Corporate social performance - financial performance ... 9
Large shareholders - corporate social performance ... 10
4. Theoretical model... 13
5. Hypothesis ... 16
6. Descriptive statistics... 17
Corporate social performance... 17
Large shareholders ... 18
Control variables ... 22
7. Correlations ... 24
Shareholder data correlations ... 24
CSR data correlations... 25
Correlations between CSP and types of shareholders ... 26
8. Regressions ... 27
9. Results... 30
Corporate social performance and measures of shareholder concentration ... 32
Type of shareholders and corporate social performance... 34
10. Conclusion... 37
11. References ... 39
Appendix A... 43
Appendix B... 44
List of Abbreviations
MNE MultiNational Enterprise
CSR Corporate Social Responsibility
CSP Corporate Social Performance
CFP Corporate Financial Performance
1.
Introduction
Publicly traded corporations can not choose who becomes a large shareholder, but some might wish that they could. Over the last 10 years, shareholder influence seems to have grown despite management resistance and their influence extents further than the annual meeting of shareholders. Recent research (Fahlenbrach & Cronqvist, 2006) has shown that large shareholders influence corporate profitability and policies. The question remains whether this interference also has a social side.
This paper analyses whether corporate social performance is influenced by the level of ownership concentration. Do large shareholders have an influence on a company’s social performance? Do large shareholders use their influence on the business to affect corporate social performance or does it not matter to them? Or do only certain types of investors stimulate or deter corporate social performance?
2.
Corporate social responsibility
In this section the literature on corporate social responsibility and performance will be reviewed. To understand what the variable corporate social performance in this research is, it is useful to investigate the history of corporate social responsibility and performance. The history of corporate social responsibility is short but confusing. There has often been debate about what corporate social responsibility actually is and how it can be operationalized.
Over the last century a lot of research has been done in the field of Corporate Social Responsibility. Corporate Social Responsibility (CSR) has in fact been around as a scholar term for over 50 years. However, since the beginning of the industrial revolution certain companies have had social policies to help the community and especially their workers. Since the 1950’s, (corporate) social responsibility has been a topic of formal writing. Social Responsibilities of the Businessman (Bowen, 1953) is widely recognized as one of the first influential writings on corporate social responsibility.
short term) value for the company. The results of this research are therefore useful in adressing the answer to this large-scale discussion.
As mentioned before, corporate social responsibility and corporate social responsibility has been defined and redefined many times. Carroll (1991) produced an influential paper in which he divided corporate social responsibility in 4 aspects: Philanthropic, Ethical, Legal and Economic. To use corporate social responsibility correctly as a variable in this research, this paper will now look at how it can be used as a multidimensional construct.
Corporate social responsibility is a vaguely defined construct but treating it as a multidimensional construct has proven useful to describe all aspects of it (Carroll, 1999). Zarha and LaTour (1987) already identified different dimensions in 1987. Recent empirical research has shown that treating CSR as one uniform construct is not correct. Pirsch et al (2007) suggest that different CSR initiatives can yield different results. They state that a company can actively pursue short-term results by selecting certain CSR initiatives. Shareholders can be more interested in certain CSR initiatives than in others. This is interesting for this research. This research will therefore identify various different aspects of CSR and expects that the results will differ for the different dimensions that were identified. The unifying approach of all CSR initiatives in one contruct has received much criticism. According to Griffon and Mahon (1997) the studies do not capture firms’ social performance as they “inadequately reflect the breadth of the construct” and thus lack an appropriate level of validity.
In this research corporate social responsibility is sometimes used interchangeably with corporate social performance. This is because various researches use different terms to identify the same thing. This research considers corporate social performance to be the concrete actions and results that follow from corporate social responsibility initiatives. Corporate social performance is an extension of theoretical thinking on CSR (Carroll, 1999, Wartick and Cochran, 1985) and allows for more empirical research.
Now that an overview of the literature on corporate social responsibility has been given, it is useful to step back and allow for a look at how this can be useful for this research. The literature on defining corporate social responsibility/performance has important implications for this research. It has proven vital to measure corporate social performance correctly. This is attempted by the scoring technique used by the Ethical Investment Research Service, EIRIS. It is important to use this EIRIS information correctly in this research, by applying the right aggregation techniques. By applying the same factor analysis as other widely appreciated studies that use EIRIS, this research attempts to do this. It is also important to keep in mind that different means of operationalizing CSR can yield different results.
3.
Ownership concentration
This paper is about corporate social responsibility and ownership concentration. In the previous section corporate social responsibility has been discussed. This section will further explore the literature related to ownership concentration. Conceptually this is what we are going to investigate.
Figure 1 Conceptual Model of Research Question
The question is whether concentrated ownership has an influence on corporate social performance. This influence can be both positive and negative or not present. Since there is very little research in this field, a few areas of research that hold some common ground with this research area are explored in order to get some insight into what can be expected from the regressions. In this literature section the agency problems between large shareholders and management will be discussed. This will give us insights in the possible dynamics of the hypothesized C arrow which describes the influence blockholders have on corporate social performance. Furthermore, the relationship between financial performance of the company and the presence of large shareholders is discussed in order to give more insight in the A arrow and to understand the interests of large shareholders. Then the link between financial performance and corporate social performance is discussed. If large shareholders prefer superior profits, this area of reserach will provide an insight in the sign of the relationship between
large shareholders and corporate social performance in this research. Finally, the C arrow, the link between ownership concentration and corporate social performance will be discussed, by reviewing the few studies that have been conducted in this field.
Ownership concentration - agency theory
In this subsection the potential agency problems between large shareholders, small shareholders and management with regard to corporate social responsibility issues will be investigated. This is essential to the research since the presence of these agency problems can be the explanation of the outcome of the empirical research. Therefore, the rationality developed in this subsection is particularly vital to the research.
firm. Therefore, this research will investigate whether large shareholders, or blockholders, influence the corporate social performance of companies.
Financial performance - ownership concentration
In order for us to understand the (possible) effect of blockholders on corporate social performance it is vital to understand the larger dynamics by studying the research field on ownership concentration in relation to corporate performance in general. This is important because there is little research on the specific link between ownership concentration and corporate social performance. Dominating this research is the financial performance- stock ownership link.
