Corporate Social Responsibility and Financial Performance:
Analyzing the Impact of the Level of Femininity
in the Country
Abstract
This thesis explores the relationship between Corporate Social Responsibility and financial performance of a company in the context of the femininity dimension of national culture. My sample cover the period 2002-2017 and consists of 35,367 observations. The data is retrieved from Datastream and the Gender Data Portal of the World Bank Group. I find that CSR performance positively affects the financial performance. Furthermore, I confirm the hypothesis that the above relationship is positively affected by country’s femininity.
Keywords: corporate social responsibility, financial performance, femininity, cultural dimensions
Student number: s3757188 Name: Zuzanna Cyganowska Study Programme: MSc IFM
1
1. Introduction
Nowadays, the pressure exercised on a company to be more socially and environmentally conscious and to conduct its operations in an ethical way increases as people become more aware of the environmental and social issues. Corporate social responsibility (CSR) arose already in the 1950s, however its importance and definition has expended throughout the years (Kolk, 2016). CSR can be broadly defined as the company’s responsibility to act in a sustainable manner and respect the interests of all its stakeholders and it includes economic, social, and environmental activities (Kolk, 2016). We observe the departure from the neoclassical economist perception that the sole responsibility of a company is maximizing its profits and shareholder wealth (Friedman, 1970) in favor of shared value creation. Creating shared value was defined by Porter and Kramer (2011), pp 64 as: “creating economic value in a way that also creates value for society by addressing its needs and challenges”. The authors argue that it is crucial for the company to respect the needs of the society in order to gain the support for its operations. They also expect that shifting the focus from profit maximization to shared value creation will provoke innovation and growth in the global economy. The companies, however, are still primarily evaluated based on their financial performance which signifies its outmost importance. As CSR is becoming a wide-spread practice and the number of companies engaged in it is increasing worldwide, the question as to what extent it impacts the financial performance of a company arises.
2
(2017) acknowledge that culture can act as a moderator on the link between CSR and financial performance as it influences the effectiveness of CSR activities realization. Culture influences the way people think and react to the outside world. Based on the above assumption, Swaidan (2012) conducted a research on the effect of national culture and consumers’ ethics and behaviors and found a significant relationship between them. Since culture influences ethics and choices, it can also influence the reactions to CSR. The main aim of the thesis is to find out whether the femininity/ masculinity dimension of the national culture impacts the relation between CSR and financial performance of the company. I expect that the pressure to conduct ethically as well as the reactions to CSR differ across countries, especially between the masculine and feminine countries. Specifically, if it is proven that country’s femininity affects a relationship between a company’s CSR engagement and its financial performance, it would constitute a valuable piece of information for managers who could adjust company’s CSR policy accordingly. This knowledge would be especially important for managers working abroad as it would allow them to better grasp the culture of the country they are working in. The above leads to the research question stated as follows:
Does country’s femininity/masculinity affect the relationship between CSR and financial performance of a company?
3
2. Literature review
2.1 Corporate social responsibility and financial performance of a company
There is a lack of consensus in the literature as to the impact of CSR on financial performance. Horváthová (2010) attributes the discrepancies in the empirical results of the studies on the subject to different methodology adapted by the researchers and to the sample selection. Blasi, Caporin, and Fontini (2018) conducted a multidimensional study on the impact of CSR on financial performance. They implemented both the accounting and the market indicators of financial performance, i.e. ROA, ROE, ROS, ROI, total stock return and financial risk. The authors acknowledge that accounting indicators contain only historical information and can suffer from biases and different accounting practices, whereas market- based indicators represent only the expectations of financial stakeholders and can be biased due to information asymmetry or market distortion. The authors also distinguish seven different aspects of CSR, i.e. environmental, community, human rights, employee relations, diversity, product and governance. The empirical analysis shows that total stock return is highly and positively correlated with CSR while financial risk is negatively correlated. As for the accounting- based indicators, the result are less univocal for ROA and ROS together with ROE being only correlated with the diversity aspect and ROI being negatively correlated with virtually all the categories of CSR.
4
to ameliorate the communication between different stakeholders and lessen conflicts between them. Their empirical results show positive link between CSR and firm value. The above findings are consistent with the study on CSR and organizational processes and performance by Eccles, Ioannou, and Serafeim (2014) in which the authors argue that highly sustainable firms find it crucial to understand and address the needs of its stakeholders. These practices lead to the creation of long-term relationships based on cooperation and trust.
The focus on CSR is becoming even more substantial as the environmental awareness of shareholders is constantly increasing. Flammer (2013) finds that shareholders react to CSR performance of the companies and it is reflected in the stock prices which rise when the company is reported to act responsibly and decrease when misconduct is reported. Additionally, CSR is more often considered a norm of behavior as the plunge of stock prices in reaction to environmentally harmful events has intensified and the rise of stock prices resulting from environmentally friendly events has weakened. The link between environmental CSR and financial performance is further explored in the study by Lioui and Sharma (2012). The authors argue that CSR items constitute a cost for the company and therefore negatively impact financial performance. Their empirical findings point at the direct negative impact of environmental CSR on financial performance, however they find that CSR has a positive impact on R&D efficiency and through that can improve the financial performance. Nevertheless, the above study encompasses only the environmental aspect of CSR without accounting for the entire, more diverse nature of CSR activities (as described by Blasi, Caporin, and Fontini, 2018).
