Do CEO age and gender
impact CSR expenditure
Jelmer van der Ende
Msc International Business and Management
Rijksuniversiteit Groningen
S2168022
Abstract
In 2012 at a global forum about sustainability a key speaker stated that companies no longer need to be the best in the world, but they have to be the best for the world. Due to this change in objective, corporate social responsibility (CSR) is becoming an increasingly important phenomenon in modern day business. But what is the definition of CSR and who decides within a company what the CSR spending has to be?
In recent years a number of studies have been conducted focusing on the definition of CSR and determination of the motives for CSR spending. Furthermore, emphasis is given to the effects of CSR on company value, image and profit. Research with a focus on external characteristics of a companies’ leadership, in particular the impact of the CEO in relation to CSR spending, is however very limited. This paper will touch upon this subject. The main research question in this paper is: “What is the effect of CEO age and gender on CSR
investments within Global Fortune 500 companies”. It is concluded that neither CEO age nor CEO gender has a statistically significant effect on CSR spending.
Key words: CSR, CEO age, CEO gender, Global Fortune 500.
The Global Fortune 500 companies have spend over 20 billion
US$ in 2013, an amount of money that can make a huge
Table of contents
1. Introduction
5
2. Literature review
8
2.1 CSR
8
2.1.1 Defining CSR
8
2.1.2 Motives for CSR
9
2.1.3 Effects of CSR
10
2.1 CSR in relation to CEO’s
12
2.1.1 The Upper Echelon Theory
12
2.1.2 Previous research
13
3
Conceptual model & hypothesis
14
1. Introduction
In 2012 there was a Global Forum about sustainability at Rio de Janeiro. Over 2500 CEOs and high executives discussed the responsibilities of their respective corporations at this Global Forum. One of the key speakers stated that there is an on-going change in the objectives of corporations. Companies no longer need to be the best in the world but they have to be the best for the world, which poses a difficult challenge. Companies have many stakeholders to take into account whilst still needing to be profitable. Due to this change in objective, corporate social responsibility is an increasingly important phenomenon in the modern day business structure.
The transparency of companies has increased a lot over the past decade. Consumers have more information about the origin of the products they buy and the production process
associated with the products (Bandsuch et al., 2008). Because companies are more transparent many consumers opt for more corporate social responsibility in terms of a ‘green’ production process, fair employee working conditions and other social and environmental aspects.
It is however hard to believe that only consumers find corporate social responsibility
important, since other stakeholders, CEO’s and other high-ranking executives are also people with morals and a tendency to do the ‘right’ thing. This leads to believe that CSR is not just a concept pushed by the consumers, but is also affected by CEO’s.
The main objective of every company is to make profit. Profit ensures company survival, which is the primary goal. A company that makes profit often invests a part of these profits to enhance their production process, increase their market share through marketing campaigns, moves abroad for international expansion or other ways to increase company value and with this yearly profits. Though many companies also invest in corporate social responsibility, whilst this is not a production enhancer and the costs associated to corporate social responsibility are often higher than the direct profits it brings.
have looked at certain CEO characteristics and whether these have an influence on the CSR initiatives of a corporation (Petrenko et al., 2015; Wang et al., 2015; Maak et al., 2016).
Most of these studies look at the inner characteristics of a CEO, which a CEO either has or doesn’t have, and link these characteristics to overall or specific CSR practices. Though little research has been done linking the external characteristics of a CEO, for instance age,
education or country of origin, to CSR practices. As a consequence, little is known about the effect of these external characteristics on CSR.
Due to a lack of time and resources, there are only two external characteristics, which will be addressed in this research. The first one is CEO age. It is interesting to see what the
differences are between CEO’s who grew up in a world where CSR was not that important yet and CEO’s that followed CSR courses during their studies and see CSR as an important factor of doing business from the day that they started working.
Besides looking at the CEO age, there is also a possibility that gender differences influence the decision for CSR and overall CSR expenditures. Females are considered having a more nurturing character, caring more for other people than men do broadly speaking. It is in the character of women to take into account other people in their decisions instead of looking just at profits. Naturally, men do not look solely at profits but the degree to which they take into account other people is regarded as lower than the degree that women take other people into account when making decisions. Since CSR is considered something that does not directly affect the profits of a company, but that it is something to help (local) communities or environmental problems, women are expected to have a higher tendency towards CSR than men.
For this paper data is collected of Global Fortune 500 companies, the selected target group.
The main research question addressed in this paper is:
Prior to answering the main research question, a literature review is carried out to address several sub-questions:
1. How can CSR be defined and measured? 2. What are the motives for CSR?
3. What are the effects of CSR?
4. What is the effect of CEO’s on CSR?
5. What is the previous research of CEO effect on CSR?
Based on the literature review, control or dummy variables are identified. The dummies used in this paper are (1) whether a company participates in either business to consumer
transactions or business to business transactions, (2) the business segment in which
companies operate, (3) the number of years a CEO is already working on his or her respective position and (4) The country of origin (the headquarters location) of the company.
This paper will do a statistical study trying to link CEO age and CEO gender to CSR spending. This paper will only use information sources available on the Internet. No interviews will be conducted.
2. Literature review
A literature review will be conducted with the aim to get more insights in the definition of CSR, the motives of CSR, the effects of CSR and the relation between CSR and a companies’ CEO. Furthermore, the literature review will be carried out to answer sub-questions.
2.1 CSR
2.1.1 Defining and measuring CSR
Finding an appropriate universal definition for CSR is a struggle for many researchers because there are many issues which can be included in CSR.
