1
The impact of foreign board member diversity: importing
Corporate Social Responsibility scores.
Sven van Oenen
University of Groningen
Faculty of Economics and Business
MSc Finance
Supervisor Dr. N. Selmane
June 2019
Abstract
This study examines the effect board diversity with regards to nationality has on various measures of corporate social responsibility (CSR) performance. A distinction between civil and common law foreign countries is also made. The results that are obtained using a sample of 349 S&P 500 companies between 2002 and 2016 show a positive effect of diversity on CSR performance. We also observe differences between the outcomes of civil and common law foreigners, civil law foreigners have no significant effect on governance performance while common law foreigners show a strongly significant effect.
Keywords: Board diversity, nationality, common law, civil law, corporate social responsibility.
2
Introduction
From the start of the 21st century, board diversity has been a hot topic in society. Many have spoken out against the relatively homogenous board rooms in terms of gender and race in large corporations. This goes so far even that in the U.S. and several EU countries board diversity practises and disclosure concerning these practises is required to be improved by law. Due to the popularity of the subject, and the political debate surrounding it, many scholars took towards studying board diversity and its impact on firm value (e.g. Carter et al. 2003) or Corporate Social Responsibility (CSR) performance (e.g. Harjoto et al. 2015, Bear et al. 2010). Research concerning board diversity often finds positive relationships between firm value, CSR scores, and various diversity dimensions such as gender (Post et al. 2011, Harjoto et al. 2015, Bear et al, 2010) and tenure (Harjoto et a. 2015). Other diversity dimensions like age are less conclusive, Harjoto et al. (2015) finds no significant relationship between CSR performance and age while Post et al (2011) find a positive relationship between environmental CSR scores and boards with older members on average. Despite this previous research, relating diversity in terms of foreign board membership to CSR performance has so far been mostly overlooked, while a clear argument can be made for its present day importance. I attribute this to the fact that while the pressure of both governments and society towards increasing board diversity in terms of for instance gender or race has increased, there is no such pressure regarding diversification in terms of nationality. However, despite this lack of social and regulatory pressure, companies have already started diversifying their boards in terms of nationality. In the data sample introduced in this study, I find that the percentage of foreign board members increases from 4.36% to 12.32% from 2002 until 2016 (See Figure I) – a 183% increase!
3 stakeholder conflicts that might arise due to different interests. Nowadays, corporations exist within an increasingly global business environment, where multinational firms are now held accountable for a wider range of stakeholders, a global audience (Werther and Chandler, 2010). Considering this argument, our research question is: Do we improve recognition of the interests and needs of this global audience of stakeholders by having a more diverse board in terms of nationality?
Figure I:
The percentage of board members with a foreign nationality measured over S&P 500 firms between 2002 and 2016.
It is obvious that ‘foreign’ is a very broad specification of board members. After all, some of these foreign board members may indeed be much more ‘foreign’ to an S&P 500 company than others. Perhaps a more culturally distant foreign board member might offer different diversification options, therefore broadening the perspective of the board more with regards to some stakeholder needs, and less in others. In order to create some nuance, I will make a distinction in this study within the group of foreign board members. I will divide them into 2 groups: common law and civil law foreign board members. These groups are then separately tested in their own diversity indices.
Oxelheim and Randøy (2003) mention that the Anglo-American corporate governance system is often regarded as the most demanding. Therefore, this may lead to an increase in governance score. We know civil law is not used in countries with an Anglo-American governance system, while a very large part of the common law foreign board members are
0,00% 2,00% 4,00% 6,00% 8,00% 10,00% 12,00% 14,00% 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
4 from countries with Anglo-American governance culture (Canada, for instance). Therefore, we might expect that diversifying by hiring board members who have a civil law nationality might not increase governance performance. The reason I use common law and civil law as our distinction, and not Anglo-American governance system versus non-Anglo-American governance system is that it is very hard to categorise how much countries would fall under this Anglo-American governance system. In fact, there is not even one permanent perspective on what constitutes “Anglo-America”. Canada would be obvious, but what about, for example, Jamaican or Dominican foreign board members. There is a very clear distinction between countries that follow common law and countries that follow civil law, therefore I do not rely on my own judgement to identify countries as one or the other when I use this criterion to categorize countries.
