Corporate social responsibility, corporate political activity and corporate financial performance : an empirical study on the environmental dimension of CSR

(1)

1

MSc. Thesis:

Corporate Social Responsibility, Corporate Political Activity

and Corporate Financial Performance: An empirical study on

the environmental dimension of CSR

Student: Kevin Perquin

Student Number: 10358080

MSc Business Administration, Strategy Track, University of Amsterdam

Supervisor: Dr. Pushpika Vishwanathan

Date: 22-06-2018

Word count: 14,440

(2)

2

Statement of Originality

This document is written by Student Kevin Perquin who declares to take full

responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and

that no sources other than those mentioned in the text and its references have

been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision

of completion of the work, not for the contents.

(3)

3

1. Abstract

Scholars have found that Corporate Social Responsibility and Corporate Political Activity can both enhance the firm’s financial performance. For a long period of time, these fields of research have been studied independently. Recently, scholars have started to conceptualize that from the resource based view an integration of CPA and CSR strategies can enhance their effectiveness. In this study the effect of CPA on the relationship between CSR and Corporate Financial Performance was tested. Moreover, this relationship was also tested for the

environmental dimension of CSR. Hierarchical regressions on 393 US based Fortune 500 listed firms were run to test these propositions. Findings, suggest that CPA does not affect the relationship between CSR and CFP. Unexpectedly, a negative impact of CPA on the

relationship between Corporate Environmental Performance Strengths and CFP was found.

Key words: Corporate Social Responsibility (CSR), Corporate Political Activity (CPA),

Corporate Environmental Performance (CEP), Corporate Financial Performance (CFP), Resource Based View (RBV).

(5)

5

2. Introduction

During the European refugee crisis millions of immigrants entered the European Union. Society wanted to help them, however this was not an easy task at hand. They had trouble adjusting to their new living environments and especially finding jobs. Starbucks noticed these societal problems and responded upon them by setting up a specific programme to help these immigrants. They provided them with jobs, helped them integrate in the local

communities and supported them with finding appropriate housing. What this example illustrates is that firms can contribute to solving societal problems (Mohn, 2017; CNN, 2018.).

Over time there has been increasing pressure on firms by society to become more socially responsible while conducting business. Firms respond to this by engaging with their stakeholders from both inside and outside the firm, through corporate social responsible actions, as illustrated by Starbucks. They for example reduce their environmental footprints, while producing their products or are improving the working conditions for their employees. These initiatives can improve the overall image and perception of the firm by its stakeholders (Harrison, Bosse, & Phillips, 2010; McWilliams, 2000; Wang, Dou, & Jia, 2016).

Furthermore, it is argued that these CSR activities can enhance the firm’s financial performance through cost reductions and revenue increases. Cost reductions for example refer to lower labour costs as people are more willing to work for a respectable firm, while

revenues can increase through increased willingness to buy from reputable firms by

consumers. These reputational benefits and additional economic outcomes make it interesting for firms to invest into their CSR programmes, as is for example visible in many annual reports of publicly listed firms in which their CSR activities are disclosed. Scholars have over time obtained strong empirical evidence, which states that CSR enhances Corporate Financial

(6)

6 Performance (Aguinis & Glavas, 2012; Barnett & Salomon, 2006; McWilliams, Siegel, & Wright, 2006; Orlitzky et al., 2003; Tsoutsoura, 2004; Wang et al., 2016).

On the other hand, firms can be active in the political landscape with the intention of improving their own market position, which can come at the cost of society (Oberman, 2004). This statement is illustrated by the recent Net Neutrality act, at the ending of 2017 the Federal Communications Commission led by Chairman Ajit Pai repealed the Net Neutrality Rules in the United States of America. In essence, it is the removal of regulation that prohibits

broadband providers from artificially managing internet traffic. For example, they are now allowed to block or even charge additional fees for accessing certain content and are able to adjust internet user speeds based on the content provider or user. The repeal of Net Neutrality has led to public outcry as people perceive it as taking away the freedom of the internet, negatively impacting society as a result. Besides, US citizens many internet based firms have protested against this repeal as it will most likely increase their costs and reduce their users’ experience. Although, the large majority of citizens, firms and politicians were strongly against the removal of Net Neutrality, the FCC decided to go through with it. The general consensus is that through strong political activities, like lobbying and donations, the broadband providers have been able to sway members of the FCC into their favour with as result a policy outcome that favours them over society (Kang, 2018; Opensecrets, 2018).