A study by Carroll (1979) identifies four categories of responsibility for corporations, these four categories are:
-‐ The economic responsibility -‐ The legal responsibility -‐ The ethical responsibility -‐ The voluntaristic responsibility
The economic responsibility of corporations includes being profitable and producing goods. The legal responsibility is to follow the law and not breaking (inter)national rules and regulations. Ethical responsibility goes beyond the legal responsibility of firms, it requires firms to do more than abiding the law and is therefore not an unequivocal defined
A study done by Sheehy (2015) argues that it is very difficult to find a universal definition for CSR and defines CSR as an international private business regulator, which is political of nature (Sheehy, 2015). The study by Sheehy (2015) therefore confirms that there is no real universal definition possible for CSR. This is also stated by Okoye (2009). He uses the essentially contested concepts (ECC) theory developed by Gallie (1956) to test whether CSR can have several conceptions instead of one universal definition. The ECC theory points out that certain concepts are through their nature inevitably contested. Okoye (2009) concluded that CSR is one of these concepts, which cannot be universally defined, and that there are multiple conceptualizations possible. These conceptualizations could be on a multicultural context (Sachs et al., 2005), a developing country context (Jamali & Mirshak, 2006), a
specific sector context (O’Riordan & Fairbrass, 2013) or on the identification of the skills and capabilities required for successful CSR implementation and management (Kakabadse et al., 2009). Some researchers try to build a framework, which is more universally applicable (Yuan et al., 2011).
For quantifying the CSR initiatives, this paper makes use of reported CSR investments. As there is no real universal definition of CSR, companies determine themselves what they define and apply as CSR. In addition, many companies do not go into depth what the exact CSR initiatives are that they spent money on. Therefore, it is chosen that to assess the companies’ commitment to CSR, this paper will look at total money spent.
The consequence of only looking at total money spent on CSR is that no distinction can be made between economical, legal, ethical and voluntaristic CSR expenditure.
2.1.2 Motives for CSR
media could also have a large effect on CSR investments from companies. Both NGOs and the media can pressure companies into investing in CSR initiatives motivated by, among other things, negative publicity. Institutional pressures are also mentioned in a paper of Kablak & Trendafilova, 2010.
Besides pressures from outside the company, there could also be pressures and motives from inside the company (Dare, 2016). A multilevel theoretical framework has been developed by Aguilera et al. (2007) to explain the internal motivations for CSR. Three motivations are proposed:
-‐ Instrumental motive -‐ Relational motive -‐ Moral motive
The instrumental motive is followed to gain favourable business outcomes, including wealth maximization. The instrumental motive focuses on CSR initiatives with short-term outcomes. The relational motive is used to “build, maintain, and restore legitimacy” (Dare, 2016). This can be compensation for bad publicity or increasing company image. The moral motive goes with and beyond business importance. The moral motive is due to a believe that helping the humanity is the higher purpose for corporations without economic gain. The moral motive sees virtue as a reward (Dare, 2016).
This shows that company motives towards CSR are not always formed from the desire to make profit or that CSR initiatives are incorporated by a company from an ethical point of view. Pressure from different stakeholders can also be the reasoning. These pressures can vary from shareholders, to governments, to environmental organizations, the media and NGOs.
2.1.3 Effects of CSR
Though there are several underlying effects of CSR, that are less obvious, which have to be discussed.
Firstly, CSR initiatives can indicate that a company is doing well, investing in purposes, which are not necessary for company survival or profit generation but for the mere purpose of doing the ‘right’ thing, and with a moral motive (Dare, 2016), showing that a company has spare money it can invest into CSR initiatives (Adam and Shavit, 2008). Investing this way gives a positive message to potential investors and consumers, even in times when a
corporation’s profits are declining. Even though moral motives have no intent whatsoever of maximizing wealth, this could still be the outcome since most of the CSR initiatives derived from moral motives are long-term. A longitudinal study by Lamberti and Lettieri (2009) was conducted, looking at gaining stakeholder trust by investing in CSR, in which trust is defined by Giddens (1990) as having faith in behaviours and practices of a company with limited knowledge about the company. The study by Lamberti and Lettieri (2009) shows that long-term CSR investments can increase the trust from stakeholders (Lamberti and Lettieri, 2009). Trust and image enhancements can ultimately lead to an increased company value (Bihari and Pradhan, 2011).
CSR is often used as a corporation image enhancer (Lii & Lee, 2012). By participating in the right initiatives at the right time a company can increase its public image, however this is closely related to being proactive instead of reactive. If companies take a reactive stance on CSR than this is often to overcome setbacks. Minor and Morgan (2011) looked at how
effective CSR is to overcome major setbacks. They concluded that companies with more CSR initiatives and higher CSR ratings are more successful in overcoming setbacks than
companies who participate less in CSR. Consequently, CSR can be very useful for large corporations in times of economical crisis or as a counter measure for negative publicity (Vanhamme & Grobben, 2009). This however begs the question whether it could be avoided by enhancing image and reputation in an earlier stage, as in being more proactive. It is
important to state that reactive CSR relational motives are initiated to reach the previous trust and image level, whereas proactive CSR relational motives can give a company an edge in trust and image (Dare, 2016).
company uses CSR or other social initiatives to hide certain aspects or to give the consumer a positive company image, which is not representative. It is very difficult to verify whether a company has the image it deserves and whether all the initiatives presented to the public are with ‘right’ intentions (Laufer, 2003). This gives CSR a very strategic function. A company can give a positive signal to potential investors when the company spends a lot on CSR whilst this may not be expedient. As described previously, there are several motives for CSR
initiatives (Dare, 2016). However, ‘green washing’ is hard to categorize in any of these motives. The closest motive is the relational motive of enhancing company image. Though the framework expects that the enhanced image is representative for the CSR initiatives that the company invests in (Dare, 2016). However, ‘green washing’ is not representative and is a tool to mislead the consumer and hide operational misconduct.