Large shareholders (blockholders) have more influence on the company than dispersed shareholders. For example, Gorton and Schmid (2000) show that German bank-type shareholders improve performance. There is an abundance of literature on this subject. Studies done on Anglo-Saxon companies show profitability of a company is enhanced by the presence of large shareholders (Stiglitz, 1985; Shleifer and Vishny, 1986). Studies on continental Europe companies show a more mixed view indicating that the presence of large shareholders does not necessarily enhance shareholder value. (Lehmann and Weigand, 2000). According to various sources a trade-off between control and liquidity exists (e.g. Pagano and Roell, 1998, Bolton and von Thadden 1998). The results are mixed but significant in most cases; shareholders seem to have an influence on financial performance of a company, that is contingent or situation dependable. For this research this is important because it shows that large shareholders have an influence on a company
Corporate social performance - financial performance
In this subsection CSP will be linked to corporate financial performance. This is a well researched area and the insights provided by these researches may shed some light on the topic of this paper; the link between ownership concentration and corporate social performance. If a clear link does exist between CSP and CFP, profit-oriented shareholders will act upon that knowledge. That is why this subsection will shortly review the knowledge in this area.
between shareholders and corporate social performance. In the aggregate, the view has persisted that results are inconclusive regarding this subject. (Jones and Wicks, 1999, McWilliams and Siegel, 2000, Roman et al. 1999).Seifert, Morris and Bartkus (2004) find evidence that corporate philanthropy is a discretionary social responsibility and they find evidence for the traditional thinking about corporate social responsibility—that doing well enables doing good. The findings imply no significant effect on profits from corporate generosity. Orlitzky (2003) on the contrary, in an extensive meta-analysis, of all known studies, found a bidirectional and simultaneously positive relationship between CSR and financial performance. Margolis and Walsh (2001, p.10), in their survey paper on the link between corporate social and financial performance find that; “corporate social performance is found to have a positive relationship to financial performance in 42 studies (53%), no relationship in 19 studies (24%), a negative relationship in 4 studies (5%), and a mixed relationship in 15 studies (19%).”
Overall, the conclusion by Ullman, 1985, that it is “data in search of a theory” remains an interesting statement. It does seem to be the case that slack resources and higher profitability lead to more corporate social initiatives. The conclusion of this short overview is that there is no clear sign in the relationship between CSP and CFP as there are many researches that find a positive relationship but also many researches that find a negative relationship. It is a shame for this research that academic progress in this area has not resulted in clear conclusions because if they would have been developed the results would also direct this research. We assume that large shareholders are profit maximizers. Therefore a positive relationship in this field would also mean positive relationship expectancy for this research and vice versa. More research exists on the CSP-CFP relationship and the effect of ownership structure on firm performance than research in the field of ownership structure on corporate social performance. The next section will deal with this issue.
Large shareholders - corporate social performance
In 1988, Atkinson and Galaskiewicz found evidence that company giving is influenced by the percentage of stock held by the CEO. They only investigate stockholdings by CEO’s and attribute their findings to the same hypothesis as proposed in this paper. Barnea and Rubin (2006) is one of the only researches directly investigating the link between CSP and ownership. They employ a data set that categorizes the largest 3,000 U.S. corporations as either socially responsible or socially irresponsible. They find that insiders’ ownership and leverage are negatively related to the firm’s social rating, while institutional ownership is uncorrelated with it. The weakness of their research is in the corporate social responsibility data. By characterizing a company as either socially responsible or not, their findings are not as sophisticated as the researches based on more sophisticated databases such as EIRIS and KLD.
suggest that it is necessary to consider different types of institutional investors to more precisely examine their different impacts. Bartkus, Morris and Seifert (2002) find strong evidence that powerful owners discourage excessive philanthropy. They investigate the U.S. market and determine blockholders as >5% ownership of shares. However, their sample is limited to a mere 66 companies. Brammer and Millington (2005) offer partial support for the hypothesis that agency conflicts play a role by showing a significantly lower propensity to become involved in charitable giving among firms with highly concentrated ownership (<50%). The definition of highly concentrated ownership is unusual in literature on blockholdings, as it usually described as 5%, 10% or 20%. Rahbein, Waddock and Graves (2004) study shareholder activism in the field of social performance. Their empirical evidence suggests that shareholder activists are submitting social-policy resolutions with the intention to alter the social behavior of poor corporate performers. However, they conclude in the end that: “…taken as a whole, perhaps philanthropy is an issue in which many active investors (i.e., institutions and blockholders) may prefer to make individual decisions about charitable giving and recipients…”.
Overall, the research in the field of large shareholders in relationship to corporate social performance is marginal. The studies conducted to date show that large (institutional) shareholders tend to restrict corporate social spending. If corporate social performance comes at a cost that is not returned in extra profit, this view is in line with the shareholder maximization view which was defined earlier in the paper. Especially the literature with regard to manager shareholdings which lead to smaller CSR investments show that there might be an agency conflict between managers and shareholders with regard to corporate philanthropy and corporate social responsibility. Managers want to increase their social status by giving corporate slack resources to social initiatives whereas most company shareholders are more interested in a higher return. However, treating all groups of large shareholders as equal might do injustice to the complex motivations and the diversity of large shareholders.
4.
Theoretical model
Based on the literature review, it is necessary to develop more theory on the relationship between large shareholders and corporate social performance. In this section a visual explanation of the investigated relationships is presented as well as a theoretical model. The section explains the rationale behind investigating this area and lays the groundwork for the proposed hypothesis in the next section.
Figure 2 Relationships between the variables
The figure above describes the relationships between the variables in this research. It describes that shareholder concentration can be present due to different types of shareholders and that corporate social responsibility is a multidimensional construct that is composed of stakeholders, governance and
environment. Based on the literature review and on general economic theory this research will now hypothesize what the influence between shareholder concentration and corporate social responsibility might be.
Corporate social performance as a single construct encompasses many aspects. It benefits many or even all stakeholders in the company. Shareholders are only one of the stakeholders in the company. Assuming that corporate social performance comes at an initial cost, the shareholders pay for corporate social performance. Other stakeholders do not directly pay for corporate social performance. The benefits of corporate social performance are for all stakeholders and not only for the shareholder. If the benefits, whatever they may be, outweigh the costs for the company, the company will be more likely to invest more in social performance. However, the benefits as a whole are only partly for the shareholders whereas the costs are completely for the shareholders. Therefore, the larger the stake of a shareholder is in the company, the less likely he is to accept large investments in corporate social performance. Now there are two important questions with regard to this process. 1. How much benefits are created by investing in CSP, and do they outweigh the costs? In the literature review we have seen that results are inconclusive with regard to this subject. And the next question is: 2. How much of these created benefits go to the shareholders and how much goes to the other stakeholders in the company? These questions are situation dependable and will not be answered directly here.
5.
Hypothesis
In this section the main hypothesis will be proposed and explained. It will be tested in the next sections using the empirical data selected.
Hypothesis 1:
High ownership concentration in European multinational companies is negatively related to corporate social performance.