5
intrinsic motivation and compassion at work. It indicates that CSR is valued by the employees and has indirect positive impact on their performance. Higher creativity of employees translates to innovation and can lead to positive financial outcomes. Furthermore, in a study conducted in Scandinavia, Vidaver-Cohen and Brønn (2015) prove that CSR performance is positively associated with company’s reputation and stakeholder support. On average, people have more trust in companies with strong CSR performance and were more willing to buy their products and services. These findings are consistent with a study on consumers’ responses to CSR conducted by Fatma, Khan, and Rahman (2018) which shows that CSR positively affects customer satisfaction and brand loyalty through enhanced identification with the company. Additionally, the study of Vidaver-Cohen and Brønn (2015) shows that CSR has positive impact on respondents’ eagerness to work in a company or invest in it. Taking all the above into consideration, my first hypothesis is as follows:
H1: Corporate social responsibility has a positive influence on the financial performance of
a company.
2.2 Femininity of a country and reactions to CSR
6
implement “outside-in” approach meaning they first consider the needs of the society and try to align their operations accordingly as opposed to the “inside-out” approach prevalent e.g. in the USA where companies only having identified their own interests try to interact them with the society (Strand, Freeman, and Hockerts, 2015).
The roots of the outmost importance of CSR in the Scandinavian countries compared to other countries can be found in the culture. Denmark, Norway, and Sweden score really low on the masculinity dimension of Hofstede’s cultural dimensions with values of 16, 8 and 5 respectively which indicates that they are among the most feminine countries in the world (Hofstede, 1998). Hofstede (1998) defines masculinity as a society which focuses on material success and femininity as society in which the focus on the quality of life prevails. Masculine societies pursue ego-goals, meaning they value money, competition and performance, whereas feminine societies pursue social-goals, meaning they promote equality and solidarity and are more people and relationship-oriented. There is a high level of sympathy for the weaker and caring for others in the feminine societies and the quality and balance in the work life are more desirable than achieving material success and being better than others. Furthermore, the traditional gender roles are more evident in the masculine societies with men pursuing career goals and providing for the family and women taking care of the house and children. These differences are less pronounced in feminine societies which are deemed to be more egalitarian (Hofstede, 1998).
7
been widely discussed as a source of competitive advantage of the company as it leads to higher productivity, higher managerial efficiency and lower turnover rates of the employees (Gupta, and Sharma, 2016). Additionally, interpersonal relationships in the workplace foster the feeling of meaningfulness which is associated with enhanced psychological safety and positively affects engagement (Kahn, 1990). Although femininity may positively impact the financial performance through channels mentioned above, the fact that people in feminine societies are less performance-oriented indicates that the overall effect of femininity on the financial performance will be negative.
It is found that the feminine cultures strongly support CSR performance compared to the masculine cultures which is in line with the emphasis on the overall quality of life and cooperation in these cultures (Strand, Freeman, and Hockerts, 2015). Moreover, culture has a moderating effect on the relationship between CSR and employees’ affective organizational commitment which in turn positively affects employees’ performance (Mueller, Hattrup, Spiess, and Lin-Hi, 2012). Author of the above study find that the relationship between CSR and affective organizational commitment is stronger in cultures with high human orientation as opposed to cultures with high performance orientation. As feminine cultures are more human oriented, it indicates that employees in more feminine societies are more responsive to CSR performance of their companies.
8
The main aim of this thesis is to research how femininity moderates the relationship between CSR and financial performance by influencing how stakeholders react to CSR. As it was previously stated, people in feminine societies value CSR more so it can be assumed that they support the companies that are socially responsible more than people in less masculine societies would which lead to additional advantages of good CSR performance of companies based in feminine countries. Nevertheless, femininity may not only affect the way CSR translates into financial performance but also the CSR itself. Companies do not operate in a vacuum and the national culture has a big influence on both the organizational culture and the individuals within the company who make decisions concerning CSR (Cray, and Mallory, 1998). For the above reason, the values embedded in the national culture can translate to the actions of the company, i.e. a company based in a feminine country can focus more on CSR and perform better in this area because the social goals that are important in the feminine culture are ingrained in the mission and philosophy of the company. As both CSR and financial performance of the company are influenced by femininity, the endogeneity bias arises while testing the hypotheses stated in this thesis. This issue will be further discussed in Section 3.
As CSR is deeply embedded in feminine countries, it can be expected that ceteris
paribus companies that engage in CSR activities are more rewarded in the feminine than in
the masculine countries. Furthermore, a company that is not socially responsible would be more criticized by a feminine society and it is possible that it would have implications for its financial performance. Based on the above, my second hypothesis is as follows:
H2: Femininity of the society has a positive effect on the relationship between CSR and
9
3. Methodology and data
I apply an OLS panel regression using companies from multiple countries. The dependent variable is financial performance. I follow Lioui and Sharma (2012) and Buchanan, Cao, and Chen (2018) and define financial performance as Tobin’s Q calculated as market value of equity plus liquidating value of preferred stock plus book value of debt minus balance sheet deferred taxes divided by total assets.