In order to measure CSR performance I use the Environmental, Social, Governance (ESG) scores retrieved from Thomson Reuters EIKON. This ESG score is a score compiled of three separate ‘pillars’: the respective firm’s Environmental score, Social score, and Governance score. It is also important to note that board structure, and more specifically foreign board membership is one of the measurements used in the governance pillar. For this reason, and in order to test one of our hypotheses, I test both the combined ESG score and the three separate pillars. These separate pillars also offer us additional added value in our inference. If a company would want to improve their Environmental score for instance, I might be able to tell them if diversifying their board with regards to nationality could help them in doing so. And if so, whether they should hire a common law or civil law board member.
In this study, I use a sample consisting of S&P 500 companies only. Information regarding their board consistency is retrieved from the WRDS BoardEx database. Unfortunately, the nationality of a substantial minority of board members is missing in this database, and therefore some concessions had to be made and over 150 firms were removed from the dataset due to unavailability of information regarding their board members. Also, since our time period is quite long (from 2002 to 2016), for many firms some years of data are missing.
5 common law country, this essentially counts all foreign common law board members as domestic board members. Therefore there will always be lower or the same diversity score as in the first index. The third index measures diversity with respect to all foreign common law board members compared to all common law members in the board. The civil law foreign board members are thus excluded from this third index.
Using our dataset of S&P 500 companies between 2002 and 2016 I find interesting results. I find that, for the first index, there are highly significant results (<1% significance) between foreign board member diversity, the total ESG Score and the social pillar. For the other pillars I also find significant positive result, however only at a 5% significance. This leads us to believe that diversifying your board with regards to nationality leads to an overall better CSR performance. Therefore I reject the null hypothesis for hypothesis 1, in line with what I might expect from previous literature.
For the second index, I find that the overall ESG score and the environmental pillar are still highly significant. However, the social score has dropped slightly in significance (<5%). But the largest difference between index 1 and this index is clearly the governance pillar, whereas it was significant in the first index, there is no convincing significance to be found in the second index. Moreover, when we test our third index the results are highly significant. This leads us to reject the null hypothesis in hypothesis 2, the increase in governance score that we found in our first index is because of diversification with regards to nationality is due to diversifying by hiring foreign common law board members, not foreign civil law boards members. This is in with what we could expect following the reasoning by Oxelheim and Randøy (2003).
6
Literature review
Stakeholder theory and Corporate Social Responsibility
Stakeholder theory and its approach towards corporate strategy first emerged around the mid 1980’s, where one of the main outings of the movement was Freeman’s Strategic Management: A Stakeholder Approach in 1984. Here, he argued firms should attempt to align their strategy to not only the interests of their stockholders but to a broader audience: the stakeholders. Freeman (1984) defines stakeholders as “any group or individual who is affected by or can affect the achievement of an organization’s objectives”. There are clear links between a strategy that attempts to recognize the needs and interests of all stakeholders (stakeholder management) and Corporate Social Responsibility practises. Stakeholder management’s principal argument is that companies should act in the interests of all constituents who can influence or be influenced by obtaining the companies objectives (Freeman, 1984). Firms who follow the fundamentals of stakeholder management can also be considered to behave in a responsible manner socially (Benson and Davidson, 2010).
Following the reasoning above, studies like Hillman and Keim 2001, Benson and Davidson 2010 and Harjoto et al. 2015 have used corporate social responsibility as a proxy for having an effective corporate strategy with regards to different stakeholders. In their studies, they used the KLD database. While I do not have access to this database, I do have access to the Thomson Reuters ESG scores, which are constructed very similarly to the KLD database scores. Both take environmental, social and governance performance into account. It is argued that, due to the increased amount of different interests and needs stakeholder have compared to stockholders, a diverse board offers a broader perspective and is therefore better equipped to recognize and act upon these different interest and needs stakeholders might have (Harjoto et al. 2015).
Diversity and CSR
7 both an increase in CSR strengths and decreases CSR concerns. Bear et al. (2010) also finds a strongly significant relationship between gender diversity and CSR scores.