This thesis is aimed at exploring the relationship between ownership concentration and the level of corporate social performance within European companies. Based on literature review and general economic theory we propose that large shareholders are profit maximizers. High scores on corporate social responsibility come at a cost. The benefits are not only available to the large shareholders but to all shareholders and stakeholders. Therefore, large shareholders benefit equally, but pay more for high corporate social responsibility scores. To add strength to this line of reasoning Bartkus, Morris and Seifert (2002) find strong evidence that powerful owners discourage excessive philanthropy. Therefore, large shareholders will be less in favor of corporate social initiatives and will have a negative influence on the corporate social performance of companies. This is in line with previous research (Atkinson and Galaskiewicz, 1988, Coffey and Fryxell, 1991, Bartkus, Morris and Seifert, 2002, Hamilton and Millington, 2005, Barnea and Rubin, 2006)
This paper will also distinguish between the different types of shareholders as research shows that different types of shareholders will have different interests. Based on the classification used in the database, AMADEUS, we have included 5 groups of shareholders which can be grouped: Banks, Individuals, State, Financial Companies (institutional investors) and Industrial Companies.
6.
Descriptive statistics
In this section, the data used in this paper is presented and an explanation is given for the choices in the use of the variables. By reading this section a thorough understanding can be developed of the whole dataset and the choices made in the regressions used later in this research.
For this research, data was collected on shareholdings and on corporate social performance of European companies. The corporate social performance indicators were collected for the year 2005 from a world-wide database on corporate social responsibility based in the UK, the Ethical Investment Research Service, EIRIS. The data on shareholdings was collected from AMADEUS, a large database on European company information.
From the EIRIS dataset, 898 companies are selected that are in a list of the largest publicly quoted European companies. Of these companies, data was extracted from AMADEUS, a large database filled with financial and other information of European companies. For the year 2005 the data was extracted. Not all necessary data was available (128) and not all data was complete (79). From the sample of 898 companies, after filtering the data this is what remained:
Years Complete Data Incomplete Data Deleted during research Total 2005 691 770 128 898
Research sample of European Multinational Enterprises
Table 1 Research sample of MNE's
Corporate social performance
By factor analysis the scores were grouped to three constructs which were named: “Stakeholders”, “Governmental” and “Environment”.
Factor analysis was again applied on the three indicators to generate a single factor, named “CSR”, and use this factor in our econometric analysis. This was previously done in other research (Dam and Scholtens, 2007, Dam, Scholtens and Sterken, 2007)
The collected data and judgment by EIRIS will be the basis of the analysis of corporate social performance. Their way of scoring corporate social responsibility is similar in method to the widely used KLD database.
Large shareholders
Data on ownership percentages is extracted from the publications of European firms. To get this information AMADEUS is used, a database that contains accounting information for a large number of European firms. Only European companies will be investigated in our analysis. Not all information is disclosed in AMADEUS, Therefore we have excluded those European companies that do not disclose their main shareholders. This research created a pooled and balanced cross-section data set for 2005 of 691 companies.
Shareholders of VOLKSWAGEN A G (2005)
Shareholder name Ownership (%) Ownership^2
1 FOREIGN INSTITUTIONAL
INVESTOR
28.20% 0.080
2 PORSCHE AUTOMOBIL HOLDING
SE
14.00% 0.020
3 STATE OF LOWER SAXONY 13.60% 0.018
4 VOLKSWAGEN BETEILIGUNGS-GESELLSCHAFT MBH 13.00% 0.017 5 BRANDES INVESTMENT PARTNERS, L.P. 10.70% 0.011 6 SELF OWNED 9.80% 0.010 7 GERMAN INSTITUTIONAL INVESTORS 7.00% 0.005
8 CAPITAL GROUP COMPANIES
INC, THE
5.12% 0.003
Herfindahl index 0.163
In this case we measure the concentration of shareholders to assess their combined bargaining power in influencing the firm.
Table 2 Volkswagen example of Herfindahl Index
Below the descriptive statistics of the herfindahl index for the European Multinational Enterprises in the sample are presented.
Volkswagen AktienGesellschaft, a large division of Volkswagen, the German manufacturer of cars, has reported shareholders like many European companies. At the 10% definition a high number of 5 blockholders are
present. These shareholders will have significant influence on the day-to-day business of the company. The herfindahl index reflects this highly
Descriptive statistic Herfindahl Index of Multinational Enterprises 2005 Valid 770 Mean 12% Std. Deviation 19.5% Variance 3.8%
Table 3 Descriptive statistics Herfindahl Index of MNE's
The mean Herfindahl index over all companies is about 12% in 2005 with a standard deviation of 19.5%. To control whether the Herfindahl index gives a correct picture this paper assesses the average shareholding of the largest 5 shareholders. The average shareholding of the first 5 companies is about 11% with a standard deviation of 8.1. It shows that the Herfindahl index has a higher standard deviation than the standard deviation of the average first 5 shareholders. This is logical since the average filters out some of the deviation.
Industrial Companies Financial Companies Individuals Bank Employees State 25.00 20.00 15.00 10.00 5.00 0.00 To ta l a ve ra ge p er ce nt ag e of s ha re ho ld in g pe r c om pa ny 25.3% 45.95% 10.09% 15.12% 1.25% 2.29%
Average Shareholding by Type in European Multinational Enterprises
Figure 4 Average shareholding by type in European MNE’s
The figure above gives an overview of the average percentage of total shareholdings per group. It shows that financial companies have the largest shareholdings on average in companies.Many studies have investigated the effect of the type of shareholders on companies such as Coffey and Fryxell (1991), Johnson and Greening (1999), and Lehmann and Weigand (2000). We follow this line of research and assume that the various types of shareholders have similar interests on average, within their type.
This study also subdivided the types of shareholders in groups of companies that are dominated by one type of shareholder. The following table presents the most interesting findings when subdividing the multinational companies in groups according to their dominant shareholder type.
Herfindahl Turnover Herfindahl Turnover Herfindahl Turnover Herfindahl Turnover Herfindahl Turnover Herfindahl Turnover Valid 21 21 73 73 7 7 62 61 452 443 159 144 Mean 0.16 € 31,990,175 0.06 € 8,082,962 0.05 € 6,956,931 0.18 € 5,740,710 0.08 € 4,473,946 0.31 € 6,846,329 Median 0.14 € 11,047,087 0.03 € 2,627,000 0.01 € 2,689,958 0.12 € 456,538 0.05 € 876,013 0.2 € 2,385,537
Shareholder Types and Characteristics of Multinational Enterprises
State Bank Employees Individual Financial Company Industrial Company
Herfindahl (herfindahl index) represents the level of shareholder concentration; this will be explained in the next section. Turnover presents the average turnover in euros of the companies.
There are very few (21) companies who were dominated by state shareholders but when they are, the shareholder concentration is generally higher. Most companies have financial companies as their dominant shareholder; 452. When a bank or financial company is dominant, the Herfindahl index is usually much lower. When industrial companies (129) are dominant the Herfindahl index is exceptionally high (.31) compared to the average Herfindahl index of .14. State ownership tends to dominate very large companies (average turnover 31 million euro), whereas the average turnover of a company dominated by financial companies is about 4 million.