The independent variable of interest is CSR. In choosing a measure of CSR, I follow Cheng, Serafeim, and Ioannou (2014) and Eccles, Ioannou, and Serafeim (2014) and use the ESG score (TRESGS) provided by Thomson Reuters Asset 4 database which is available via Datastream. The score contains information on the environmental and social performance of the company as well as its corporate governance. It is constructed using over 400 different measures from the areas of: resource use, emissions, innovation, management, shareholders, CSR strategy, workforce, human rights, community, and product responsibility. The score was first introduced in 2002 and the number of firms covered by the rating is growing each year. Currently over 7000 firms globally are included. The ESG score takes values between 0 and 100 and the higher the score, the better the CSR performance of the company. Moreover, Thomson Reuters provides another score, i.e. ESG combined score (TRESGSC) which is calculated based on the ESG score and the ESG controversies score. If the company was involved in some controversies in the last period, it is penalized and the ESG combined score is lower than the basic ESG score.1 As the controversies in the area of CSR often attract the public attention, I want to check how they affect the financial performance of a firm and therefore I repeat the empirical analysis with the ESG combined score as the main independent variable instead of ESG score.
1 Based on: ESG scores methodology [PDF File]. February 2019. Retrieved from
10
A significant problem arises when trying to operationalize the second variable of interest, i.e. femininity. Geert Hofstede described four dimensions of national culture, i.e. individualism-collectivism, uncertainty avoidance, power distance, and masculinity-femininity, and calculated the scores for individual countries based on extensive interviews conducted in years 1967-1973. He published the results of his research together with the theory of cultural dimensions in 1980 (Hofstede, 1980). Due to the workload it requires as well as the assumption that the national culture does not change that quickly, he did not measure the scores on the cultural dimension yearly. He extended the list of the countries included in the study as well as added additional dimensions, i.e. long-term orientation and indulgence, in the publications of 2001 and 2010 respectively (Hofstede, 2001; Hofstede, Hofstede, and Minkov, 2010) . As the result, there is no consistent annual data available in order to successfully conduct the analysis of the moderating effect of culture on the relationship between CSR and financial performance.
11
Consequently, women are less likely to pursue career and focus on professional life in a masculine society. Instead they are expected to focus on the family (Hofstede, 1998). Therefore, the number of women in workforce as compared to the number of men in workforce is lower in masculine societies. Based on the above, I proxy the femininity of a country by constructing the femininity variable as a ratio of labor force participation rate of the female population of the country to the labor force participation ratio of the male population of the country, where higher value of the variable indicates that the national culture is more feminine. As the main aim is to examine how the femininity of a country influences the relationship between CSR and the financial performance, I implement an interaction of the CSR and femininity variables in the model.
As it was mentioned in Section 2.2 of this thesis, femininity can influence both financial performance and CSR. As both financial performance and CSR contain an element that is dependent on femininity, it leads to the correlation of CSR which is the explanatory variable with the error term. As the result, CSR is not truly exogenous and it is difficult to assign causality, i.e. state that it is CSR that affects financial performance and not the opposite. One way of fixing the problem of endogeneity would be to separate the variance in the CSR variable that is exogenous from the financial performance by finding an Instrumental Variable (IV) which is highly correlated with CSR but uncorrelated with the error term. Nevertheless, finding a good IV is extremely difficult as it not only needs to meet the above requirements but it also has to be rooted in theory. Endogeneity results in biased estimates and causes the results of the empirical analysis to be less believable, therefore, I acknowledge that the causal relationship cannot be truly confirmed without solving the endogeneity bias.
12
expenditures to total assets ratio, fixed assets to total assets ratio, leverage calculated as total debt to total assets ratio, cash holdings calculated as the ratio of cash and short-term investment to total assets and R&D intensity calculated as R&D expenses to total assets. If the value for the R&D expenses is missing, I set the variable to zero (Buchanan, Cao, and Chen, 2018). I also control for industry and year fixed effects by creating dummies for each industry classified by 2-digit SIC code and each year (Cheng, Serafeim, and Ioannou, 2014). Industry fixed effects help control for the within industry correlation but not for the additional within firm correlation. For the above reason, I cluster the standard errors at the firm level in order to control for the within firm error correlation as the failure to do so may result in misleadingly small standard errors and large t-statistics.
In order to test the hypotheses, the following specification is used:
Firm performance,i,t = α0 + β1 x TRESGSi,t-1 + β2 x Femininityc,t-1 + β3 x TRESGSi,t-1 x
Femininityc,t-1 + Firm-level controls + Industry F.E. + Year F.E.
+ εi,t (1)
where i is a subscript for each firm, t is a subscript for each year and c is a subscript for each country. In order to control for reverse causality, both CSR variable and femininity variable are lagged by one year (t-1).