Other indices were also considered, Tenure and expertise are reported to reduce CSR concerns (Harjoto et al. 2015), while results with regards to age are quite inconclusive. Harjoto et al. does not find a significant relationship, Post et al. (2011) finds a positive relationship between specifically environmental CSR scores and more elderly boards.
The underlying principle of many of these studies is that a board which represents the stakeholders of a firm are better at recognizing the needs and interests of these stakeholders. Considering the global audience and the increasingly international business described by Werther and Chandler (2010), looking at diversity from an international perspective makes sense. But addressing this question from a stakeholder perspective has so far not been done. However, the effects of “importing” governance systems have been addressed by Oxelhaim and Randøy (2003), and their research is one of the foundations for my second hypothesis. Considering the previous literature’s result with regards to diversity and CSR performance, and the link to stakeholder theory. I formulate our first hypothesis:
Hypothesis 1: Ceteris paribus, foreign board membership is positively associated with CSR performance.
Common law, Civil law and the Anglo-American governance system
8 Oxelheim and Randøy (2003) examine the effect of Anglo-American foreign board membership on the value of Swedish and Norwegian firms. Anglo-American board membership might have an effect on firm value due to “importing” of the Anglo-American corporate governance system along with these board members. This corporate governance system is often regarded as the most demanding, and therefore “importing” it will signal a firms’ willingness to improve its corporate governance practises, increasing reputation. A positive relation is found between firm value and Anglo-American board membership. While it might also be an obvious choice to make a distinction between Anglo-American and non Anglo-American in this study, I have reasons not to do so. First of all, defining what the Anglo-American governance system entails, which countries follow it and to what extend these countries follow it is very difficult. In fact, there is not even a consensus on what countries belong to Anglo-America, as many Caribbean countries (of which nationality there are also board members in the sample used in this study) are included or excluded depending on the source. Also, if I were to test our third index using “US versus non-US Anglo-American”, I would receive a very small fraction for non-US Anglo-American board members. This brings along with it severe skewness and normality issues. I can already see some skewness in the diversity groups I am using now, increasing with smaller groups. Therefore, I choose civil law and common law as our way of making a distinction within the foreign board members in our sample. Common law countries have been heavily influenced by British culture, just like the Anglo-American countries, it is not that far of a stretch to say that common law countries are more likely to adhere to the Anglo-American governance system to some extent. Because of this, board members from foreign common law countries bring with them both the diversity of foreignness and the quality of the Anglo-American governance system, and thus are in theory good candidates to increase corporate governance performance. Foreign board members from civil law countries still offer the diversification of being foreign, however firm would be importing a less demanding governance system along with this board member. Therefore, I might expect that the impact a civil law foreign board member might have on governance performance to be less than that of a common law foreign board member. This conclusion leads us to our second hypothesis:
9
Research method and data
The dataset used in this study concerns S&P 500 firms over the time period between 2002 and 2016. There are clear reasons for choosing S&P 500 firms, some of the largest firms on the planet. First of all, foreign board membership is not too widespread in the US, and even our sample including the largest, presumably most internationally oriented firms only had a 4.35% fraction of foreign board members at the start of our sample in 2002, increasing to 12.32% in 2016 over the course of 15 years of data (see Figure I). In order to run our tests, I need a reasonable fraction of foreign board members in our dataset in order to avoid too much skewness in our data and obtain a reasonably normal distribution in our diversity indices. This holds especially true if I am going to make a distinction in foreign board members later with regards to common or civil law, decreasing the fractions further. I can see this effect in our own variables too, see Table III. The diversity indices with the lowest means have the highest skewness and kurtosis values. Also, the data availability concerning board member nationality is somewhat limited, and using boards from the largest firms gives us a better chance at obtaining as much information per company as possible. Still, while researching these large companies data availability was a problem, and many firms did not have sufficient data to be of use. Data unavailability is a possible cause for bias in research, if there is a clear reason the data is not made public. I do not believe however that this is the case in this study, because the data is actually not unavailable, it is just not included in the data source we use. Companies could not hold this data confidential even if they desperately wanted to. The data concerns some of the board members of the largest companies in existence, and plenty of information including their nationality can be found online. It would be possible to look up each board member individually and obtain a more complete sample that way, however considering the time constraints that is not feasible for this study, so concessions had to be made. 151 companies were removed from the list of S&P 500 companies, leaving 349 in our sample. There was a clear criterion that I used to assess whether I would include a company or not: the average number of board member nationality observations per year had to be greater than 2. The reasoning for this has to do with the diversity indices I use, and will be explained after I explain the diversity indices in this section.