It can be concluded that CSR is a relevant factor and can often lead to high benefits. Long-term CSR investments can increase the trust from stakeholders and contribute to image
enhancements, ultimately leading to an increased company value. On the other hand, CSR can be misleading due to ‘green washing’. The concept of ‘green washing’ will not be taken into account in this research because of the difficulty to assess and will be listed as a limitation.
2.2 CSR in relation to CEO’s
2.2.1 The Upper Echelon TheoryThe upper echelon theory is a theory that links the personality of an executive to the company strategy. It is argued that the personality of the executive is reflected in the companies’ strategic plan. Therefore, it could be argued that the strategic CSR implementation of a company is also linked to the personality or characteristics of an executive or CEO. The definition of the upper echelon theory is as follows:
according to the theory, organizations become reflections of their top executives.” (http://www.palgraveconnect.com/esm/doifinder/10.1057/9781137294678.0713).
This theory is very relevant for organisations as stated by Hambrick and Mason (1984). If the experiences, values and personality of a potential CEO candidate are properly aligned with a companies’ values, beliefs and culture, it can lead to firm specific competitive advantages. Firm specific competitive advantages are needed for a firm to be successful over its competitors. According to the resource-based view, especially knowledge sources are essential to gain these advantages (Wang et al., 2009). These knowledge sources can be personal connections, product specific knowledge, market knowledge, next to the ability to communicate, to plan, how well somebody deals with stress and the ability to motivate
people. Since a CEO is the overall executive of an organization it is key for a company that he is well equipped with relevant skills and knowledge.
In addition, the knowledge management of a firm has to be aligned with the strategic
management of a firm (Casselman & Samson, 2007). Furthermore, there must be alignment between a CEO’s knowledge and experience and the competitive strategy of the company to have a truly positive effect on the results of company (Bagnoli & Giachetti, 2014; Shih & Chiang, 2005). For example, when a firm culture is based upon being very innovative and having a large CSR expenditure with a green image, selecting a CEO who is specialized in doing business the ‘traditional’ way can harm the company and its reputation. So for selecting the proper CEO the upper echelon theory is of relevance and the election of a CEO can make or break the success of a company. It is a challenge for a company to align its knowledge management to its corporate strategy.
2.2.2 Previous research – internal characteristics
Another paper by Petrenko et al. (2015) draws upon the concept of narcissism as a driver for CSR initiatives. They argue that CEO’s can engage in CSR initiatives because of personal needs for attention and enhancement or reinforcement of personal image.
There was one paper by Van der Duijn-Schouten et al. (2014), which examined the effects of religion on CSR.
All these three papers try to find if there are internal CEO characteristics or personal traits (respectively integrity, narcissism and religion) which affect the prioritization of CSR
initiatives within corporations. Obviously this can be very useful for corporations even though it is difficult to analyse internal characteristics and personal traits.
This research paper will touch upon the external characteristics of CEO’s in relation to CSR initiatives.
3. Conceptual model & hypothesis
A widely researched field is the field of gender differences and board diversity within
corporations and the effect it has. Traditionally only men have been the centre of the business world, but this is rapidly changing. Women are getting more and more prominent board positions and due to personality characteristics women are sometimes chosen above men for certain executive positions. Though full equality in the business world between men and women has not been reached yet. A paper by Bear et al. (2010) looks at the effect of gender and board composition on CSR and firm reputation. They found out that women present in a company board not only enhances critical board processes, but also improves the rating for CSR, which consequently results in a better corporate reputation with the advantages accompanying a better reputation.
Given that having women in a company board has positive effects on CSR activities and perceived CSR activities, it will be interesting to see whether having a woman as a CEO versus having a man as CEO also results in a significant difference in CSR initiatives.
influencing the decision process and numerous other company forces. Therefore, it is the question whether in Global Fortune 500 companies gender difference at CEO level has a significant impact on CSR spending.
A survey has been conducted by Ruegger and King (1992) among students, to see whether age and gender had an effect on the perceptions of ethical conduct. They found significant differences in perceptions of ethical conduct in both age and gender differences. Ruegger and King (1992) concluded that women are more ethical than males and that older students had better ethical perceptions than young students. This outcome indicates that age might influence CSR initiatives. Apart from this study no further research was found.
Taking the outcomes from previous research into account the following hypotheses are developed:
H1: Female CEO’s spent more on CSR than male CEO’s in Global Fortune 500 companies
H2: Older CEO’s spent more on CSR than younger CEO’s
The conceptual model below shows the dependent and independent variables used in this research. The independent variables are CEO gender and CEO age, effecting the dependent variable, CSR spending.
Global Fortune 500 Companies
CSR
spending
CEO
The next section will discuss the sample, the variables, the statistical model and the method assumptions.