This example illustrates how policies can change the business environment and how firms can manipulate these changes through being politically active. The political landscape is thus filled with opportunities, but also creates a lot of uncertainty for firms. New

governmental policies can completely reshape industries and business environments (Hillman & Hitt, 1999). Engaging in corporate political activity is a strategic approach by firms in order to deal with this uncertainty. More specifically, corporate political activity is as stated by Hillman & Hitt (1999): “an attempt to use the power of government to advance private ends.

(7)

7 The overall objective of political behaviour is to produce public policy outcomes that are favourable to the firm's continued economic survival and success” (Hillman & Hitt, 1999, pp. 826).

Within the CPA field of research it is argued that these political activities can enhance value creation and improve financial performances. As it can create a business environment that favours certain firms over the competition, potentially even leading to a competitive advantage (Cooper, Gulen, & Ovtchinnikov, 2010; Hillman & Hitt, 1999; Hillman, Keim, & Schuler, 2004; Lux, Crook, & Woehr, 2011; Mathur, Singh, Thompson, & Nejadmalayeri, 2013; Schuler, 1996).

The resource based view of the firm has been used to explain these performance differentials for both CSR and CPA (Barney, 1991). McWilliams & Siegel (2001) argue that CSR is a bundle of resources a firm has access to and that these can become firm specific. In essence, this means that there is a difference among firms with regards to their bundle of CSR resources. For example, some firms have more and better CSR resources than others, as can be reflected in better environmental production methods. As the RBV is an outcome based theory, it is especially relevant for explaining the relationship between CSR and CFP, as CFP is an outcome based metric(McWilliams & Siegel, 2001; Russo & Fouts, 1997).

Similarly, scholars in the field of CPA have used the resource based view to explain differences in CPA outcomes between firms (Dahan, 2005; Hillman et al., 2004). Dahan, (2005) argues that CPA can be seen as a bundle of political resources a firm has access to and can become firm specific. These political resources differ among firms, which can explain the differences in performance regarding CPA. For example, some firms have built strong

relationships with political actors over time, giving them direct access to affect certain policy making processes. As a result of these superior political resources firms can achieve an

(8)

8 improved financial performance (Dahan, 2005; Hillman et al., 2004, 2004; Lux et al., 2011; Mathur et al., 2013).

From the perspective of the RBV it would make sense to integrate CSR and CPA, as they can both be seen as bundles of resources, which can potentially impact each other (Mellahi, Frynas, Sun, & Siegel, 2016). In addition, both theories are classified as nonmarket strategies, which in essence means that firms are trying to improve their performance through strategies outside of the market place (Mellahi et al., 2016). However, for a long period of time scholars have separated these fields of research, as they were often perceived as

conflicting strategies. Society views CSR generally as a good thing firms participate in, while CPA is often seen as an activity that negatively impacts society (Mellahi et al., 2016).

Furthermore, these strategies can lead to conflicting goals within the firm. For example, profiling the firm as an environmental friendly firm, while being politically active for reducing stricter environmental rules (Dahan, Hadani, & Schuler, 2013; Oberman, 2004; Brammer & Pavelin, 2006).

Nowadays, scholars are trying to integrate these two non-market strategies as it is suggested that they can enhance each other’s effectiveness(Hond, Rehbein, Bakker, & Lankveld, 2014; Mellahi et al., 2016; Rasche, 2015; Rehbein & Schuler, 2015). More specifically, it is theorized that an integration of these strategies can enhance the CFP of the firm. This can be realized by using CPA resources for strengthening the effectiveness of CSR programmes (David, Bloom, & Hillman, 2007; Hond, Rehbein, Bakker, & Lankveld, 2014; Mellahi et al., 2016).