10 these three pillars. Not every pillar has the exact same amount of indicators, but each pillar has equal weight of 1/3rd in determining the ESG score, therefore some indicators weigh more heavily than others. Mean total ESG and individual pillar scores over the years 2002 to 2016 are shown in Table I. It is clear to see that CSR performance as measured by the ESG scores has increased over this time period. The total ESG score increases from 48.09 to 65.16, an increase of 35.5%. This is also reflected in the individual pillar Environmental, Social and Governance scores, showing increases of 42.7%, 37.5% and 27.8% respectively. The relatively lower increase in governance performance is mainly due to the higher mean performance in the governance pillar initially. Also, even though governance scores start relatively high, they actually score lower on average than the social score, scoring only 54.55 compared to an average social score of 57.08. on average, firms in our sample tend to score worst on the environmental score, with an average of 53.63. This all adds up to an average total ESG score of 55.38.
Table I
Mean total ESG and individual pillar scores of 349 S&P 500 companies, 2002-2016.
“ESG” indicates the total ESG score, whereas “Environmental”, “Social” and “Governance” indicate the respective environmental, social and governance pillar scores.
Year ESG Environmental Social Governance
11 In order to measure the effect of the different forms of diversity on CSR performance, I must use a diversity index. In this study, we use Blau’s index of heterogeneity as defined in Blau (1977):
Where B refers to Blau’s diversity index, i is the number of categories and P is the proportion of board members within the category. The index can, depending on the number of categories, take on values between 0 and . There are 2 categories in our all of our indices: domestic and foreign in the first index, civil law and common law in the second index, and American and foreign common law in the third index. Therefore, all three indices can take values between 0 (indicating no diversity in this dimension) and 0.5 (indicating perfect diversity in this dimension). So, the higher the value of the index, the more diversity in the board with regards to that diversity dimension. It is important to note that, while the first and second index use the entire board as denominator to calculate the proportions, the third index uses only the common law board members (both American and foreign) as denominator to calculate the proportions. The characteristics of this index lead me to my data selection criteria. I removed 151 companies from the total list of 500, and the criterion was that a company needed to have an average number of board member observations higher than 2 per year. The reason I specifically wanted more than 2 is as follows: when you only have 2 observations, this can give a very extreme picture of the diversity in the board. If the 2 observations are both of domestic origin, then you will have a diversity score of 0, whereas if you would have 1 of them being foreign you would already obtain a perfect diversity score of 0.5. Of course having for instance only 3 board member observations is not ideal either if the actual board consists of more members, but I had to draw the line somewhere and balance between keeping a reasonable sample size and getting complete information.
12 average common foreign diversity increased from 0.05 to 0.10. We can also see that over the time period we investigate boards are on average more diverse in terms of foreign nationality than with regard to foreign civil law and foreign common law board members. This is to be expected of course, since foreign civil law and foreign common law are just different subsections of being foreign nationality. The average diversity score of ‘Diversity common foreign’ is higher than that of ‘Diversity Civil”, but this does not necessarily have to mean that the fraction of foreign common law board members is larger than that of foreign civil law members. This is because as previously mentioned the denominator used for calculating the common foreign diversity foreign is smaller than the denominator used to calculate the other diversity scores.
Table II
Mean diversity indices in 349 S&P 500 companies, 2002-2016.
“Foreign”, “Civil” and “Common Foreign” indicate Blau’s diversity Indices scores with respect to foreign board members, civil law board members and foreign common law board members respectively.
Year Diversity Foreign Diversity Civil Diversity common foreign
13 These indices are then used as inputs in a fixed effects Ordinary Least Squares (OLS)
regression to test the effect of different diversity indices on different measures of CSR performance. The regressions all follow a similar form:
Where denotes the different CSR measures (total ESG score and the environmental, social and governance pillars). denotes the constant and the ’s denote coefficient factors for the respective variables. denotes the different diversity indices (the foreign, civil law and foreign common law indices), denotes the board size, denotes the firm age and denotes CEO duality. The denotes a firm-fixed effect and a time-fixed effect.
denotes the error term. Finally, stands for the respective firm and stands for the respective time .