4. Research methods
4.1 Sample
The main objective in this paper is to see whether CEO age and gender affects CSR spending within Global Fortune 500 companies. To test whether there is a statistically significant correlation between these variables a new database was created. The database contains the annual CSR spending between 2011 and 2013 of in total 130 Global Fortune 500 companies. The average annual CSR spent in US$ will be used. Since the order of magnitude of the CSR spending is related to the size of a company, indicated by the companies’ revenue, the
percentage of average annual CSR spending in comparison to companies’ revenue will be applied.
Due to missing information of one CEO over the period 2011 and 2013, one company was excluded from the database of 130 companies, leaving 129 companies. 129 companies from the Fortune 500 might seem a minority and therefore not representative, though this is not really the case. The 129 companies researched in this paper have a combined annual CSR spent of 17,9 billion US$. The entire Global Fortune 500 has a combined annual CSR spent of approximately 20 billion US$ (2011 – 2013 data). The CSR and revenue data was collected from the Unesco and Global Fortune website. For details the respective companies’ financial statements were used.
average profit gives no additional insights. Furthermore, annual profit data shows a relatively large variance over the years 2011 to 2013 compared to the variance in revenue.
Fortune 500
The chosen target group for analysis is the Global Fortune 500, a target group for which a lot of information is available in the public domain. Since almost all of the Global Fortune 500 companies are listed at the stock exchange, the companies have to publicly report a high level of detailed information to potential investors, etc. and thus it was relatively easy to find the required information on the web. Furthermore, since the (financial) reporting is subject to defined standards and reviewed by independent third parties, the information available is reliable.
Global Fortune 500 companies are active in many different industries and markets. This in combination with the total size of the target group in terms of revenue, employees, etc. and the public role these companies have, makes it an interesting target group.
4.2 Dependent variable
The dependent variable is the amount of average annual CSR spent by the respective Fortune 500 company versus the respective annual revenue (in percentage). When using the
percentage average annual CSR spent versus annual revenue, the database will consist of 129 companies. The dependent variable will be assessed on a numeric basis.
4.3 Independent variables
CEO age will not be split into groups and will be assessed on a numeric basis, like the dependent variable.
4.4 Control variables
Control variables are necessary for testing the validity of the research. To clarify or assess whether there is any relationship between the dependent and independent variables, control variables have to be included. The control variables will remain constant. To exclude that the dependent variable, the average annual CSR spent vs. revenue, is influenced by other factors, several control variables were identified based upon the outcome of the literature review, addressed in this and previous paragraphs:
(1) Whether a company participates in either business to consumer transactions or business to business transactions,
(2) Market segment of operations,
(3) The length of a CEO’s tenure at his or her respective company and (4) The country of origin (head office location) of the company.
This paper will make a distinction between the companies operating in the business to consumer (B2C) sector and the business to business (B2B) sector. A prominent reason to invest in CSR is due to consumer pressure as stated by Groza et al., (2011) and Kim & Lee (2012). This consumer pressure is more relevant for B2C than for B2B. Consequently, it can be argued that there is a fundamental higher percentage of expenditures in CSR at companies doing business with consumers which should be taken into account in this research.
Therefore, the companies in the dataset will be divided amongst two groups, namely B2C and B2B. There are several companies which operate in both B2C and B2B segments. Since it is reasonable to assume that these will still feel the consumer pressure from the B2C market segment, these companies are put in the B2C group. If the company operates in the B2C group, the dummy attached to B2C will equal 1. If a company does not operate in the B2C market, the dummy will equal 0.
initiatives. The Standard Industrial Classification (SIC-code) will be applied resulting in eight different industries:
- Mining and construction (SIC 1000-1499) -‐ Manufacturing (SIC 2000-3999)
-‐ Transportation and communication (SIC 4000-4999) -‐ Wholesale and retail (SIC 5000-5999)
-‐ Finance, insurance & real estate (SIC 6000-6799) -‐ Service industry (SIC 7000-8999)
-‐ Public administration (SIC 9100-9999)
If the company operates in manufacturing, the dummy attached to manufacturing will equal 1. If the company does not operate in manufacturing, the dummy will equal 0. All other industry variables as described above are constructed the same way.
The third dummy is the length of the CEO’s tenure at its respective company. Chung and Liu (2013) state that many newly appointed CEO’s put great emphasis on increasing the share price in a short amount of time.
On the other hand, it is argued that CEO’s with a long tenure at their respective company will have had a longer time to successfully implement their ideas and strategy. Concluding, it is important to take into account the length of the tenure of the CEO at its respective company. The length of the tenure will not be categorized in groups but will be used as a numeric dummy variable.
The fourth and final dummy variable added is the country/region of origin for the respective Global fortune 500 company. The literature review shows that the context in which a
company operates - cultural, sector specific, economical and political (see also the paragraph motives for CSR and more specific pressure from government) – affects CSR commitment. Therefore, location of the head office is one of the dummy variables, since the location determines to a large extend the cultural and political context. Instead of applying all the different countries that are present in the database, clusters are made of these countries with a comparable cultural and political context. By doing so, the number of clusters is limited to 9 (still a high number; instead of 23) with the aim to avoid dividing the dataset too much. The split in clusters/regions is as follows:
- USA/Canada
- Far East Asia (Singapore, South Korea, Taiwan and Japan) - Australia
- China
- South America (Brazil, Colombia) - Russia
- India
- Indonesia (remark: concerns only one company active in the oil industry (petroleum, refining), being Pertamina).
If the company is home-based in the USA/Canada area, the dummy attached to USA/Canada will equal 1. If the company is not home-based in USA/Canada, the dummy will equal 0. All other industry variables as described above are constructed the same way.