An article from the New York Times by Boudette (2017) illustrates this integration, as several car companies in the US were actively lobbying, such as General Motors and Ford, for reverting the emission standard changes introduced by the Obama administration, when the

(9)

9 Trump administration took office. At the same time, companies like Hyundai and Toyota were lobbying for stricter emission rules, because they have the required technologies and wanted to strengthen their market position (Boudette, 2017). Firms like Toyota and Hyundai are trying to use their political resources to build further on the strong results of their CSR programmes and gain a competitive advantage.

On the other hand, it is theorized that CPA resources can be used to reduce the negative effects of a weak CSR programme. An example for this is the case of Nestlé in Canada, they are extracting ground water on long expired permits harming the natural

environment in the process. However, through their political activities they have been able to continue their operations on these expired permits without large repercussions (CBC Canada, 2017).

Both of these examples illustrate that some firms are using their CPA resources to affect the outcome of their CSR activities. This is in line with the argumentation of several scholars that there is an overlap between CSR and political resources, which can be used to strengthen or compensate each other (Hond et al., 2014; Mellahi et al., 2016; Rasche, 2015; Rehbein & Schuler, 2015). However, these studies are based on practical examples and theorizing, while lacking empirical evidence to support these claims. Empirical evidence is necessary to further assess this theorized interaction between CSR and CPA. More

specifically, how this interaction may impact the firm’s financial performance. Therefore this study attempts to assess whether this relationship exists and how strong it may be. As a result the following research question is proposed:

Does Corporate Political Activity affect the relationship between Corporate Social Responsibility and Corporate Financial Performance?

(10)

10 In addition to testing the complete CSR variable, this study will also examine this proposed relationship through the environmental lens of CSR. Within the literature a recurring theme is that the Corporate Environmental Performance is a strong predictor with regards to firm financial performance. Many scholars have provided empirical evidence to support this theory (Dixon-Fowler, Slater, Johnson, Ellstrand, & Romi, 2013; Ambec & Lanoie , 2008). Similarly, in the CPA stream of research it is argued that the effectiveness from political activities is strong with regards to environmental issues (Child & Tsai, 2005; Cho et al., 2006; Peterson & Pfitzer, 2009). Furthermore, by looking at the examples of Nestlé, Toyota and Hyundai it illustrates the proposed relationship between CPA and Corporate Environmental Performance. Therefore, it is interesting to also analyse the relationship between CEP and CPA with regards to CFP.

In this study the research question is examined by conducting an empirical study on 393 publicly listed Fortune 500 firms averaged over the years 2014-2016. CSR was measured by using a KLD dataset, CPA was measured by PAC and Lobbying Expenditures, a new CPA variable was introduced for assessing the impact of environmental issues on the CEP-CFP relationship and CFP was measured through the market metric Tobin’s q. Findings of this study suggest that the environmental performance positively affects the firm’s financial performance. Furthermore, they suggest that Lobbying expenditures negatively impact the positive effect of a strong environmental performance on the firm’s financial performance. This study contributes to the Resource Based View, by addressing the interplay between CSR and CEP with regards to CPA resources. The results challenge the theoretical assumption that firms can strengthen their strong environmental performance through their political resources, as the opposite effect was found. This strengthens the theoretical assumptions that firms have problems integrating their CEP and CPA resources. The practical implication of these results

(11)

11 suggests that firms with strong environmental performances should be careful with the use of their political resources as this may reduce their financial performance.

This study is structured in the following order. First, theoretical arguments have been built to form hypotheses for testing the proposed research question. Second, the methodology of the empirical research has been described. Third, the empirical results are analysed in the results section. Fourth, in the discussion the results are analysed with regards to existing literature and limitations of this study are provided. Finally, a conclusion is made to summarize this study.

3. Theoretical Framework

3.1 Corporate Social Responsibility and Firm Financial Performance

This stream of research has been developing since the 1950s and is becoming an increasingly important business element for firms (Carroll, 1999). McWilliams & Siegel (2001) define CSR as the following: “Here we define CSR as actions that appear to further some social good, beyond the interests of the firm and that which is required by law.” (McWilliams & Siegel, 2001, pp. 117). CSR consists of many different dimensions that combined result in the broader term CSR. Dahlsrud (2008) conducted a meta-analysis on the definitions of CSR and concluded that CSR consists of five dimensions; environmental (the natural environment), social (the relationship between business and society), economic (socio-economic or financial aspects), stakeholder (stakeholders or groups) and voluntariness (actions not prescribed by law) (Dahlsrud, 2008). In each of these dimensions firms can have different levels of

performance, which combined create a total CSR performance. This is for example measured by Kinder, Lydenberg & Domini institution (KLD, 2016).