Control variables
The paper also includes control variables. In related literature (e.g. Ulmann, 1985, McWilliams and Siegel, 2000) size, risk, industry and R&D investments are important control variables for corporate social performance. This research will use operating revenue, return on total assets, country of origin, gearing and liquidity ratio of the firm as control variables. Size, measured by operating revenue, is a relevant control variable, as current literature provides evidence that smaller companies are less concerned with CSP (Waddock and Graves, 1997; Burke a.o., 1986). Return on assets is also a relevant control variable as is shown in the literature review; more profit can lead to more social initiatives. This is the same reason why the liquidity ratio is used. Leverage, as a control variable, is related to the level of risk management is willing to take (Barnea and Rubin, 2006). This research will control for this characteristic and take gearing in percentage as control variable.
Valid Missing Mean Median Employees Turnover Gearing
Return on Assets Liquidity ratio Employees 717 181 23998 5088 1 Turnover 750 148 € 6,173,985 € 1,229,289 0.71 1 Gearing 715 183 138 80 0.04 0.03 1 Return on Assets 0.02 0.02 -0.07 1 Liquidity ratio 0.08 -0.06 -0.07 0.09 1
Descriptive statistics on control variables for Multinational Enterprises (2005)
Data in posession of Lammertjan Dam
7. Correlations
Now that the dataset has been presented, this paper will examine the correlations between the variables.
Shareholder data correlations
First, this paper analyses the shareholder data. The correlations are presented below.
Table 6 Correlation between types of shareholder in MNE's
There are interesting significant correlations between the different groups of shareholders which are presented in the table above. Financial company’s shareholdings are negatively correlated to industrial company’s shareholdings, but positively correlated to bank shareholdings. State shareholdings are positively correlated to financial company shareholdings while they are negatively correlated to individual holdings. Individual shareholdings are negatively correlated with bank holdings. After analyzing the correlations and the preliminary regression results using banks as a separate category, it is decided that banks in the final regressions will be grouped under financial companies since the results are similar.
Correlation between types of shareholders in Multinational Enterprises
Herfindahl
Index State Employees Bank Individuals
Financial Companies Industrial Companies Herfindahl Index 1 State .079(*) 1 -. Employees 0 .090(*) 1 Bank -0.06 -0.066 -0.02 1 Individuals .178(**) -.085(*) 0.005 -.082(*) 1 Financial Companies 0.051 -.119(**) -0.048 .219(**) -0.012 1 Industrial Companies .743(**) -0.054 -0.008 -0.065 -0.017 -.156(**) 1
CSR data correlations
Variable Observations Mean
Standard
Deviation CSR Stakeholders Governmental Environment
CSR 761 0.19 1.13 1
Stakeholders 761 0.31 1.06 0.95 1
Governmental 761 -0.15 1.03 0.82 0.7 1
Environment 761 0.16 0.97 0.82 0.7 0.55 1
Descriptive Statistics Corporate Social Responsibility of Multinational Enterprises (2005) Correlations
Table 7 Descriptive statistics Corporate Social Responsibility of Multinational Enterprises (2005)
This table describes the correlations between the subconstructs of corporate social responsibility. Naturally, correlation between the variables and the CSR construct is high, since they are part of it. The correlation between the various subconstructs is also high, indicating that companies that score high on one aspect also score high on the other aspects of corporate social performance. The mean of the variables is around zero, indicating that many companies score low on the EIRIS scale from -2 to 3.
Table 8 Descriptive statistics Corporate Social Performance and control variables (2005)
This table describes the correlations including the proposed control variables. Interestingly, the correlation between the control measures for size, employees and operating revenue, are highly correlated with all aspects of corporate social responsibility. The bigger a company, the better the corporate social performance. It is therefore vital to include these control variables in our analysis. However, to prevent problems with multicollinearity, either employees or turnover has to be included. Therefore we take one measure of size, which is operating revenue. The other correlations in the table are not very high. This knowledge will be put to use
herfindahl index csr environment governance stakeholders employees operating turnover gearing return on assets liquidityratio
in the regression analysis which will focus more on the correlation between the Herfindahl index and CSR.
Correlations between CSP and types of shareholders
Now that this research has determined the correlations within the two datasets and has investigated the effect of the proposed control variables, this subsection will investigate the correlations between the constructs and subconstructs of the two datasets. The table below presents the results.
Table 9 Correlations Corporate Social Performance and types of shareholders (2005)
The table shows a high correlation between CSR and individuals. A high correlation between financials and corporate social responsibility also exists. The correlations are negative. In all other cases, the correlations are overall negative but not very high. In the regression extra attention will be devoted to the negative relationship between corporate social responsibility and individuals and financials.
The correlations are a very basic measure to investigate relationships. However, the results found here show interesting leads for further investigation. Overall, the correlations show that the size effect is important, but that there are interesting negative correlations that deserve more attention. Now that the correlations have been investigated, the next section will focus its attention on the most interesting correlations and will further investigate this by applying regression analysis.
csr environment governance stakeholders state selfowned bank individual financials industrial
8.
Regressions
In this section the methodology used to get the results is explained as well as its drawbacks. This section will help to understand the methodology and reproduce the research.
This thesis tests whether a statistically significant directional relationship exists between corporate social performance and ownership concentration. Ordinary least square regression is used, which is a statistical technique that can be used to estimate relationships between variables. Ordinary least squares estimation applies to the following linear multiple regression model used:
Corporate social performance (Yi) = a0 + a1 (X1i) + a2 (X2i) + a3 (X3i) + a4 (X4i) + a5 (X5i) + a6 X6i.... a21X21i + ei
X1i= Herfindahl Index X2i= Operating revenue X3i= Gearing
X4i= Return on total assets X5i= Liquidity ratio
X6i through X21i = Country dummy (15 European countries)
estimators (BLUE, Gauss-Markov Theorem). These estimators are unbiased (i.e. expected to be the same as the true value) and efficient (i.e. estimated with smallest variances or confidence intervals). All of the control variables have been widely used in other researches, as explained earlier in the paper.
Since there are various relationships possible a subdivision can be made within both types of shareholders and corporate social performance. This research ran several different regressions to test the effects. The data has been collected for the year 2005. The regressions have also been reverted so ownership concentration is the dependent variable and corporate social performance the independent variable. Furthermore, the control variables are excluded one by one to test whether they were necessary to get significant results. Only operating revenue, a control variable for size of the multinational enterprise was necessary for significant changes in the results. However, employees, instead of operating revenue, could also be used to arrive at the conclusions drawn below. The regression has been modified to also test for types of shareholders. To test for the effect of types of shareholders the total shareholdings of one type of shareholder have been added per company. This total has been used in the regression with corporate social performance to see whether certain types of shareholdings lead to bad or good corporate social performance. Furthermore, subconstructs of corporate social performance, earlier used by (Dam and Scholtens, 2007 and Dam, Scholtens and Sterken, 2007) have been used in regression analysis, instead of the overall construct of corporate social performance to test whether shareholder concentration has a specific influence on certain types of corporate social performance. For a full overview of the regressions, see the appendix. The main results are presented below.