All the information needed for the dummy variables was found searching the Internet, using multiple sources. Below a table summarizing all the variables:
Type of variable
Name of variable Description
Dependent CSR Numeric scale of the percentage of average annual CSR spent versus revenue of the respective Fortune 500 company.
Independent CEO Age Dummy variable with a numeric scale of the CEO’s age
Independent CEO Gender Dummy variable which equals 1 for a male CEO and 0 for a female CEO
Control B2C/B2B Dummy variable which equals 1 if the respective company operates in the B2C market and will equal 0 if the company operates in the B2B market only Control Market segment Dummy variable indicating in which market segment
2000-3999), transportation and communication (SIC 4000-4999), wholesale and retail (SIC 5000-5999), finance, insurance and real estate (SIC 6000-6799), service industry (SIC 7000-8999) and public administration (SIC 9100-9999). If the company operates in manufacturing, the dummy attached to manufacturing will equal 1. If the company does not operate in manufacturing, the dummy will equal 0. All other industry variables as described above are constructed the same way.
Control CEO Tenure Dummy variable with a numeric scale of the CEO’s tenure at its respective company
Control Country Dummy variable indicating what the home-based cluster/region is for the Fortune 500 companies. Clusters are made of these countries with a comparable cultural and political context. The following clusters are identified: Western-European countries, USA/Canada, Far East Asia, Australia, China, South America, Russia, India and Indonesia. If the company is home-based in the USA/Canada area, the dummy attached to USA/Canada will equal 1. If the company is not home-based in USA/Canada, the dummy will equal 0. All other industry variables as described above are constructed the same way.
4.5 Statistical model
The hypotheses are tested using the following model, containing one dependent variable, Y, two independent variables, X1 andX2 and Xci utilized control variables:
Y indicates the annual average CSR spent vs. revenue, B0 is a constant,
X1 measures the CEO’s age and X2 measures the CEO’s gender.
Xci stands for all the control variables, which are,
(a) A dummy taking the value of 0 for B2B companies and 1 for B2C companies (Xc1),
(b) Dummies for the market segment the company operates in (Xc2), (c) The length of the CEO’s tenure at its respective company (Xc3) and (d) Dummies for the country of origin of the company (Xc4).
Finally, E is the error term.
The dependent variable is measured on a scale level, and the independent variables as well as the control variables are measured on a nominal or scale level. This makes linear regression an applicable method to test the two hypotheses. The regression is conducted in two steps. First, a regression is carried out by a run including only the dependent variable and the control variables. The second step is a regression including also the independent variables.
4.6 Method assumptions
There are several critical assumptions, which have to be satisfied before ordinary least squares (OLS) is allowed. These assumptions have to be met for providing the best linear unbiased estimates (BLUE). The assumptions, which are tested, are multicollinearity, normality, endogeneity and homoscedasticity. If one of the assumptions is violated, this can create problems of reliability in the results.
4.6.1 Homoscedasticity
will also be a bias in the statistics of the test, resulting in incorrect hypotheses test results, p-values and confidence intervals. Heteroscedasticity gives more weight to observations, which potentially have larger error terms. This means that the observations in the analysis furthest removed from the actual, true regression line, provide the least information about the actual, true regression line. In case there is heteroscedasticity detected, there are several
methodological ways to deal with heteroscedasticity, such as making a correction for the weight of large error variances.
Normally, the heteroscedasticity is assessed using a scatterplot. Though by assessing the scatterplot you cannot be certain. To be sure the Breusch-Pagan test has to be conducted. The outcome of this test can be seen in appendix 1 and confirms that heteroscedasticity is not present in this model. Given is a significance level of 0,1210, which is above 0,05, the confidence interval. This means that the model is not affected by the assumption of heteroscedasticity.
4.6.2 Endogeneity
4.6.3 Multicollinearity
The next assumption, which has to be met for OLS is multicollinearity. Multicollinearity relies on that the independent variables amongst each other are not perfectly correlated; the values are not allowed to be exact linear functions of other explanatory variables in the model (Hill et al., 2009). In case the assumption of multicollinearity is violated, it means that
variables are collinear. Having collinear variables makes it very difficult to isolate any relationship between variables in the model. The OLS outcomes will be unbiased and show the best linear unbiased estimates, though the outcomes might be of low quality with unrepresentative information about the parameters that are unknown due to large standard errors. It might be difficult to predict true parameters; this issue can become even more problematic in case there is little variation amongst the explanatory variables.
To test for multicollinearity, the variance inflation factor (VIF) has to be calculated. The VIF is an index that estimates the inflation of the variance of an estimated regression coefficient due to collinearity. The outcome of the VIF shows that the assumption of multicollinearity is met. The tolerance needs to be above 0,1 and the VIF has to be below 10 to meet the
assumption. As can be seen in Appendix 2, the VIF varies between 1 and 6. It can therefore be concluded that the assumption of multicollinearity is not an issue in this model.
4.6.4 Normality
There is a theorem from Gauss and Markov, which states that OLS might still give BLUE outcomes and estimates even when there are errors that are not normal, independent or identically distributed (Hill et al., 2009). This makes normality the least essential of all the assumptions and violation effects are not as severe as other assumptions.
If there is non-normality, as in this case, than it can be resolved through transformation of the dependent variable. The robustness checks have to give clarity.
It is concluded that the important OLS assumptions are met, with the remarks made regarding normality, and therefore it is appropriate to use OLS to estimate in this model.