Nowadays, firms are often open about their CSR activities and even report them in the annual reports to illustrate their engagement in CSR. This trend has developed over time, in

(12)

12 which stakeholders started to put more pressure on firms to take CSR into account. Many stakeholders, like employees and customers were increasingly expecting firms to conduct business in a responsible way (Aguinis & Glavas, 2012; Campbell, 2007). In addition, during the 1990s the instrumental stakeholder theory became prevalent in this field of research (Jones, 1995). This theory suggests that having good relationships with the stakeholders in the firm is essential for the firm’s performance and it was not just destroying shareholder value. It was therefore argued that investments in building these relationships should be beneficial for the firm’s performance (Freeman & Evan, 1990; Jones, 1995).

Furthermore, scholars like McWilliams & Siegel (2001) argue that CSR also improves the societal value, by for example reducing emissions beyond the legal standards or farming without the use of pesticides. Engaging CSR thus not only affects the firm’s performance, but a wide variety of business aspects. This is supported by Aguinis & Glavas (2012)’s meta-analysis regarding 588 journal articles and 102 books of CSR, in which they concluded that CSR for example also leads to an improved reputation, better evaluations of the company and product and customer loyalty

CSR researchers have also shown that besides society, firms can also financially benefit from their CSR initiatives. McWilliams & Siegel, (2001) build on the resource based view for explaining these potential economic benefits from CSR. They argue that CSR can be seen as a bundle of resources within the firm and that these resources can be valuable, rare and non-imitable and thus potentially leading to competitive advantage (Barney, 1991). The main point of this view is that CSR is suggested to be a strategic asset of firms to improve their economic performance (McWilliams & Siegel, 2001; McWilliams et al., 2006). The results of these CSR strategies are suggested to strengthen the firm’s financial performance.

During the recent years many studies have found significant increases in the firm’s financial performance through their CSR programmes (Aguinis & Glavas, 2012; Barnett &

(13)

13 Salomon, 2006; Brammer; McWilliams et al., 2006; Orlitzky et al., 2003; Tsoutsoura, 2004; Wang et al. 2016). Interestingly, Barnett & Salomon (2006) have concluded that CSR positively affects firm’s performance, but within certain boundaries. As they argue that this positive relationship is dependent on expenditure pattern of a firm with regards to CSR. Investing in CSR does not automatically result in higher financial returns in comparison with the competition. Firms need to choose the right investments or it can result in negative financial results (Barnett & Salomon, 2006).

In response to the many emerging studies in the field of CSR and its relation to CFP, Orlitzky, et al. (2003) conducted a meta-analysis of these studies on CSR in relationship to CFP. They included both proponents and opponents of the relationship, and found that the relationship is generally positive between CSR and CFP. The amount of empirical evidence collected strongly exceeded the amount of empirical evidence that was arguing for no relationship or even a negative relationship between CSR and CFP (Orlitzky et al., 2003).

In following of Orlitzky et al. (2003)’s meta-analysis, Wang et al. (2016) have also conducted a meta-analysis based on more recent studies. Their results were similar to those of Orlitzky et al. (2003), finding an exceeding amount of studies that argue in favour of the positive relationship. However, they provide several arguments why some scholars argue against this relationship. Most notably from the perspective of neoclassical economics and the agency theory. Neoclassical economists state that investing in CSR creates a competitive disadvantage, because the firm is using resources for activities other than improving operational efficiencies. From the perspective of the agency theory, it is argued that is an interest pursued by managers in their own interest, thus taking money away from shareholders (Wang et al., 2016). Another interesting finding of Wang, et al. (2016) was the difference in measuring CSR among the studies. These differences in measurement can cause significant differentials with regards to the relationship between CSR and CFP.