13
the indices are reviewed on quarterly bases and all the newly included companies are added to the coverage. The data coverage starts in 2002 with the companies from the following indices: SMI, DAX, CAC 40, FTSE 100, FTSE 250, S&P 500, NASDAQ 100. Starting in 2008, more indices have been included, i.e. DJ STOXX, MSCI World, S&P/TSX, COMPOSITE, Russell 1000, MSCI Emerging Markets, Bovespa, S&P ASX 300, S&P NZX 50, Russell 2000*, IPC 35, IPSA 40, MERVAL, COLCAP, PERU GENERAL INDEX, MSCI Emerging Markets - China (179 new companies). As of February 2019, Thomson Reuters reported being in the process of adding the Russell 3000 Index companies as well. Over 70% of global market capitalization is covered in the ESG data set. The ESG score is data-driven and Thomson Reuters ensures that the firm size and transparency biases are minimal
14
Table 1 Variable description
Variable Description
Tobin’s Q (Market value of equity + Liquidating value of preferred stock + Book value of debt - Balance sheet deferred taxes) / Total assets
TRESGS The environmental, social and governance score for a firm as provided by Thomson Reuters
TRESGCS The environmental, social and governance combined score including controversies for a firm as provided by Thomson Reuters
Femininity The female employment ratio to male employment ratio of the population in a country in which a firm is based
Firm size Logarithm of total assets
Sales growth Net sales in year t to net sales in year t-1 ratio Capex Capital expenditures to total assets ratio Fixed assets Fixed assets to total assets ratio
Leverage Total debt to total assets ratio
Cash Cash and short-term investment to total assets ratio R&D R&D expenditures to total assets ratio
This table describes the variables used in the empirical analysis. All the data is retrieved from Datastream with the exception of the data on employment used to calculate the Femininity variable which is retrieved from the Gender Data Portal of the World Bank Group.
15
Additionally, I winsorize the firm-level data on the 1% level to diminish the effect of outliers. Table 1 presents the descriptions of the variables used in the analysis and table 2 presents the summary statistics for the variables. It shows the mean, standard deviation, minimum and maximum value for each of the variables.
Table 2 Summary statistics
Variable Obs Mean Std. Dev. Min Max
Tobin’s Q 35,367 2.238 1.259 1.049 8.294 TRESGS 35,367 50.524 17.301 0 97.89 TRESGCS 35,367 45.33 15.807 0 95.59 Femininity 35,367 0.776 0.092 0.233 0.913 Firm size 35,367 15.433 1.426 11.962 18.954 Sales growth 35,367 0.078 0.243 -0.509 1.224 Capex 35,367 0.055 0.182 0 32.234 Fixed assets 35,367 0.605 0.208 0.089 0.959 Leverage 35,367 0.256 0.177 0 0.78 Cash 35,367 0.134 0.127 0.001 0.633 R&D 35,367 0.016 0.033 0 0.181
This table reports the summary statistics of the the key variables. Tobin’s Q is calculated as market value of equity plus liquidating value of preferred stock plus book value of debt minus balance sheet deferred taxes divided by total assets. TRESGS is the environmental, social and governance score for a firm as provided by Thomson Reuters. TRESGCS is the environmental, social and governance combined score including controversies for a firm as provided by Thomson Reuters. Femininity is the femininity of a country in which a company is based defined as the female employment ratio to male employment ratio of the population in this country. Firm size is defined as the logarithm of total assets in USD. Sales growth is the sales growth rate calculated as net sales in year t divided by net sales in year t-1 minus one. Capex is capital expenditures to total assets ratio. Fixed asset is the fixed assets to total assets ratio. Leverage is calculated as the total debt to total assets ratio. Cash is the ratio of cash and short-term investment to total assets. R&D is R&D intensity defined as R&D expenditures to total assets ratio. I provide mean, standard deviation, minimum and maximum. The data period is 2002-2017.
16
Table 3 Observations by country
Country Freq. Percent Cum.
Argentina 24 0.07 0.07 Australia 2,176 6.15 6.22 Austria 142 0.40 6.62 Belgium 241 0.68 7.30 Brazil 544 1.54 8.84 Canada 2,298 6.50 15.34 Chile 162 0.46 15.80 China 567 1.60 17.40 Colombia 50 0.14 17.54 Cyprus 1 0.00 17.54 Czech Republic 27 0.08 17.62 Egypt 55 0.16 18.50 Finland 323 0.91 19.41 France 1,055 2.98 22.39 Germany 946 2.67 25.07 Greece 183 0.52 25.59 Hong Kong 1,180 3.34 28.92 Hungary 24 0.07 28.99 India 525 1.48 30.47 Indonesia 196 0.55 31.03 Ireland 146 0.41 31.44 Israel 91 0.26 31.70 Italy 431 1.22 32.92 Japan 4,323 12.22 45.14 Kenya 2 0.01 45.15 Kuwait 3 0.01 47.14 Luxembourg 35 0.10 47.24 Malaysia 300 0.85 48.08 Mexico 211 0.60 48.68 Morocco 14 0.04 48.72 Netherlands 369 1.04 49.76 New Zealand 186 0.53 50.29 Norway 272 0.77 51.06 Oman 6 0.02 51.08 Peru 38 0.11 51.18 Philippines 125 0.35 51.54 Poland 139 0.39 51.93 Portugal 105 0.30 52.23 Qatar 6 0.02 52.24 Russia 260 0.74 52.98
17
Table 3 (continued)
Country Freq. Percent Cum.