9.
Results
In this section the results are presented of the ordinary least square regressions used to answer the following question: Does ownership concentration affect a company’s social performance?
Table 8 Corporate Social Performance of MNE's and Shareholder concentration in 2005 Corporate Social Performance of MNE’s and shareholder concentration in 2005
Measure 1 2 3 4
Read: Coefficient CSP Governance Environmental Stakeholder
(T-Value) Herfindahl index -0.534** -0.316 -0.362* -0.569*** (-2.39) (-1.48) (-1.90) (-2.65) Operating revenue 0.000*** 0.000*** 0.000*** 0.000*** (-9.55) -7.08 (-9.14) (-8) Gearing 0 0 0 0 (-1.08) (-0.42) (-1.2) (-0.89)
return on total assets 0.005 0.001 0.006* 0.004
(-1.36) (-0.35) (-1.95) (-1.16)
Liquidity ratio 0.006 0.007 -0.008 0.008
(-0.32) (-0.4) (-0.56) (-0.44)
R^2 0.169 0.116 0.192 0.141
Number of observations 691 691 691 691
The estimated ordinary least square regression is for measure 1: Corporate social performance (CSP) (Yi) = a0 + a1 (X1i) + a2 (X2i) + a3 (X3i) + a4 (X4i) + a5 (X5i) + a6 X6i... a21X21i + ei
This table shows the corporate social performance of multinational enterprises measured in several ways, related to the ownership concentration of the European multinational enterprises. In the left hand column first the control variables used in the regression are presented. The interaction term, the herfindahl index is presented below that and the most interesting results are printed in bold letters. The first row under the title presents the measures of corporate social performance used in the regression, namely the overall CSP construct and the subconstructs: Governance, Environmental, and Stakeholder. The * signals the level of significance. The R^2 indicates that there is a good fit in all models. Below the R^2 the number of companies used in each regression is given.
Corporate social performance and measures of shareholder concentration To test the robustness of the results presented in the first table, this study uses blockholdings as another way to measure ownership concentration. The figures below show the ownership concentration as measured in blockholdings.
Blockholders in European Multinational Enterprises in 2005
Blockholder No Blockholder 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% P e r c e n t 92.1% 7.9%
Presence of a blockholder (>5%) in European MNE's
Blockholder No Blockholder 60.0% 40.0% 20.0% 0.0% P e r c e n t 69.0% 31.0%
Presence of a blockholder (>10%) in European MNE's
Blockholder No Blockholder 60.0% 40.0% 20.0% 0.0% P e r c e n t 35.6% 64.4%
Presence of a Blockholer (>20%) in European MNE's
Figure 5 Blockholders in European Multinational Enterprises in 2005
The figure above shows that blockholdings are common in European firms. Even at the 20% level almost 40% of the firms have blockholdings of at least 20%. However, blockholdings only show the largest shareholding in a company. The table below presents the results using this alternate way of measuring ownership concentration.
2005 Blockholder No Blockholder
Number Percentage Number Percentage
Measure CSP 2005 Herfindahl index -0.534** (-2.39) 5% Blockholding -0.424*** (-2.81) 10% Blockholding -0.299*** (-3.49) 20% Blockholding -0.328*** (-3.72) Operating revenue 0.000*** 0.000*** 0.000*** 0.000*** (-9.55) (-9.62) (-9.78) (-9.89) Gearing 0 0 0 0 (-1.08) (-1) (-1.08) (-1)
return on total assets 0.005 0.004 0.004 0.005
(-1.36) (-1.03) (-1.02) (-1.25)
liquidity ratio 0.006 0.007 0.002 0.004
(-0.32) (-0.4) (-0.12) (-0.25)
R^2 0.169 0.172 0.177 0.179
Number of observations 691 691 691 691
The estimated ordinary least square regression is: (Yi) = a0 + a1 (X1i) + a2 (X2i) + a3 (X3i) + a4 (X4i) + a5 liquidity ratio (X5i) + a6 X6i... a21X21i + eiFor brevity sake, the individual control variables included in the calculations have not been reported here. For
further detail see the appendix. * indicates significance at 10, ** at 5, and *** at 1%, respectively
Corporate Social Performance and Measures of Shareholder Concentration in European MNE's (2005)
Table 8 Corporate Social Performance and Measures of Shareholder Concentration in European MNE's (2005)
These tests serve as robustness checks for the Herfindahl index regression. Overall, the results show strong support at the 1% level for a negative relationship between blockholdings and corporate social performance. At the 5, 10 and 20% definition there is a significant (1%) negative relationship between companies with blockholdings and corporate social performance. The group of companies with blockholdings significantly underperforms with regard to corporate social performance compared to the group without blockholdings. The outcomes provide further support for the conclusion that there is a negative relationship between ownership concentration and corporate social responsibility, even when measured in different ways over different time periods. The results for the control variables can be found in the appendix.
Type of shareholders and corporate social performance
Type of Shareholders and Corporate Social Performance in European MNE’s (2005)
Measure Corporate Social Performance
2005 Coefficient (T-Value) State -0.004 (-0.55) Self Owned -0.029 (-1.54) Financials -0.004** (-2.24) Individuals -0.011*** (-3.70) Industry -0.002 (-1.04) Operating revenue 0.000*** (-9.38) Gearing 0 (-1.01)
Return on total assets 0.006
(-1.51)
Liquidity ratio 0.006
(-0.37)
R^2 0.188
Number of Observations 691
The estimated ordinary least square regression is: (Yi) = a0 + a1 (State) + a2 (Self Owned) + a3 (Financials) + a4 (Individuals) + a5 (Industry) + a6 X6i... a21X21i + a22 (operating revenue) + a23 (gearing) + a24 (return on total assets) + a25 (liquidity ratio) + ei. For brevity sake, the country fixed effects are not reported. * indicates significance at 10, ** at 5, and *** at 1%, respectively
Table 9 Type of Shareholders and Corporate Social Performance in European MNE's
shareholdings and individuals show a strong and significantly negative relationship with corporate social performance.
State shareholdings have an insignificant negative effect. Therefore it can be assumed that states, through their shareholdings, do not affect a company’s corporate social performance within European multinational enterprises.
In contrast with existing literature on the topic of insider ownership and corporate social performance, no significant negative relationship can be established based on this research with regard to employee ownership and CSP.