(14)

14 Based on the arguments presented above, it is argued that the relationship is generally positive between CSR and CFP. Although, there are some studies which argue against this relationship and should be taken into account. However, these studies are largely exceeded by the amount of studies which provide empirical support for a positive relationship. Therefore this study adapts a similar view that CSR enhances the firm’s CFP. As a result, the following hypothesis is proposed:

H1: Corporate Social Responsibility enhances the firm’s Corporate Financial

Performance.

3.2 Corporate Environmental Performance and Corporate Financial Performance In the previous section the CSR relationship with CFP was explained. However, as stated by (Dahlsrud, 2008) CSR is a bundle of multiple dimensions, one of these dimensions the natural environment. Several scholars have argued that the environmental dimension of CSR has a strong effect on the firm’s financial performance (Dixon-Fowler et al., 2013; Russo & Fouts, 1997; Ambec & Lanoie, 2008).

Similarly to McWilliams & Siegel (2001)’s integration of CSR and the resource based view, Russo & Fouts (1997) have integrated the RBV with regards to the firm’s

environmental and financial performance. The rationale behind this integration is that the RBV is an outcome orientated theory that focuses on firm performance. They propose that three elements of the RBV are directly impacting this relationship; physical assets and technology, human resources and organizational capabilities and intangible resources (Russo & Fouts, 1997). As firms need these resources in order to have the ability to obtain good environmental performances, it can explain performance differences amongst firms with regards to the environmental dimension (Russo & Fouts, 1997). In addition, their empirical

(15)

15 results support these statements as firms with better environmental performances outperform firms with lower environmental performances with regards to CFP.

Ambec & Lanoie (2008) analysed multiple studies within the Corporate Environmental Performance field of CSR and concluded that being environmentally

responsible can improve the firm’s economic returns through two main mechanisms; reducing costs and improving revenues. An interesting perspective they offer is that firms can also focus on selling products, which offer environmental solutions. As is argued this green focus can significantly improve the financial performance of firms (Ambec & Lanoie, 2008).

Furthermore, Dixon-Fowler et al. (2013)’s study argues that this CEP improves the CFP aspect of the firm. They have moved beyond the question if CEP enhances CFP and instead looked at when this relationship is the strongest. Their results indicate that CEP positively affect CFP for both small and large firms, sometimes even more for smaller firms. Another interesting finding was that this relationship seems to be the strongest for US based firms (Dixon-Fowler et al., 2013).

Moreover, Etzion (2007)’s study also takes into account the cost saving aspect of being a green firm. Especially, the effectiveness of the environmental programmes within a firm are an important driver for the positive effect on CFP. Cost saving is for example the most effective if employees are aligned with the values and mission of the firm with regards to the environmental programme. Participation of stakeholders in the organisation is

necessary to achieve the strongest positive effect of CEP on CFP (Etzion, 2007).

Finally, (Orlitzky et al., 2003) conducted a meta-analysis of CSR in relationship with CFP studies. As part of this analysis, they have looked at the environmental aspect of CSR. Especially studies within this field of research provided strong empirical evidence supporting

(16)

16 this theorized relationship. Based on these studies and the general consensus within this field of research, it is theorized that CEP improves CFP.

Based on the arguments presented above, this study argues that Corporate

Environmental Performance enhances the Corporate Financial Performance. Therefore the following hypothesis is proposed:

H2: Corporate Environmental Performance enhances the firm’s Corporate Financial

Performance.

3.3 Corporate Political Activity and Corporate Financial Performance

Corporate political activity is a non-market strategy for firms to reduce uncertainty in their external environment caused by governmental institutions (Hillman et al., 2004). There are many different political activities firms can undertake, the most commonly used are PAC contributions and Lobbying expenses. These are direct financial investments with the goal of affecting legislators and supporting preferred candidates within governmental institutions. The reasoning behind these investments is that it will help the firm operate within its market by reducing governmental interference (Hadani & Schuler, 2013; Mathur et al., 2013).

Furthermore, firms also indirectly invest in CPA to obtain additional information about the legislation making process or be informed in an early stage of its development. Agrawal & Knoeber (2001) show that this process can for example be achieved by having outside members with political experience in the board. Their results indicate that having these political actors on board is especially beneficial for firms acting in industries with a lot of governmental interference (Agrawal & Knoeber, 2001).