I estimate my model using firm-fixed and time-fixed effects as this helps in controlling for unobserved heterogeneity. I have performed the Hausman test on our data to test for random effects, and the null hypothesis of random effects was rejected at a 1% significance level. Also, I adjust for clustered standard errors (CSEs), CSEs are observed typically when some observations within the sample have a relationship to one another. When unaccounted for, CSEs can lead to standard errors which are too small, making for too small p-values and eventually, possibly biased inference. I account for CSEs because panel data such as the data used in this study is particularly prone to CSEs.
Control variables
Some other variables then board diversity could have an impact on the ESG scores of a company. In this study, I control for board size, firm age, firm size and CEO duality.
14 Board size, the total number of board member observations I had per firm. Research in the past by Post et al. (2011) finds a small, insignificant positive impact of board size on CSR performance. However, previous research has found that a larger board size leads to smaller firm value (Mak & Kusnadi, 2004). As firm value and CSR performance are positively related (Jo & Harjoto, 2011) we might expect an increase in board size might be able to decrease CSR performance as well.
The natural logarithm of firm size, measured by taking the natural logarithm of the firms previous years’ total revenues. I use the natural logarithm because these revenues do not follow a normal distribution. The reason for inclusion of firm size is Udayasankar’s (2008) argument that larger firms tend to be more socially responsible due to an increased exposure to the public. Also, Harjoto et al. (2015) finds a significant relationship between firm size and CSR performance.
The natural logarithm of firm age (years). Again, I take the natural logarithm to make this variable more normally distributed. Harjoto et al. (2015) finds no significant relationship between firm age and CSR performance, but other studies by Badulescu et al. (2016) and Withisuphakorn & Jiraporn (2015) find evidence for younger firms being less involved in CSR activities.
Descriptive statistics
Figure V presents the descriptive statistics of the different variables in our dataset. The firms in our sample have a mean foreign diversity score of 0,11, a mean civil diversity score of 0,05 and mean common foreign diversity score of 0,08. As I can see from the skewness and kurtosis values, the distribution of the diversity indices becomes more skewed and higher in kurtosis value when the mean diversity score is lower.
The mean total ESG score in our sample is 56.26. The mean Environmental, Social and Governance score are 54.77, 58.11 and 55.91 respectively. The firms in our sample thus perform relatively well in the Social score, and relatively poor in the Environmental score.
15 For most variables we have 4336 observations, however some data regarding revenues, firm age and CEO Dualitywere missing, resulting in 4128, 4000 and 4326 observations for these variables respectively. As these are control variables, and some data unavailability is bound to occur when you are combining different data sources like I am in this study, I do not think this is a serious problem. For a short summary of the variables and their sources, see Appendix I.
Table III
Descriptive Statistics of the diversity indices, ESG total and pillar scores and control variables.
N Mean Median
Std.
Dev. Skewness Kurtosis Min Max
Foreign diversity 4336 0.11 0.00 0.15 1.02 -0.20 0.00 0.50 Code diversity 4336 0.05 0.00 0.10 2.19 4.49 0.00 0.50 Common foreign diversity 4336 0.07 0.00 0.13 1.61 1.45 0.00 0.50 ESG score 4336 56.26 56.33 17.34 -0.03 -0.82 8.56 97.89 Environment Score 4336 54.77 53.78 22.05 0.07 -1.05 4.90 98.69 Social Score 4336 58.11 58.72 19.83 -0.07 -0.87 4.77 98.93 Governance Score 4336 55.91 57.05 20.17 -0.18 -0.84 7.73 98.81 Board size 4336 9.28 9.00 3.60 1.40 6.22 1.00 40.00 LN Firm age 4000 3.13 3.13 0.97 -0.36 -0.08 0.05 5.09 LN Revenues T-1 4128 22.74 22.75 1.38 -0.44 3.96 11.70 26.91 CEO Duality 4326 0.76
16 regression, this should not cause any issues. I also find quite strong correlations between the total ESG score and the individual Environmental, Social and Governance pillars. This was again to be expected, as the total ESG score is comprised of these 3 individual pillars. I also see that there is correlation between these 3 pillars, indicating that companies that score well in one pillar tend to score better in other pillars as well. This again was to be expected. Correlation between the different CSR scores should not cause any issues because these scores are never used in the same regression. Also, there is some correlation between the natural logarithm of firm age and the CSR scores, the natural logarithm of previous years’ revenues and the CSR scores, and board size and CSR scores. The correlation between CSR scores and diversity indices are relatively small.