Saudi Arabia 27 0.08 53.06 Singapore 422 1.19 54.25 Spain 455 1.29 55.53 South Korea 701 1.98 47.13 Sri Lanka 7 0.02 55.55 Sweden 570 1.61 57.17 Switzerland 608 1.72 58.89 Thailand 164 0.46 59.35 Turkey 143 0.40 59.75
United Arab Emirates 24 0.07 59.82
United Kingdom 3,428 9.69 69.51
United States 10,782 30.49 100.00
Total 35,367 100.00
The table presents the tabulation of the observations by country. The data is for period 2002-2017.
18
Table 4 Femininity by country
Country Mean Min Max SD Range
Argentina 0.648 0.626 0.662 0.008 0.036 Australia 0.809 0.769 0.840 0.021 0.070 Austria 0.793 0.735 0.832 0.032 0.097 Belgium 0.771 0.706 0.814 0.035 0.108 Brazil 0.700 0.675 0.713 0.011 0.038 Canada 0.856 0.827 0.871 0.014 0.045 Chile 0.600 0.483 0.680 0.070 0.197 China 0.825 0.808 0.848 0.012 0.039 Colombia 0.657 0.553 0.712 0.055 0.159 Czech Republic 0.740 0.724 0.761 0.012 0.037 Denmark 0.862 0.832 0.886 0.017 0.053 Egypt 0.292 0.263 0.310 0.015 0.046 Finland 0.877 0.859 0.891 0.010 0.033 France 0.817 0.773 0.841 0.019 0.068 Germany 0.791 0.745 0.832 0.029 0.087 Greece 0.690 0.624 0.751 0.041 0.127 Hong Kong 0.759 0.713 0.796 0.030 0.083 Hungary 0.739 0.716 0.756 0.011 0.040 India 0.380 0.337 0.443 0.040 0.106 Indonesia 0.602 0.586 0.620 0.012 0.034 Ireland 0.750 0.693 0.788 0.033 0.094 Israel 0.820 0.777 0.858 0.025 0.081 Italy 0.643 0.597 0.678 0.027 0.081 Japan 0.676 0.646 0.716 0.023 0.070 Kenya 0.885 0.862 0.913 0.018 0.051 Kuwait 0.690 0.663 0.714 0.015 0.051 Luxembourg 0.568 0.547 0.579 0.010 0.032 Malaysia 0.748 0.643 0.823 0.055 0.180 Mexico 0.597 0.564 0.657 0.037 0.092 Morocco 0.530 0.474 0.559 0.028 0.085 Netherlands 0.338 0.321 0.349 0.007 0.028 New Zealand 0.806 0.746 0.838 0.028 0.092 Norway 0.824 0.786 0.853 0.022 0.067 Peru 0.884 0.861 0.901 0.015 0.040 Philippines 0.807 0.752 0.835 0.024 0.083 Poland 0.635 0.608 0.660 0.017 0.052 Portugal 0.754 0.742 0.775 0.010 0.033 Qatar 0.814 0.764 0.838 0.023 0.074 Russia 0.543 0.468 0.620 0.055 0.152
19
Table 4 (continued)
Country Mean Min Max SD Range
Saudi Arabia 0.804 0.789 0.822 0.011 0.033 Singapore 0.250 0.213 0.280 0.021 0.067 South Korea 0.730 0.659 0.788 0.044 0.129 Spain 0.744 0.635 0.818 0.064 0.183 Sri Lanka 0.471 0.433 0.483 0.014 0.050 Sweden 0.886 0.874 0.902 0.009 0.028 Switzerland 0.811 0.775 0.848 0.024 0.073 Thailand 0.797 0.783 0.815 0.009 0.032 Turkey 0.389 0.332 0.450 0.043 0.118
United Arab Emirates 0.436 0.382 0.462 0.028 0.080
United Kingdom 0.806 0.777 0.834 0.019 0.057
United States 0.813 0.801 0.822 0.007 0.022
Mean 0.695 0.653 0.728 0.025 0.075
This table reports the mean, minimum, maximum, standard deviation and range of the femininity variable by country. The data is for period 2002-2017.
20
Table 5 Correlation table
21
4. Empirical findings
4.1 Main results
Table 6 presents the results of the OLS regression with Tobin’s Q as the dependent variable. The first column presents the results of regression using TRESGS as the main independent variable. It shows that CSR performance of a firm has a small, positive and highly significant effect on the financial performance. As Tobin’s Q contains information about shareholders’ expectations, the above results are consistent with the findings of Flammer (2013) that shareholders react to CSR activities and these reactions affect stock prices. The above confirms the hypothesis H1 stating that CSR has a positive impact on the financial performance of the company.
The second column of Table 6 presents the results of the OLS regression that used TRESGCS instead of TRESGS as the measure of CSR performance. Also in this case, there is a small, positive and highly significant effect of CSR performance on the financial performance of a firm which confirms the hypothesis H1 stated in this thesis. Nevertheless, contrary to my expectations, CSR performance when proxied by the ESG combined score has a lower impact on the financial performance than when it is proxied by the ESG score. It implies that the financial performance is affected more by the overall CSR performance along with the idea that CSR controversies do not have a substantial effect on the financial performance.