Financial company’s shareholdings and individual shareholdings are significantly negatively related to corporate social performance of European multinational enterprises. Financial company shareholdings are common among European MNE’s. In line with our predictions these shareholders either influence a company’s social performance negatively or take stakes in companies that perform worse socially.
10. Conclusion
In this section the results will be summarized, conclusions will be drawn and recommendations for further research will be given. The conclusion gives a short overview of the most important results and conclusions.
This paper analyses, using ordinary least square regression techniques, whether corporate social performance has a relationship with ownership concentration for European multinational enterprises. Ownership concentration is measured by the Herfindahl index. Corporate social performance is based on the EIRIS 2005 database of European companies. The results show that concentrated ownership leads to significantly worse corporate social performance. Large shareholders determine the social agenda of multinational European enterprises.
This paper then subdivides corporate social performance in three subconstructs, Environmental, Governance and Stakeholders and relates it to ownership concentration. Both Stakeholder social performance issues and Environmental social performance issues are significantly negatively related to the presence of large shareholders.
This paper also uses different classifications of ownership concentration and shows that even when using different definitions the results are robust without any unexpected insignificant result. To understand which kind of investor is more likely to influence the social performance of multinational enterprises negatively the paper investigates the relationship between groups of types of shareholders and corporate social performance. It shows that financial companies and individuals have a significant negative relationship to corporate social performance of the companies in which they invest.
conclusion is that ownership concentration is negatively related to a company’s social performance. These findings are absolutely new in the academic field. Never before, with such a substantial set of companies and use of logical variables, has this field been investigated. The results are convincing and robust.
11. References
Atkinson, L., Galaskiewicz, J., ‘Stock Ownership and Company Contributions to Charity’ Administrative Science Quarterly, Vol. 33, No. 1, 82-100
Barnea, A., Rubin, A (2006) ‘Corporate Social Responsibility as a Conflict Between Shareholders’ EFA 2006 Zurich Meetings working paper SSRN: http://ssrn.com/abstract=686606
Bartkus, B.R., Morris, S.A., Seifert, B., (2002) Governance and Corporate Philanthropy: Restraining Robin Hood? Business Society, Vol. 41, No. 3, 319-344
Bolton, P von Thadden, E.-L., (1998) ‘Blocks, Liquidity, and Corporate Control The Journal of Finance Vol. 53 No. 1, 1-25
Bowen, H.R. (1953), ‘Social Responsibilities of the Businessman’, Harper and Brothers, Publishers, New York, NY.
Brammer, S. Millington, A., (2005) ‘Corporate Reputation and Philanthropy: An Empirical Analysis’ Journal of Business Ethics, Vol 61, No. 1, 29-44
Burke, L., J. M. Logsdon, W. Mitchell, M. Reiner (1986) ‘Corporate community Involvement in the San Francisco Bay Area’, California Management Review, XXVIII(3), pp. 122–141.
Carrol, A.B. (1991), ‘The Pyramid of Corporate Social Responsibility: Toward the
moral management of organizational stakeholders’, Business Horizons, Vol 34, No. 4, 39-48.
Carrol, A.B. (1999), ‘Corporate Social Responsibility: Evolution of a Definitional Construct’, Business and Society, Vol 38, No. 3, 268-295.
Cochran, P.L., Wood, R.A. (1984), ‘Corporate Social Responsibility and Financial Performance’, Academy of Management Journal, Vol. 27, No. 1, 42-56.
Coffey, B., Fryxell, G.E., (1991) ‘Institutional ownership of stock and dimensions of corporate social performance: An empirical examination’ Journal of Business Ethics Vol. 10, No. 6, 437-444
Cronqvist, H., Fahlenbrach, R., (2006) ‘Large Shareholders and Corporate Policy’ SSRN Working Paper Series, 891188, Review of Financial Studies, forthcoming
Dam, L., (2008) ‘Corporate Social Responsibility and Financial Markets’ proefschrift Rijksuniversiteit Groningen
Dam, L., Scholtens, L. J. R., Sterken (2007) ‘Corporate Governance and International Location Decisions of Multinational Enterprises’ Corporate Governance, Vol 15, No. 6, 1330-1347
Donaldson, T., Preston, L.E. (1995), ‘The Stakeholder Theory of the Corporation: Concepts, Evidence, and Implications’, Academy of Management Review, Vol. 20, No. 1, 65-91
Freeman, R.E. (1984), Strategic Management: A Stakeholder Approach, Pitman, Boston, MA.
Friedman, M. (1970), ‘The Social Responsibility of Business is to increase its profits’, The New York Times Magazine, September 13, 122-126.
Gorton, G., Schmid, F.A., (2000) ‘Universal banking and the performance of German Banks’ Journal of Financial Economics, Vol. 58, No. 1-2, 29-80
Griffon, J. J. and Mahon, J. F. (1997) “The corporate social performance and corporate financial performance debate: twenty five years of incomparable research”, Business and Society, Vol. 36, No. 1, 5-31
Guerard, J.B. (1997) “Is there a cost to being socially responsible in investing?”, Journal of Forecasting, No. 16, 475-490
Heinkel, R., Kraus, A. and Zechner, J. (2001) ‘The effect of green investment on corporate behavior’ Journal of Financial and Quantitative Analysis, Vol. 36, No. 4, 431-439
Jensen M.C., Meckling, W.H. (1976) ‘Theory of the firm: Managerial behavior, agency costs and ownership structure’ Journal of Financial Economics. Vol. 3, No. 4, 305-360
Jones, T.M., Wicks, A.C., ‘Convergent Stakeholder Theory’, The Academy of Management Review, Vol. 24, No. 2, 206-221
Johnson, R.A., Greening, D.W., (1999) ‘The Effects of Corporate Governance and Institutional Ownership Types on Corporate Social Performance’ The Academy of Management Journal, Vol. 42, No. 5, 564-576
Korf, F.B., (2007) The Multidimensionality of CSR: The Effect on Short-term Financial Performance’, Master Thesis Rijksuniversiteit Groningen
La Porta, Lopez-De-Silanes, F., Shleifer, A. (1999), ‘Corporate Ownership Around The World’, The Journal of Finance, Vol. Liv, No. 2, 471-517
Mahoney, L., Roberts, R.W., (2007) ‘Corporate social performance, financial performance and institutional ownership in Canadian firms’ Accounting forum, Vol. 31, No. 3, 233-253
Margolis, J. D., Walsh, J. P. (2001). ‘People and profits? The search for a link between a company’s social and financial performance.’ Greenwich, CT: Erlbaum.
McWilliams, A. Siegel, D (2000), ‘Corporate Social Responsibility and Financial Performance: Correlation or Misspecification’, Strategic Management Journal, Vol. 21, No. 5, 603-609.