Table IV
pair wise correlation of the variables
17
Results
The statistical results of this study are presented in this section. I discuss the outcome of foreign board member diversity first, civil law diversity second and common law diversity last. Table V presents the results of 4 OLS regression of board diversity in terms of nationality on various Corporate Social Responsibility (CSR) performance measures. The first column uses the total ESG Score as a performance measure, where the second, third and fourth columns use the Environmental, Social and governance scores respectively as CSR performance measures. In parentheses below the coefficients cluster adjusted standard errors are shown.
The results in Table V suggest that there is a strong positive relationship between board diversity in terms of nationality, the total ESG score performance and the social score performance of the firm, significant at the 1% level. An increase in board diversity by 1 unit would lead to an increase of 11.381 points on the overall ESG score, while this increase of board diversity would lead to a 11.811 point increase on the social score. The effect on the individual Environmental and Governance scores seems to be positive as well, both at a 5% significance level. A 1 unit increase in board diversity would respectively lead to a 11.188 and 11.143 point increase on the individual pillars. Since the effect of diversity in terms of nationality on the overall ESG score is highly significant and the effect on the individual pillars is significant as well I would argue Hypothesis 1 is supported by these results. The explanation behind these results according to previous literature would be that the foreign board members broaden the perspective of the board and make the board more aware of the different needs that the (global) stakeholders might have. This leads to the board making decisions more in line with the stakeholders’ needs and subsequently increases CSR scores. This effect applies to all CSR dimensions, including Environmental, Social and Governance.
18 Table V:
Corporate Social Responsibility scores and nationality diversity in US boards from 2002 to 2016, Fixed Effects OLS
This table displays the output from a Fixed Effects Ordinary Least Squares regression. The sample includes 4132 observations from 362 S&P 500 companies over the time period between 2002 and 2016. Data regarding board nationality diversity was retrieved from the WRDS BoardEX database, while data regarding the dependent variables “ESG score”, “Environmental Score”, “Social Score” and “Governance score” was retrieved from Thomson Reuters EIKON. The column “ESG score” denotes the total CSR score, “Environmental score” denotes the CSR score regarding the environmental pillar, “Social score” denotes the CSR score regarding the social pillar and “Governance score” denote the CSR score regarding the governance pillar. The row “Foreign Diversity” denotes the Blau’s diversity index score regarding board member foreignism, “Board size” denotes the total number of observations per board, “LN Firm age” denotes the natural logarithm of the age in years of the firm at each time t, “LN Revenues T-1” denotes the natural logarithm of the previous years’ (T-1) revenues of the firm, “CEO Duality” is a dummy variable denoting whether the CEO of the firm is also the chairman of the board, it takes 1 if there is CEO duality and 0 if there is no CEO Duality. *, ** and *** indicate statistical significance at the 10%, 5% and 1% level respectively, with clustered standard errors being displayed below the different variable and constant coefficients.