In addition, some firms mobilize their employees or interest parties like labour unions to affect or withhold certain pieces of new legislation. This method is especially effective for

(17)

17 changing the public opinion and forcing governmental institutions to react (Baumgartner, Berry, Hojnacki, Leech, & Kimball, 2009).

The commonality between these different methods of participation in political

activities is that they are non-market strategies with the goal of maintaining or improving the firm’s position within the market with regards to governmental issues. These non-market strategies can even turn into a competitive advantage, through stimulating certain policies and or legislators, which favour themselves over the competition (Hillman & Hitt, 1999; Hillman et al., 2004).

Hillman et al., (2004) state that the resource based view can be used as an explanation for these performance differentials between firms with regards to CPA. As CPA can be seen as a bundle of resources a firm has. Dahan (2005) further reviewed this relationship and states that these resources can be seen as political resources a firm has access to, for example strong relationships with certain policy makers. Furthermore, firms can mobilize these political resources for obtaining favourable outcomes regarding the political environment or use them to gain an advantage over the competition through their political activities. As a result, the RBV can explain performance differentials between firms based on their political activities, as these can be firm specific (Barney, 1991; Dahan, 2005; Grant, 1999).

To further explain the performance differentials between firms based on their political activities, scholars have provided empirical evidence that supports a positive relationship between CPA and CFP. Schuler (1996) illustrates how strongly CPA can affect an industry. He found that US Steel firms in the 1970s and 1980s were successfully lobbying against the import of foreign steel, which enabled them to obtain returns that were higher than the global prices for a long period of time, thus obtaining abnormal financial returns in the process (Schuler, 1996).

(18)

18 Moreover, Cooper et al. (2010) provide empirical evidence over a twenty five year period that political contributions by firms significantly improve their future stock returns. This effect was even stronger if the firm was supporting candidates in the state in which it is active.

In addition, Lux et al. (2011) further build on the theories proposed by Hillman, et al. (2004) that CPA enhances the firm’s financial performance with empirical evidence. Their results suggest that CPA does enhance the firm’s financial performance (Lux et al., 2011). Combining all of the previous arguments, an argument can be made that political resources can be used as a value creating non market strategy (Cooper et al., 2010; Dahan, 2005; Hillman & Hitt, 1999; Hillman et al., 2004; Lux et al., 2011; Mathur et al., 2013; Schuler, 1996).

Although CPA seems like a beneficial non-market strategy for firms, it is not feasible for all firms within a sector. Hillman, et al. (2004) suggest that there are four main drivers for a firm’s participation in CPA; Firm level, Industry type, Institutional type and relevance of the Issue. These four main drivers are acknowledged by most of the scholars within this field of research (Hadani & Schuler, 2013; Hillman et al., 2004; Lawton, McGuire, & Rajwani, 2013; Lux et al., 2011). Lux, et.al. (2011) have tested these drivers of CPA and found significant results for Firm level, Institutional type and Industry. Within these drivers, Firm level for example refers to firm size and strategy; Institutional type contains for example local and national governmental institutions and Industry refers to the level of regulation of the firm’s market. Relevance of the issue is significant but not as strong as the other three (Lux et al., 2011). Lenway & Rehbein (1991) have shown the effects of firm size on the firm’s

willingness and ability to engage in CPA. Larger firms are more likely to engage in CPA than smaller firms. Most of the time, the industry leader or leaders will become active in the political environment, while the rest will either follow them or not engage in CPA at all. This

(19)

19 is for example explained by the differences in resources and acknowledgement by the

government (Lenway & Rehbein, 1991).

Furthermore, Hadani & Schuler (2013) argue that firms in regulated industries are more likely to engage in CPA than in non-regulated industries, because of their ability to directly affect the market through CPA. This is also reflected in the results which indicate that CPA enhances financial returns in regulated industries (Hadani & Schuler, 2013).