22
Table 6 Regression results
(1) (2) (3) (4)
Variable TobinsQ TobinsQ TobinsQ TobinsQ
TRESGS 0.00336*** 0.00336*** (0.000600) (0.000596) TRESGCS 0.00239*** 0.00234*** (0.000380) (0.000377) Femininity -0.982*** -1.028*** (0.203) (0.201) TRESGS x Femininity 0.0267*** (0.00581) TRESGCS x Femininity 0.0184*** (0.00379) Firm size -0.365*** -0.357*** -0.372*** -0.362*** (0.0174) (0.0170) (0.0174) (0.0170) Sales growth 0.451*** 0.445*** 0.450*** 0.444*** (0.0240) (0.0239) (0.0240) (0.0238) Capex 0.305 0.307 0.306 0.309 (0.239) (0.241) (0.238) (0.240) Fixed assets -0.499*** -0.506*** -0.471*** -0.482*** (0.103) (0.104) (0.103) (0.104) Leverage -1.076*** -1.076*** -1.086*** -1.087*** (0.0823) (0.0823) (0.0821) (0.0821) Cash 1.374*** 1.369*** 1.384*** 1.373*** (0.146) (0.147) (0.145) (0.146) R&D 2.980*** 3.019*** 3.098*** 3.129*** (0.678) (0.679) (0.679) (0.679) Constant 8.394*** 8.267*** 8.480*** 8.338*** (0.290) (0.285) (0.291) (0.286)
Industry fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Observations 35,367 35,367 35,367 35,367
Number of firms 5,044 5,044 5,044 5,044
23
and significant, whereas the effect of femininity is stronger and negative. Focusing on social goals and wellbeing more than on the performance and material success may not be the only explanation for the negative effect that Femininity has on Tobin’s Q as it can partially be explained by characteristics of investors. In more feminine countries, both males and females participate in making financial decisions as opposed to the situation in masculine countries where gender roles are more pronounced and men handle the finances (Hofstede, 1998). Women are known to be less optimistic about the future and more risk-averse than men which influences their behavior on the stock market (Jacobsen, Lee, Marquering, and Zhang, 2014). Tobin’s Q is higher when the investors’ predictions about the future of the company are positive so when there is more female investors on the market or men make financial decisions in consultation with their female partners, it is expected that the level of optimism will be lower. Based on the above, the negative impact Femininity has on Tobin’s Q might partially result from the unwillingness of investors to invest in stocks which they consider overvalued which disables the market value to vary too substantially from the book value.
As for the interaction between TRESGS and TRESGCS with Femininity, there is a positive and highly significant effect on Tobin’s Q. The above means that for any given level of femininity in the country, CSR positively affects financial performance and that firms that are based in more feminine countries achieve better financial performance than firms characterized by the same level of CSR which are located in less feminine countries ceteris
paribus. The results confirm the hypothesis H2 stating that femininity has a positive impact
24
4.2 Robustness check
The empirical results in section 4.1 confirm the hypothesis H1 stating that CSR has a positive impact on financial performance as well as the hypothesis H2 stating that the above relationship is positively moderated by the level of femininity. In this section, I present the findings obtained from robustness check. Firstly, I check if the results suffer from sample selection bias. As the United Kingdom and the United States together constitute 40,18% of the original sample, there is a chance that the results presented in Table 6 are biased.
Panel A of Table 7 presents the results of the OLS regressions that use the same specifications as the models presented in Table 6 but the two countries mentioned above are excluded from the sample. The effect TRESGS and TRESGCS has on Tobin’s Q in notably smaller than the one presented in Table 6, however, it is still positive and highly significant which confirms the hypothesis H1. The coefficients of Femininity are negative, highly significant and very close to the coefficients obtained in Table 6 where the whole sample was included. As presented in Column 3 and Column 4, the interaction of TRESGS and TRESGCS with Femininity has a positive and highly significant effect on Tobin’s Q which confirms the hypothesis H2 stating that there is a positive moderating effect of femininity on the relationship between CSR and financial performance.
25
are within the confidence intervals of the coefficients showed in Table 6. Both hypothesis H1 and hypothesis H2 are confirmed.