(1) (2) (3) (4)
VARIABLES ESG Score Environmental Score Social Score Governance Score
19 Table VI:
Corporate Social Responsibility scores and civil law diversity in US boards from 2002 to 2016, Fixed Effects OLS
This table displays the output from a Fixed Effects Ordinary Least Squares regression. The sample includes 4158 observations from 364 S&P 500 companies over the time period between 2002 and 2016. Data regarding board nationality diversity was retrieved from the WRDS BoardEX database, while data regarding the dependent variables “ESG score”, “Environmental Score”, “Social Score” and “Governance score” was retrieved from Thomson Reuters EIKON. The column “ESG score” denotes the total CSR score, “Environmental score” denotes the CSR score regarding
the environmental pillar, “Social score” denotes the CSR score regarding the social pillar and “Governance score” denote the CSR score regarding the governance pillar. The row “Civil Diversity” denotes the Blau’s diversity index score regarding board member civil law country origin, “Board size” denotes the total number of observations per board,
“LN Firm age” denotes the natural logarithm of the age in years of the firm at each time t, “LN Revenues T-1” denotes the natural logarithm of the previous years’ (T-1) revenues of the firm, “CEO Duality” is a dummy variable denoting whether the CEO of the firm is also the chairman of the board, it takes 1 if there is CEO duality and 0 if there is no CEO Duality. *, ** and *** indicate statistical significance at the 10%, 5% and 1% level respectively, with clustered standard
errors being displayed below the different variable and constant coefficients.
(1) (2) (3) (4)
VARIABLES ESG Score Environmental Score Social Score Governance Score
20 Table VI presents the results of 4 OLS regression of board diversity in terms of civil on various Corporate Social Responsibility (CSR) performance measures. The first column uses the total ESG Score as a performance measure, where the second, third and fourth columns use the Environmental, Social and governance scores respectively as CSR performance measures. In parentheses below the coefficients cluster adjusted standard errors are shown. The result for the total ESG score remains significant, however only at a 5% level, whereas it was significant at a 1% level for foreign diversity. The coefficient indicates that a 1 unit increase in civil law diversity would lead to a 15.687 unit increase in ESG score. Some of the individual pillar scores have also changed. The Environmental score is now significant at a 1% level, while in the foreign diversity index it was significant only at a 5% level. the diversity effect on Social score is significant at a 10% level, while it was highly significant with regards to the foreign diversity index. Another difference occurs with regards to the governance pillar: whereas it was significant at a 5% level with regards to foreign diversity, civil law diversity results are not significant. This supports hypothesis 2.
21 Table VII:
Corporate Social Responsibility scores and common law diversity in US boards from 2002 to 2016, Fixed Effects OLS
This table displays the output from a Fixed Effects Ordinary Least Squares regression. The sample includes 4132 observations from 362 S&P 500 companies over the time period between 2002 and 2016. Data regarding board nationality diversity was retrieved from the WRDS BoardEX database, while data regarding the dependent variables “ESG score”, “Environmental Score”, “Social Score” and “Governance score” was retrieved from Thomson Reuters EIKON. The column “ESG score” denotes the total CSR score, “Environmental score” denotes the CSR score regarding
the environmental pillar, “Social score” denotes the CSR score regarding the social pillar and “Governance score” denote the CSR score regarding the governance pillar. The row “Common Diversity” denotes the Blau’s diversity index score regarding board member civil law country origin, “Board size” denotes the total number of observations per board, “LN Firm age” denotes the natural logarithm of the age in years of the firm at each time t, “LN Revenues T-1” denotes
the natural logarithm of the previous years’ (T-1) revenues of the firm, “CEO Duality” is a dummy variable denoting whether the CEO of the firm is also the chairman of the board, it takes 1 if there is CEO duality and 0 if there is no CEO Duality. *, ** and *** indicate statistical significance at the 10%, 5% and 1% level respectively, with clustered standard
errors being displayed below the different variable and constant coefficients.
(1) (2) (3) (4)
VARIABLES ESG Score Environmental Score Social Score Governance Score
22
Conclusion
Corporate Social Responsibility and board diversity are nowadays important issues for publicly held firms. Governments are urging for more diversity in boardrooms, and social pressure from the public seems to incentivise companies to behave more socially responsible. While there is little to no regulatory or social pressure to diversify boardrooms in terms of nationality, in this study I observe that companies are diversifying their boards in this dimension. I argue this increase is due to the potential advantage a diverse board in terms of nationality has in being able to recognize and satisfy the needs of a globalizing stakeholder audience. I measure board diversity using three different indices: foreign versus domestic, common law versus civil law and US common law versus non-US common law. I use the Thomson Reuters’ ESG scores and its three individual pillars: Environmental, Social and governance to measure CSR performance.