5. Empirical results
5.1 Descriptive statistics
The database considered consists of datasets from 129 Global Fortune 500 companies of which only 7 are represented by a female CEO and the remaining 122 have a male as CEO. When averaging the percentage of CSR spending vs. revenue of the female CEO’s and comparing it with the average percentage of CSR spending vs. revenue of the male CEO’s, it is (as expected based upon the literature review) that female CEO’s spend more on CSR. The average percentage that women spend is 0,169 % compared to an overall average of 0,155 % (and a male average of 0,154 %). Though the outcomes of the linear regression analysis will have to show whether this is significant.
For the B2B vs B2C dummy variable there is also an interesting phenomenon. There are in total 28 B2B companies, which give a significant higher amount to CSR initiatives than the B2C companies. The B2B companies average on 0, 311% (in comparison to 0,124 % for B2C). This is unexpected given the outcome of the literature study. It suggests that what might be applicable in terms of CSR spend for B2B vs. B2C in general can not be applied for the target group Global Fortune 500 companies.
Considering the companies’ region of origin/the clusters, the following table is derived from the dataset.
The table shows that on average the USA/Canada region and Australia spend above average on CSR, whereas the others, in particular India, Russia and China are performing far below average on CSR spending.
When looking more closely at the company CEO’s it is notable that within Russia the CEO’s are also active in the countries political arena and therefore it can be argued that they are a lot more powerful and resistant to outside pressures than CEO’s from Australia and
USA/Canada. Many of the CEO’s in Russia are given that position by Russia’s president, Putin.
As well as in Russia, also in China governmental influence on doing business is arguably present, which might explain the lower CSR spending in China.
It should in addition be realised that cultural differences and organizational culture can also be the explanation of the difference in average percentage CSR spend vs. profit.
country(/region revenues(2012 avg.(CSR(spend(011913 #(companies avg.(CSR(spend(2011913/revenues(2012
When assessing the CSR expenditures on a market level/industry level, for the trend analyses a different classification is used than the SIC Codes. The SIC Codes are very useful for making a distinction in the empirical study. For the trends, the classification applied by Fortune is followed.
The table shows that 2 industries outperform. These are computer (software) and pharmaceuticals.
For the computer and pharmaceutical industry an obvious explanation can be given. Many of the CSR activities that are conducted in these market segments is the supply of NGOs, schools or small start-ups with respectively free medicine or free software. The out of pocket costs of the products supplied are relatively low for the companies operating in the computer and pharmaceutical industry, but the commercial value is very high. And as the CSR spent is based upon the market/commercial value of supplied goods (next to services, cash donations, etc.), the CSR spending is high for these companies whilst the actual out of pocket cost is substantially lower. In other words, the company spent is in fact a lower than is represented in the yearly statements.
5.2 Regression results
Appendix 4 shows the regression results in terms of correlations and overall model fit. The first model analysed only includes the dependent variable and the control variables. The second model includes the two independent variables as well. The first model explains 27,1% according to the model summary test. This means that 27,1% of the variability of the CSR spent can be explained by the control variables. The second model is expected to explain more, since two explanatory variables are added, though as can be seen in appendix 5, the increase is very small. Only 0,3% more is explained by adding both variables, CEO age and CEO gender. Though, as can be seen from the ANOVA test, the model is a good overall fit with a significance level below the confidence interval of 0,5.
Because the model has issues of non-normality (see 4.6.4), bootstrapping is performed. The outcomes of the bootstrap analysis can be found in appendix 5. The control variable for experience and the dummy for the service industry still have a statistically significant effect on CSR after bootstrapping is performed. There are no other significant changes in the regression results.
6. Conclusion
6.1 Conclusion
The total average annual CSR spend for the years 2011 to 2013 of the 129 Global Fortune 500 companies considered in the paper, amounts to 17,9 billion US$ per annum. The total CSR spend for all Global Fortune 500 companies amounts to 20 billion US$ in 2013.
(1) Whether a company participates in either business to consumer transactions or business to business transactions,
(2) The business segment in which companies operate,
(3) The number of years a CEO is already working on his or her respective position and (4) The country of origin (the headquarters location) of the company.
The trends derived from the target group, being Global Fortune 500 Companies, endorse this. Though the statistical analyses only shows a statistically significant correlation between average annual CSR spend vs. revenue with the control variable that reflects CEO’s tenure (the number of years a CEO is already working on his or her respective position). This correlation is positive meaning that the more years in function as CEO the higher the CSR expenditure. The second significant relationship is between the service industry and the CSR spend, from this can be derived that, if there are more firms operating in the service industry in this model, than the CSR spend would go up.
The trend is that female CEO’s spend more on CSR than male CEO’s. But the correlation is not statistically significant. Therefore, one of the main hypotheses of this paper – Female CEO’s will spend more on CSR initiatives than male CEO’s in Global Fortune 500 companies – is rejected.
The trends also indicate that the cultural and political context in which a company operates, affects CSR commitment. The Global Fortune 500 Companies with their main offices in the clusters USA/Canada and Australia spent more money on CSR companies with their main office in other regions/clusters.
When assessing the CSR expenditures on a market level, 2 industries outperform. The 2 industries are computer (software) and pharmaceuticals. The CSR spend for the companies in computer (software) and pharmaceuticals can be argued given that the CSR spent is based upon the market/commercial value of supplied good, whilst the actual out of pocket cost is substantially lower since the production costs for the computer software (reproduction) and medicine are relatively low.