Although most scholars agree on the drivers of CPA there is no consensus regarding the effect of CPA on the firm’s CFP. Besides the previously described positive relationship, there are two other streams of research; CPA has no effect and CPA has a negative effect on the firm’s CFP. Several studies suggest that there is no relationship between CPA and the firm’s CFP. Their results indicate that firms who actively engage in CPA do not gain abnormal financial returns. Furthermore, no negative relationship was found either, which suggest that firm’s engaged in CPA are not losing money either (Ansolabehere, Snyder Jr, & Ueda, 2004; Hersch, Netter, & Pope, 2008). Ansolabehere, et al. (2004) and Hersch, et al. (2008) argue that firms invest in political issues for the short-term with no long-term vision in mind. This is referred to as reactive behaviour of these firms to certain changes in the political environment. As a result, firms often engage in CPA to deal with these short-term political issues, without building long-term political capital. This can explain the lack of long term financial results by firms engaged in CPA (Ansolabehere et al., 2004; Hersch et al., 2008).

The other stream of research argues that CPA is negatively related to firm’s financial performance (Aggarwal, Meschke, & Wang, 2012; Hadani & Schuler, 2013). These scholars provide results indicating that the firm market value decreases by engaging in CPA.

Moreover, Hadani & Schuler (2013) concluded that CPA also negatively impacts the firm’s financial performance. Aggarwal, et al. (2012) argues that this is a result of agency theory.

(20)

20 They state that managers are investing in CPA activities for their own benefit and not for the benefit of the company, destroying firm value in the process. This is enhanced by the fact that shareholders have little control over the firm’s political activities (Aggarwal et al., 2012; Coates, 2010).

In addition, Hadani & Schuler (2013) theorize that politically active firms tend to take more risks, which can explain the negative returns. Boubakri, Mansi, & Saffar (2013) found evidence supporting this statement. Their results indicate that politically engaged firms are more likely to take additional risks than firms that are not politically engaged. This excessive risk-taking increases the chance of negative firm financial results (Boubakri et al., 2013).

Comparing all of these arguments based on the RBV, it is not clear if firms are able to improve their financial performance through their portfolio of political resources. Several scholars argue in favour of this positive relationship between political resources and firm financial performance(Dahan, 2005; Hillman & Hitt, 1999; Hillman et al., 2004; Lux et al., 2011; Mathur et al., 2013; Schuler, 1996). However, investing too much or without a clear goal in political resources can result in a value destroying activity (Aggarwal et al., 2012).

3.4 Corporate Social Responsibility, Corporate Political Activity and Corporate Financial Performance

In the previous sections of this study the relationships between Corporate Social

Responsibility and Corporate Political Activity with respect to CFP have been discussed. As stated, the positive relationship between CSR and CFP has become acknowledged over time and most of the empirical studies provide evidence to support this statement. On the other hand, the CPA relationship with CFP has provided mixed results across multiple studies, however the tendency in this field of research is to argue that it does positively enhance a firm’s financial performance.

(21)

21 Both of these relationships can be explained from the resource based perspective as, both CSR and CPA can be seen as bundles of resources firms have access to and can deploy in order to obtain improved financial returns (Dahan, 2005; Hillman et al., 2004; McWilliams & Siegel, 2001; Mellahi et al., 2016). Furthermore, they are both classified as non-market strategies, with the goal of improving the firm performance through their CSR and CPA activities. These findings suggest that CSR and CPA are focusing on similar aspects of improving the firm’s performance. However, for a long period of time these two streams of research have been kept separated, as they were not seen as overlapping and inherently different (Mellahi et al., 2016). There are several reasons for the separation of these two non-market strategies.

First, the openness of the programmes are fundamentally different, CSR is generally seen as a positive endeavour by the public. Conducting business in a sustainable way, with as little negative impact on society or even a positive impact, is perceived as a good thing by the public. This has motivated firms to be open about their CSR programmes and show their positive impacts on society, which can for example lead to an improved firm image and an increased employees’ willingness to work there (Aguinis & Glavas, 2012; Brammer &

Pavelin, 2006; Orlitzky et al., 2003). On the other hand, CPA is often perceived negatively by the public, which makes firms reluctant to be open about their political activities. The

question often raised by scholars and the public is if CPA is ethical, because the foundation of CPA rests on firms trying to improve their own market positions by affecting policies in their own favour and not necessarily in favour of society. The idea that politicians and policies can be bought is generally not positively perceived and strengthened by scandals that have

emerged over time as a result of these activities, like the Enron scandal (Alzola, 2013;. Dahan, Hadani, & Schuler, 2013; Oberman, 2004). As a result, the motives and goals of a firm’s political activity are often unclear and not publicly available, making CPA a vague concept

(22)

22 for the public (Aggarwal et al., 2012; Baysinger, 1984). These arguments show a contrast between CPA and CSR with regards to the perception by the public and relevant stakeholders.