Our results indicate that foreign board diversity is generally positively associated with CSR performance, this is consistent with previous research. More specifically, I find that foreign board diversity is positively associated with the Environmental and Governance pillars individually, as well as the total ESG score. I expect the impact of civil law diversity to be less on governance as opposed to diversifying in terms of nationality, and find results that suggest this holds true. While the total ESG score and the Environmental and Social pillars are statistically significant while testing civil law diversity, the Governance pillar no longer has a significant positive relationship to diversity.
Still, I observed from our foreign diversity index that diversity has a positive effect on the Governance pillar, so what is causing this effect? I also tested non-US common law diversity, and I found this diversity is strongly positively associated with the Governance pillar.
23 CSR attribute they would want to improve on and diversify their board accordingly having this goal in mind.
There are limitations to this research of course. First of all, the data availability of the nationalities of the board members was quite incomplete. Of course, considering these are directors from some of the largest firms on the planet, you would be able to manually search their nationalities, but for this study I did not do so due to time constraints. Another limitation is the fact that I only used US firms, and therefore I should be careful in applying conclusions from the results I have obtained in this study to firms from other countries.
24
References
Badulescu, A., Badulescu, D., Saveneanu, T., Hatos, R., 2016. SMEs and social responsibility: does firm age matter? Opportunities and Risks in the Contemporary Business Environment, 963-975.
Bear, S., Rahman, N., Post, C., 2010. The impact of Board Diversity and Gender Composition on Corporate Social Responsibility and Firm Reputation. Journal of Business Ethics 97, 207-221.
Benson, B. W., Davidson, W. N., 2010. The relation between stakeholder management, firm value, and CEO compensation: a test of enlightened value maximization. Financial Management 39, 929-963.
Blau, P.M., 1977. Inequality and heterogeneity: A primitive theory of social structure. New York Free Press, New York.
Carter, D.A., Simkins, S.J., Simpson, W.G., 2003. Corporate governance, board diversity, and firm value. Financial Review 38, 33-53.
Freeman, R.E., 1984. Strategic management: A stakeholder approach. Boston: Pitman, 46.
Harjoto, M.A., Laksmana, I., Lee, R., 2015. Board diversity and Corporate Social Responsibility. Journal of Business Ethics 132,641-660.
Hillman, A. J., Keim, G. G., 2001. Shareholder value, stakeholder management, and social issue: What’s the bottom line? Strategic Management Journal 22, 125-139.
Jo, H., Harjoto, M.A., 2011. Corporate Governance and Firm Value: The Impact of Corporate Social Responsibility. Journal of Business Ethics 103(3), 351-383.
Mak, Y.T., Kusnadi, Y., 2004. Size really matters: Further evidence on the negative
relationship between board size and firm value. Pacific-Basin Finance Journal 13(3), 301-318.
25 Oxelheim, L., Randøy, T., 2003. The impact of foreign board membership on firm value. Journal of Banking & Finance 27, 2369-2392.
Post, C., Rahman, N., Rubow, E., 2011. Green Governance: Boards of Directors’ Compostion and Environmental Corporate Social Responsibility. Business & Society 50, 189-223.
Udayasankar, K., 2008. Corporate social responsibility and firm size. Journal of Business Ethics 83, 167-175.
Werther, W.B., Chandler, D., 2010. Strategic Corporate Social Responsibility: Stakeholders in a Global Environment. SAGE publications, London 2, 20
26
Appendix
APPENDIX I: a summary of the different variables, their respective sources and a brief explanation
Variable Source Explanation
Board size BoardEx The number of board
member observations per organization
ESG score Thomson Reuters EIKON
The official ESG (Environment, Social, Governance) score of the firm
Social Score Thomson Reuters EIKON
The official social pillar score of the firm Governance Score Thomson Reuters
EIKON
The official governance pillar score of the firm Environment Score Thomson Reuters
EIKON
The official environment pillar score of the firm CEO duality Thomson Reuters
EIKON
Dummy variable which indicates whether the CEO of a company is also the chairman of the supervisory board (1 if so) or is not (0 if so)
LN revenues T-1 Thomson Reuters EIKON
The natural logarithm of the previous years’ revenues in US$.
LN age Thomson Reuters
EIKON