The main hypothesis that the age of the CEO affects CSR spending - that older CEO’s spent more on CSR than younger CEO’s - is rejected.
6.2 Limitations
The database consisted of only 129 out of 500 Global Fortune 500 companies. Furthermore, CSR spending data for only 3 years in a row was available. It can be argued that CSR data for only 3 years in a row is not representative, given that CSR initiatives are primarily long-term investments in alignment with a companies’ strategy.
There is no real universal definition of CSR. Consequently, countries, companies and business sectors may have a variety of approaches towards CSR that are not immediately obvious, and the absolute amount of CSR spending (in money terms) is not always indicative of a company’s true level of corporate responsibility. Just as it is common for a government to offer tax breaks to multinational companies, governments may also ask for social investments in return.
Some cultures have a tradition in philanthropy, but one that is discreet and private rather than publicly measured through formal CSR reporting systems. By using only reported CSR spending this element is not considered.
As stated in the literature study, the upper echelon theory can be applied to look for specific communalities between strategic CSR implementation and CEO characteristics. It is argued that the CEO’s characteristics will be reflected in the corporate strategy and therefore the CSR initiatives. Especially when it concerns Global Fortune 500 companies.
If reasoned the other way around, it might also be that companies align the CEO
It could also be that at some companies CSR spending is either high or low, irrelevant the CEO, due to a pre-established corporate strategy.
Furthermore, ‘greenwashing’ is not explicitly considered. This paper relies upon the
proposition that there are no underlying reasons for investing in CSR, but that CSR is a result of initiative in good faith/of doing ‘the right things’.
A final limitation is that CSR spending might not be representative because CSR initiatives might cost little to some companies (out of pocket money), whilst having a very high commercial value. This can be seen especially in the computer and pharmaceutical market. The reproduction of a chip or medicine injection might cost the producing company little but can have a very large commercial value and therefore there can be a very high reported CSR spending.
There are many limitations to this paper due to limited time and resources; no internal characteristics were taken into account whilst there is a possibility that they affect the outcomes of the study.
It should be considered that the data sample was taken from only Global fortune 500 companies, the target group, and therefore the results are not generalizable for small and medium enterprises (SMEs).
6.3 Further research
What we should keep in mind, the total CSR spend of the 129 Global Fortune 500 companies evaluated in this paper amounts to almost 18 billion US$ in 2013. An amount of money that can make a huge difference to the lives of millions.
7. Sources:
Adam, A.M., Shavit, T. (2008). How Can a Ratings-based Method for Assessing Corporate Social Responsibility (CSR) Provide an Incentive to Firms Excluded from Socially
Responsible Investment Indices to Invest in CSR? Journal of Business Ethics, 82(4), 899-905.
Aguilera, R.V., Rupp, D.E., Williams, C.A., Ganapathi, J. (2007). Putting the S back in corporate social responsibility: A multilevel theory of social change in organizations. Academy of Management Review. Vol 32, 836-863.
Bablak, K., Trendafilova, S. (2010). CSR and environmental responsibility: motives and pressures to adopt green management practices. Corporate Social Responsibility and Environmental Management. Vol 18(1), 11-24.
Bagnoli, C., Giachetti, C. (2014). Aligning knowledge strategy and competitive strategy in small firms. Journal of Business Economics and Management. Vol 16(3), 571-598.
Bandsuch, M., Pate, L., Thies, J. (2008). Rebilding Stakeholder Trust in Business: An
Examination of Principle-Centered Leadership and Organizational Transparency in Corporate Governance. Business and Society Review. Vol 113(1), 99-127.
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. Vol 97(2), 207-221.
Bihari, S.C., Pradhan, S. (2011). CSR and Performance: The Story of Banks in India. Journal of Transnational Management, 16(1), 20-35.
Carroll, A.B. (1979). A three-dimensional model of corporate social performance. Academy of Management Review, 4, 497-505.
Chung, C.N., Luo, X.R. (2013). Leadership succession and firm performance in an emerging economy: Successor origin, relational embeddedness, and legitimacy. Strategic Management Journal. Vol 34(3), 338-357.
Dare, J. (2016). Will the Truth Set Us Free? An Exploration of CSR Motive and Commitment. Business and Society Review. Vol 121(1), 85-122.
Fombrun, C.J. (2005). Building Corporate Reputation Through CSR Initiatives: Evolving Standards. Corporate Reputation Review. Vol 8(1), 7-11.
Gallie, W.B. (1956). Essentially Contested Concepts. Proceedings of the Aristotelian Society. Vol 56, 167-198.
Giddens, A. (1990). The Consequences of Modernity. Stanford University Press, Stanford.
Groza, M.D., Pronschinske, M.R., Walker, M. (2011). Perceived Organizational Motives and Consumer Responses to Proactive and Reactive CSR. Journal of Business Ethics. Vol 102, 639-652.
Hill, R.C., Griffiths, W.E., Lim, G.C. (2009). Principles of Econometrics 4th Edition. John Wiley & Sons Ltd.
Jamali, D., Mirshak, R. (2007). Corporate Social Responsibility (CSR): Theory and Practice in a Developing Country Context. Journal of Business Ethics. Vol 72(3), 243-262.
Kakabadse, N.K., Kakabadse, A.P., Lee-Davis, L. (2009). CSR leaders road-map. Corporate Governance: The international journal of business in society. Vol 9(1), 50-57.