Table 7 Robustness check – sample selection bias
Panel A: With the exclusion of the United Kingdom and the United States
(1) (2) (3) (4)
Variable TobinsQ TobinsQ TobinsQ TobinsQ
TRESGS 0.00167** 0.00221*** (0.000751) (0.000797) TRESGCS 0.00128*** 0.00166*** (0.000479) (0.000504) Femininity -1.004*** -1.027*** (0.217) (0.214) TRESGS x Femininity 0.0190*** (0.00647) TRESGCS x Femininity 0.0146*** (0.00416) Firm size -0.337*** -0.334*** -0.344*** -0.340*** (0.0216) (0.0213) (0.0216) (0.0212) Sales growth 0.376*** 0.374*** 0.378*** 0.374*** (0.0268) (0.0268) (0.0268) (0.0268) Capex 0.253 0.254 0.253 0.254 (0.203) (0.204) (0.201) (0.202) Fixed assets -0.570*** -0.572*** -0.534*** -0.541*** (0.119) (0.119) (0.118) (0.118) Leverage -1.314*** -1.314*** -1.317*** -1.320*** (0.108) (0.108) (0.108) (0.108) Cash 1.243*** 1.240*** 1.258*** 1.245*** (0.178) (0.178) (0.178) (0.177) R&D 1.464 1.479 1.603 1.612 (1.181) (1.183) (1.181) (1.181) Constant 7.980*** 7.928*** 8.055*** 7.993*** (0.383) (0.380) (0.383) (0.379)
Industry fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Observations 21,157 21,157 21,157 21,157
26
Panel B: With the exclusion of the countries with less than 50 observations
(1) (2) (3) (4)
Variable TobinsQ TobinsQ TobinsQ TobinsQ
TRESGS 0.00338*** 0.00333*** (0.000602) (0.000597) TRESGCS 0.00239*** 0.00231*** (0.000382) (0.000378) Femininity -1.091*** -1.132*** (0.213) (0.211) TRESGS x Femininity 0.0281*** (0.00606) TRESGCS x Femininity 0.0189*** (0.00391) Firm size -0.368*** -0.360*** -0.375*** -0.365*** (0.0175) (0.0171) (0.0175) (0.0171) Sales growth 0.452*** 0.447*** 0.452*** 0.445*** (0.0241) (0.0240) (0.0241) (0.0239) Capex 0.303 0.305 0.305 0.307 (0.238) (0.240) (0.237) (0.239) Fixed assets -0.499*** -0.507*** -0.470*** -0.482*** (0.104) (0.105) (0.103) (0.104) Leverage -1.077*** -1.077*** -1.089*** -1.090*** (0.0826) (0.0826) (0.0824) (0.0824) Cash 1.366*** 1.361*** 1.376*** 1.365*** (0.146) (0.147) (0.146) (0.146) R&D 2.969*** 3.008*** 3.087*** 3.116*** (0.678) (0.679) (0.679) (0.679) Constant 8.437*** 8.309*** 8.533*** 8.389*** (0.292) (0.287) (0.293) (0.288)
Industry fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Observations 35,129 35,129 35,129 35,129
Number of firms 4,967 4,967 4,967 4,967
27
Table 8 Robustness check – firm fixed effects
(1) (2) (3) (4)
Variable Tobin’s Q Tobin’s Q Tobin’s Q Tobin’s Q
TRESGS 0.00239*** 0.00227*** (0.000666) (0.000661) TRESGCS 0.00224*** 0.00216*** (0.000395) (0.000390) Femininity -1.828*** -1.970*** (0.587) (0.589) TRESGS x Femininity 0.0308*** (0.00639) TRESGCS x Femininity 0.0213*** (0.00395) Firm size -0.460*** -0.458*** -0.464*** -0.462*** (0.0299) (0.0299) (0.0298) (0.0299) Sales growth 0.441*** 0.439*** 0.438*** 0.436*** (0.0245) (0.0245) (0.0245) (0.0244) Capex 1.720*** 1.720*** 1.714*** 1.720*** (0.310) (0.310) (0.308) (0.311) Fixed assets -0.513*** -0.514*** -0.514*** -0.518*** (0.136) (0.137) (0.136) (0.136) Leverage -1.075*** -1.073*** -1.106*** -1.105*** (0.0939) (0.0938) (0.0932) (0.0931) Cash 1.187*** 1.187*** 1.188*** 1.180*** (0.168) (0.169) (0.168) (0.168) R&D 1.734* 1.747* 1.755* 1.748* (0.917) (0.916) (0.917) (0.914) Constant 9.735*** 9.707*** 9.829*** 9.810*** (0.449) (0.449) (0.448) (0.448)
Firm fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Observations 35,367 35,367 35,367 35,367
R-squared 0.236 0.237 0.239 0.239
Number of firms 5,044 5,044 5,044 5,044
28
The last robustness check that I perform is the OLS regressions with firm fixed effects in order to control for the firm-specific time-invariant factors that may influence the results. As it is showed by the positive and highly significant coefficients of TRESGS and TRESGCS, a positive change in CSR performance positively affects Tobin’s Q. Increase in the level of femininity in the country negatively affects the financial performance whereas increase in both femininity and better CSR performance positively affects the financial performance. The above findings are consistent with the results presented in Table 6.
5. Conclusions
CSR is becoming an inseparable part of modern business practices. The environmental and social awareness of stakeholders is constantly increasing alongside with their expectations towards companies. I find that CSR performance positively affects financial performance of the company. These findings indicate that being socially responsible and acting in a manner that complies with social norms is beneficial for the company as it ameliorates its financial performance. It is visible not only in higher returns on asset and equity but also in higher Tobin’s Q which means that investors value companies with good CSR performance and have positive expectations regarding their future. The financial reward for being socially responsible is especially strong in feminine countries as CSR is in line with the values that feminine societies keep in high regard, i.e. equality, solidarity and quality of life. The positive moderating effect of the country’s femininity on the relationship between CSR and financial performance is visible in higher Tobin’s Q.