Morck, R, Shleifer, A., Vishny, R.W., (1988), ‘Management Ownership and market valuation: An Empirical Analysis’ Journal of Financial Economics, Vol./No. 20, 293-315
Orlitzky, M., Schmidt, F.L. and Ryes, S.L. (2003), ‘Corporate social responsibility and financial performance: a meta-analysis’, Organization Studies, Vol. 24, No. 3, 403–441.
Pagano, M., Roell, A., (1998) ‘The Choice of Stock Ownership Structure: Agency Costs, Monitoring, and the Decision to go Public’ The Quarterly Journal of Economics, Vol. 113, No. 1, 187-225
Pirsch, J., Gupta, S., Grau, S.L., 2007 A Framework for Understanding Corporate Social Responsibility Programs as a Continuum: An Exploratory Study’ Journal of Business Ethics, No. 70, 125–140 _ Rahbein, K., Waddock, S.A. and Graves, S.B. (2004) ‘Understanding Shareholder Activism: Which Corporations are Targeted?’ Business Society, Vol. 43, No. 3, 239-267
Roman, R.M., Hayibor, S. Agle, B.R (1999) ‘The Relationship between Social and Financial Performance: Repainting a Portrait’ Business and Society, Vol. 38, No. 1, 109-125
Rowley & Burman, (2000) ‘A Brand New Brand of Corporate Social Performance’. Business & Society, Vol. 39, No. 4, 397-418
Seifert, B., Morris, S.A., Bartkus, B.R., ‘Having, Giving, and Getting: Slack Resources, Corporate Philanthropy, and Firm Financial Performance’ Business Society Vol 43, No. 2, 135-161
Shleifer, A., Vishny, R.W. (1986) ‘Large Shareholders and Corporate Control’ The Journal of Political Economy, Vol. 94, no. 3, 461-488
Shleifer, A., Vishny R.W. (1997) ‘A Survey of Corporate Governance’, The Journal of Finance, Vol. 52, No. 2. 737-783
Sparkes, R., (2001) ‘Ethical investment: whose ethics, which investment?’, Business Ethics: A European Review, Vol. 10 No. 3, 194-205
Thomsen, S., and Pedersen, T., (2000) ‘Ownership Structure and Economic Performance in the largest European Companies.’ Strategic Management Journal, Vol 21, No 2. 689-705
Ullman, A.A. (1985), ‘Data in Search of a Theory: A Critical Examination of the Relationships among Social Performance, Social Disclosure, and Economic Performance of U.S. Firms’, The Academy of Management Review, Vol. 10, No. 3, 540-557
Waddock, S.A. and Graves, S.B. (1997), ‘The corporate social performance-financial performance link’, Strategic Management Journal, Vol. 18, 303–319.
Wartick, S.L., Cochran, P.L., ‘The Evolution of the Corporate Social Performance Model’ The Academy of Management Review, Vol. 10, No. 4, 758-769
Appendix A
Table of Tables
Table 1 Research sample of MNE's... 17
Table 2 Volkswagen example of Herfindahl Index ... 19
Table 3 Descriptive statistics Herfindahl Index of MNE's ... 20
Table 4 Shareholder Types and Characteristics of Multinational Enterprises ... 21
Table 5 Descriptive statistics on Control Variables for Multinational Enterprises . 22 Table 6 Correlation between types of shareholder in MNE's... 24
Table 7 Descriptive statistics CSR of Multinational Enterprises (2005) ... 25
Table 10 Corporate Social Performance of MNE's and Shareholder concentration30 Table 9 CSP and Measures of Shareholder Concentration in European MNE's .. 33
Table 10 Type of Shareholders and CSP in European MNE's ... 35
Table of Figures Figure 1 Conceptual Model of Research Question ... 7
Figure 2 Relationships between the variables ... 13
Figure 3 The Relationship between Shareholders and CSP ... 14
Figure 4 Average shareholding by type in European MNE... 21
Appendix B
Measure
2004 2004 2004 2004 2005 2005 2005 2005
C ont rol Variables Coefficient (T-Value)
Operating revenue 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000***
-6.52 -6.34 -6.35 -6.5 (-9.55) -9.62 -9.78 -9.89
G earing 0 0 0 0 0 0 0 0
-1.03 -0.87 -1 -1 (-1.08) -1 -1.08 -1
return on total assets 0.007* 0.006 0.007* 0.007* 0.005 0.004 0.004 0.005
-1.68 -1.61 -1.78 -1.74 (-1.36) -1.03 -1.02 -1.25 liquidity ratio -0.048 -0.043 -0.042 -0.046 0.006 0.007 0.002 0.004 (-1.56) (-1.43) (-1.36) (-1.51) (-0.32) -0.4 -0.12 -0.25 cou1 0.239 0.169 0.018 0.035 0.48 0.472 0.406 0.255 -0.45 -0.33 -0.04 -0.07 -0.65 -0.65 -0.56 -0.35 cou2 0.137 0.141 -0.101 -0.054 0.476 0.456 0.349 0.291 -0.23 -0.24 (-0.17) (-0.09) -0.63 -0.6 -0.46 -0.39 cou3 -0.037 -0.069 -0.23 -0.201 0.347 0.32 0.274 0.169 (-0.07) (-0.12) (-0.41) (-0.36) -0.46 -0.43 -0.37 -0.23 cou4 0.014 -0.049 -0.183 -0.17 0.422 0.37 0.306 0.21 -0.02 (-0.09) (-0.33) (-0.31) -0.56 -0.49 -0.41 -0.28
cou5 0.202 0.206 (dropped) (dropped) 0.473 0.469 0.395 0.305
-0.27 -0.29 -0.53 -0.52 -0.44 -0.34
cou6 (dropped) (dropped) (dropped) (dropped) (dropped) (dropped) (dropped) (dropped)
cou7 -0.078 -0.187 -0.402 -0.3 0.489 0.466 0.37 0.295
(-0.13) (-0.33) (-0.70) (-0.52) -0.65 -0.62 -0.49 -0.39
cou8 0.26 0.235 0.099 0.11 0.722 0.609 0.594 0.576
-0.46 -0.42 -0.17 -0.19 -0.95 -0.8 -0.79 -0.76
cou9 (dropped) (dropped) -0.138 -0.125 0.243 0.12 0.159 0.096