Kim, S., Lee, Y.J. (2012). The complex attribution process of CSR motives. Public Relations Review. Vol 38(1), 168-170.
Laufer, W.S. (2003). Social Accountability and Corporate Greenwashing. Journal of Business Ethics, 43(3), 253-261.
Lii, Y.S., Lee, M. (2012). Doing Right Leads to Doing Well: When the Type of CSR and Reputation Interact to Affect Consumer Evaluations of the Firm. Journal of Business Ethics. Vol 105, 69-81.
Maak, T., Pless, N.M., Voegtlin, C. (2016) Business Statesman or Shareholder Advocate? CEO Responsible Leadership Styles and the Micro-Foundations of Political CSR. Journal of Management Studies. Vol 53(3), 463-493.
Menichini, T., Rosati, F. (2014). The Strategic Impact of CSR Consumer-company Alignment. Social and Behavioral Sciences. Vol 109(1), 260-364.
Minor, D., Morgan, J. (2011). CSR as Reputation Insurance: PRIMUM NON NOCERE. California Management Review. Vol 53(3), 40-59.
Ruegger, D., King, E.W. (1992). A study of the effect of age and gender upon student business ethics. Journal of Business Ethics. Vol 11(3), 179-186.
Okoye, A. (2009). Theorising Corporate Social Responsibility as an Essentially Contested Concept: Is a Definition Necessary? Journal of Business Ethics. Vol 89, 613-627.
O’Riordan, L., Fairbrass, J. (2013). Managing CSR Stakeholder Engagement: A New Conceptual Framework. Journal of Business Ethics. Vol 125(1), 121-145.
Petrenko, O.V., Aime, F., Ridge, J., Hill, A. (2015) Corporate social responsibility or CEO narcissism? CSR motivations and organisational performance. Strategic Management Journal. Vol 37(2), 262-279.
Sheehy, B. (2015). Defining CSR: Problems and Solutions. Journal of Business Ethics, 131(3), 625-648.
Shih, H.A., Chiang, Y.H. (2005). Strategy alignment between HRM, KM, and corporate development. International Journal of Manpower. Vol 26(6), 582-603.
Van der Duijn-Schouten, C.M., Graafland, J., Kaptein, M. (2014). Religiosity, CSR Attitudes, and CSR Behavior: An Empirical Study of Executives’ Religiosity and CSR. Journal of Business Ethics. Vol 123(3), 437-459.
Vanhamme, J., Grobben, B. (2009). “Too Good to be True!”. The Effectiveness of CSR History in Countering Negative Publicity. Journal of Business Ethics. Vol 85, 273-283.
Verissimo, J.M.C., Lacerda, T.M.C. (2015). Does integrity matter for CSR practice in organizations? The mediating role of transformational leadership. Business Ethics: A European Review. Vol 24(1), 34-51.
Wang, H.C., He, J., Mahoney, J.T. (2009). Firm-specific knowledge resources and
competitive advantage: The roles of economic- and relationship-based employee governance mechanisms. Strategic Management Journal. Vol 30, 1265-1285.
Wang, S., Gao, Y, Hodgkinson, G.P., Rousseau, D.M., Flood, P.C. (2015). Opening the Black Box of CSR Decision Making: A Policy-Capturing Study of Charitable Donation Decisions in China. Journal of Business Ethics. Vol 128, 665-683.
Yuan, W., Bao, Y, Verbeke, A. (2011) Integrating CSR Initiatives in Business: An Organizing Framework. Journal of Business Ethics. Vol 101, 75-92.
8. Appendix
Appendix 1: heteroscedasticity
Appendix 3: normality
Tests of Normality
Kolmogorov-Smirnova Shapiro-Wilk Statistic df Sig. Statistic df Sig.
CSR ,364 129 ,000 ,294 129 ,000
Appendix 4: Regression results
Model 1
ANOVAa
Model
Sum of
Squares df Mean Square F Sig.
1 Regression ,002 16 ,000 2,606 ,002b
Residual ,005 112 ,000
Total ,007 128
a. Dependent Variable: CSR
b. Predictors: (Constant), Indonesia, Dummy9199, India, Russia, Dummy7089, SouthAmerica, China, Australia, Dummy5059, Dummy4049, Experience, FEA, B2CDummy, Dummy6067, USACanada, Dummy2039
Model Summaryb Model R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson 1 ,521a ,271 ,167 ,00685 2,070
a. Predictors: (Constant), Indonesia, Dummy9199, India, Russia, Dummy7089, SouthAmerica, China, Australia, Dummy5059,
Dummy4049, Experience, FEA, B2CDummy, Dummy6067, USACanada, Dummy2039
Model 2
ANOVAa
Model
Sum of
Squares df Mean Square F Sig.
1 Regression ,002 18 ,000 2,308 ,004b
Residual ,005 110 ,000
Total ,007 128
a. Dependent Variable: CSR
b. Predictors: (Constant), Indonesia, Dummy9199, India, Russia, Dummy7089, SouthAmerica, China, Australia, Dummy5059, Dummy4049, Age, B2CDummy, MaleDummy, FEA, Dummy6067, USACanada, Experience, Dummy2039
Model Summaryb Model R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson 1 ,524a ,274 ,155 ,00689 2,056
a. Predictors: (Constant), Indonesia, Dummy9199, India, Russia, Dummy7089, SouthAmerica, China, Australia, Dummy5059, Dummy4049, Age, B2CDummy, MaleDummy, FEA, Dummy6067, USACanada, Experience, Dummy2039