29
other stakeholders whose behavior is influenced by CSR, i.e. employees and consumers, and in turn affects the financial performance. The managers should strive to use CSR in creating shared value for all of their stakeholders which will create long-term relationships with them alongside with promoting growth and innovation.
This thesis is subjected to certain limitations due to difficulties in adequately measuring the concept of femininity. I acknowledge that the measure of femininity that I implement in this thesis does not account for all the elements of national culture that are included in the masculinity/femininity dimension of Hofstede as the gap in level of employment are only one of the many components that Hofstede took into account when constructing this dimension. Nevertheless, the level of employment can be used as an approximation of femininity in the national culture as it usually reflects the division of gender roles and the prevalent ideology regarding the social goals. Furthermore, I acknowledge that the results may be subject to endogeneity bias.
30
References:
Buchanan, B., Cao, C., Chen, C. 2018. Corporate social responsibility, firm value, and influential institutional ownership. Journal of Corporate Finance, 52, 73-95.
Blasi, S., Caporin, M., Fontini, F. 2018. A multidimensional analysis of the relationship between corporate social responsibility and firms' economic performance. Ecological Economics, 147, 218-229.
Cheng, B., Serafeim, G., Ioannou, I. 2014. Corporate social responsibility and access to finance. Strategic Management Journal, 35(1), 1-23.
Cray, D., Mallory, G. 1998. Making sense of managing culture. 1st Edition. London: International Thomson Business Press.
Eccles, R., Ioannou, I., Serafeim, G. 2014. The impact of corporate sustainability on organizational processes and performance. Management Science, 60, 2835-2857.
Fatma, M., Khan, I., Rahman, Z. 2018. CSR and consumer behavioral responses: The role of customer-company identification. Asia Pacific Journal of Marketing and Logistics, 30(2), 460-477.
Flammer, C. 2013. Corporate social responsibility and shareholder reaction: The environmental awareness of investors. Academy of Management Journal, 56(3), 758-781.
31
Grewatsch, S., Kleindienst, I. 2017. When does it pay to be good? moderators and mediators in the corporate sustainability-corporate financial performance relationship: A critical review. Journal of Business Ethics, 145(2), 383-416.
Guiso, L., Sapienza, P., Zingales, L. 2006. Does culture affect economic outcomes?. Journal of Economic Perspectives, 20(2), 23-48.
Gupta, N., Sharma, V. 2016. Exploring employee engagement—a way to better business performance. Global Business Review, 17, 45-63.
Hofstede, G. 1980. Culture's consequences : International differences in work-related values (Cross-cultural research and methodology series, 5). Beverly Hills, Calif.: Sage Publications.
Hofstede, G. 1998. Masculinity/ femininity as a dimension of culture. In: Hofstede, G. (Ed.), Masculinity and femininity: The taboo dimension of national cultures. Cross-cultural psychology series, vol. 3. Thousand Oaks, London, pp. 3-28.
Hofstede, G. 2001. Culture’s consequences: Comparing values, behaviors, institutions, and organizations across nations. Thousand Oaks, CA: Sage Publications.
Hofstede, G., Hofstede G. J., Minkov, M. 2010. Cultures and organizations: Software of the mind. Revised and Expanded 3rd Edition. New York: McGraw-Hill.
Horváthová, E. 2010. Does environmental performance affect financial performance? a meta-analysis. Ecological Economics, 70(1), 52-59.
32
Jacobsen, B., Lee, J., Marquering, W., Zhang, C. 2014. Gender differences in optimism and asset allocation. Journal of Economic Behavior and Organization: Part B, 107, 630-651.
Kahn, W. 1990. Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, 33, 692−724..
Kemper, L., Bader, A., Froese, F. 2019. Promoting gender equality in a challenging environment. Personnel Review, 48(1), 56-75.
Kolk, A. 2016. The social responsibility of international business: From ethics and the environment to CSR and sustainable development. Journal of World Business, 51(1), 23–34.
Kostis, P., Kafka, K., Petrakis, P. 2018. Cultural change and innovation performance. Journal of Business Research, 88, 306-313.
Lioui, A., Sharma, Z. 2012. Environmental corporate social responsibility and financial performance: Disentangling direct and indirect effects. Ecological Economics, 78, 100-111.
Mueller, K., Hattrup, K., Spiess, S., Lin-Hi, N. 2012. The effects of corporate social responsibility on employees' affective commitment: A cross-cultural investigation. The Journal of Applied Psychology, 97(6), 1186-200.
Porter, M., Kramer, M. 2011. The big idea: Creating shared value. Harvard Business Review, 89(1-2).
Sen, S., Bhattacharya, C. 2001. Does doing good always lead to doing better? consumer reactions to corporate social responsibility. Journal of Marketing Research, 38(2), 225-243.
33
Swaidan, Z. 2012. Culture and consumer ethics. Journal of Business Ethics, 108(2), 201-213.
Van de Vliert, E. 1998. Gender role gaps, competitiveness, and temperature. In: Hofstede, G. (Ed.), Masculinity and femininity: The taboo dimension of national cultures. Cross-cultural psychology series, vol. 3. Thousand Oaks, London, pp. 117-129.