(-0.19) (-0.17) -0.29 -0.15 -0.19 -0.12 cou10 0.253 0.221 0.086 0.106 0.664 0.568 0.531 0.466 -0.45 -0.4 -0.16 -0.19 -0.89 -0.76 -0.71 -0.62 cou11 0.25 0.244 0.09 0.109 0.661 0.583 0.536 0.438 -0.46 -0.45 -0.17 -0.2 -0.89 -0.78 -0.72 -0.59 cou12 0.779 0.727 0.594 0.608 1.145 1.059 1.034 0.974 -1.29 -1.22 -0.99 -1 -1.46 -1.35 -1.32 -1.25
cou13 (dropped) (dropped) (dropped) (dropped) (dropped) (dropped) (dropped) (dropped)
cou14 0.659 0.666 0.451 0.507 1.085 1.043 1.01 0.967
-1.01 -1.04 -0.7 -0.78 -1.34 -1.29 -1.26 -1.2
cou15 0.194 0.192 0.06 0.073 (dropped) (dropped) (dropped) (dropped)
-0.22 -0.22 -0.07 -0.08
cons_ -0.228 0.358 0.173 -0.002 -0.481 -0.113 -0.236 -0.216
(-0.43) -0.67 -0.33 (-0.00) (-0.66) (-0.15) (-0.32) (-0.29) interaction term Coefficient
(t-value) Herfindahl index -0.013 -0.534** (-0.05) (-2.39) 5% Blockholding -0.586*** -0.424*** (-3.98) (-2.81) 10% Blockholding -0.274*** -0.299*** (-2.68) (-3.49) 20% Blockholding -0.108 -0.328*** (-1.03) (-3.72) R^2 0.114 0.14 0.126 0.116 0.169 0.172 0.177 0.179 Number of observations 548 548 548 548 691 691 691 691 CSP CSP
Corporate Social Performance and Measures of Shareholder Concentration in European MNE's (2004-2005)
CSP CSP
The estimated ordinary least square regression is: Corporate social performance 2005 (CSP) (Yi) = a0 + a1 Interaction term (X1i) + a2 Operating revenue (X2i) + a3 Gearing (X3i) + a4 Return on total assets (X4i) + a5 liquidity ratio (X5i) + a6 Country Dummy X6i + ei. For brevity sake, the individual control variables included in the calculations have not been reported here. For further detail see the appendix. *
Measure CSP CSP
2004 20 05
Control Variables Coefficient
(T-Value)
Operating revenue 0.000 *** 0.000***
-6.34 -9.38
G earing 0 0
-0.64 -1.01
return on total assets 0.0 07* 0.006
-1.74 -1.51 liquidity ratio -0.034 0.006 (-1.11) -0.37 cou1 0.201 0.582 -0.8 cou2 0.086 0.504 -0.14 -0.67 cou3 -0.101 0.377 (-0.17) -0.5 cou4 -0.103 0.489 (-0.18) -0.65 cou5 -0.022 0.474 (-0.03) -0.53
cou6 (dropped) (dropped)
cou7 -0.21 0.533 (-0.36) -0.71 cou8 0.29 0.795 -0.51 -1.05 cou9 (dropped) 0.275 -0.33 cou10 0.179 0.679 -0.31 -0.91 cou11 0.255 0.79 -0.46 -1.06 cou12 0.66 1.145 -1.11 -1.46
cou13 (dropped) (dropped)
cou14 0.582 1.115 -0.89 -1.39 cou15 0.108 (dropped) -0.12 _cons -0.026 -0.418 (-0.05) interaction term S tate 0 -0.004 (-0.01) (-0.55) S elf Own ed -0 .040* -0.029 -0.19 (-1.54) Financials -0.003 -0.004 ** (-.162) (-2.24) I ndividuals -0.0 16*** -0.011* ** (-4.81) (-3.70) I ndustry 0.001 -0.002 -0.44 (-1.04) R^2 0.159 0.188 Number of observations 548 691
Type of S hareh olders and Corporate Social Performance
The estimated ordinary least square regression is: Corporate Social Responsibility= β0 +β operating revenu e +β Gearing +β Return On
Measure 1 2 3 4 CSP G overnance En vironm ental St akeh older C ontrol Variables Coefficient
(T-Value)
Operating revenue 0.000*** 0.000*** 0.000*** 0.000***
(-9.55) -7.08 (-9.14) -8
G earing 0 0 0 0
(-1.08) -0.42 (-1.2) -0.89
return on total assets 0.005 0.001 0.006* 0.004
(-1.36) -0.35 (-1.95) (-1.16) Liquidity rat io 0.006 0.007 -0.008 0.008 (-0.32) -0.4 (-0.56) (-0.44) cou1 0.48 0.894 -0.44 0.576 -0.65 -1.28 (-0.71) -0.82 cou2 0.476 1.245* -0.161 0.335 -0.63 -1.72 (-0.25) -0.46 cou3 0.347 0.803 -0.057 0.231 -0.46 -1.12 (-0.09) -0.32 cou4 0.422 0.593 -0.091 0.539 -0.56 -0.82 (-0.14) -0.75 cou5 0.473 0.904 -0.216 0.47 -0.53 -1.06 (-0.28) -0.55
cou6 (dropped) (dropped) (dropped) (dropped)
cou7 0.489 1.371* -0.461 0.344 -0.65 -1.89 (-0.72) -0.47 cou8 0.722 1.350* -0.11 0.684 -0.95 -1.86 (-0.17) -0.94 cou9 9 0.243 0.494 -0.643 0.576 -0.29 -0.62 (-0.91) -0.72 cou10 0.664 0.73 0.028 0.793 -0.89 -1.02 -0.04 -1.1 cou11 0.661 1.013 -0.216 0.738 -0.89 -1.42 (-0.34) -1.03 cou12 1.145 1.392* 0.239 1.113 -1.46 -1.86 -0.36 -1.48
cou13 (dropped) (dropped) (dropped) (dropped)
cou14 1.085 0.841 0.179 1.326*
-1.34 -1.09 -0.26 -1.71
cou15 (dropped) (dropped) (dropped) (dropped)
_cons -0.481 -1.166* 0.338 -0.384 (-0.66) (-1.67) -0.54 (-0.55) interaction term Herfindahl index -0 .534** -0.3 16 -0.362* -0.56 9*** (-2.39) (-1.48) (-1 .90) (-2.65) Pseudo R^2 0.169 0.116 0.192 0.141 Number of observations 691 691 691 691
The estimated ordinary least square regression is for measu re 1 : Corporate social performance (CSP) (Yi) = a0 + a1 Herfin dahl In dex (X1i) + a2 Operating revenu e (X2i) + a3 Gearing (X3i) + a4 Return
on total assets (X4i) + a5 liqu idity ratio (X5i) + a6 Coun try Dummy X6 i + ei
Measure 2 -4 use the same regression except for the dependent variable. For brevity sake, the country fixed effects are not reported. * indicates significance at 10, ** at 5, and *** at 1%, respectively
Variable Obs Mean Std. Dev. Min Max
csr 761.00 0.19 1.13 -1.39 2.55
stakeholders 761.00 0.31 1.06 -1.21 2.29
governance 761.00 -0.15 1.04 -1.64 1.31