Second, Mellahi et al. (2016) argue that there is a difference in approach between CPA and CSR related activities. Both of these strategies are subject to either buffering or bridging mechanisms. Bridging refers to meeting the expectations of the environment, while buffering is an approach for dealing with the environment (Mellahi et al., 2016; Zheng, Singh, & Mitchell, 2015).

As previously stated, CSR is related to the instrumental stakeholder theory, as it can be seen as a means to satisfy the relevant stakeholders of the firm (Harrison et al., 2010; Jones, 1995). Continuing this line of reasoning, it is suggested that CSR is mostly a bridging

mechanism as its goal is meeting the expectations of the environment. As firms nowadays are often expected to withhold certain standards, while conducting business. This for example refers to environmental friendly production or creating healthy working environments (Aguinis & Glavas, 2012).

In contrast, CPA is seen as method for shaping the environment in the firm’s own favour or pre-emptively dealing with changes in the environment that can negatively impact the firm (Coates, 2010; Hillman & Hitt, 1999; Hillman et al., 2004). These approaches can be classified into proactive and defensive behaviour. Proactive behaviour is for example actively lobbying against a certain policy change, like stricter environmental rules, while defensive behaviour is more concerned about maintaining the political status quo (Zheng et al., 2015). As these examples illustrate, the strategies are a buffering mechanism that firms apply on the environment instead of acting on the expectations of the environment (Mellahi et al., 2016). These arguments illustrate the strategic differences between CSR and CPA.

(23)

23 As one can reason, the previously shown differences between CSR and CPA can explain why firms prefer to separate these non-market strategies. In addition, there is often a lack of alignment between these two strategies within a firm. A lack of alignment, can result in conflicting goals within the firm, for example a firm can be profiled as environmental friendly by their CSR programme, while they are politically active with the goal of reducing the emission standards. It is theorized that this misalignment can be especially hard to deal with for managers, as they do not know which strategy to follow or how to integrate them. These conflicting goals and issues can potentially negatively impact the firm’s performance (Hond et al., 2014; Mellahi et al., 2016; Rasche, 2015).

As a response to these negative effects of the misalignment between the two non-market strategies, scholars have started looking at the overlap between these strategies (Hond et al., 2014; Mellahi et al., 2016; Rasche, 2015; Rehbein & Schuler, 2015). They have built theories from the perspective of the resource based view of the firm. As previously argued both CSR and CPA can be seen as bundles of firm specific resources.

Hond et al. (2014) suggest that these resources are complementary to each other and can be used to improve the performance of both the CPA and CSR activities from the firm. From the CSR point of view, the results of these activities can build a positive and well known image of the firm, making it more interesting for politicians to become associated to the firm. This can increase the firm’s access to these political actors (Hond et al., 2014). In addition, it can become easier to build strong lasting relationships with these political actors if the CSR programme positively perceived by the public for a longer period of time. The reasoning for this is that political actors are more willing to partner with firms that are positively perceived by the public, as it enhances their own reputation (Rasche, 2015).

(24)

24 On the other hand, CPA resources can be used to strengthen the effects of the CSR activities. There are two main areas that CPA can help the CSR activities; information and policy making (Hond et al., 2014). CPA can positively affect CSR by providing information obtained through their political activities. By having access to political resources it becomes easier to obtain relevant information with regards to policy making that regards CSR

initiatives, for example stricter environmental rules or additional employee protection (Hillman et al., 2004; Rehbein & Schuler, 2015). Moreover, CPA can be used to support or create policies that enhance the firm’s CSR activities, which can be beneficial with regards to the firm’s financial performance (Hond et al., 2014). As illustrated in the introduction by the example of the US car industry, car manufacturers like Hyundai and Toyota are actively lobbying for the implementation of stricter emission regulations to obtain an advantage over the competition who do not have the required technologies (Boudette, 2017). This overlap between the CSR and CPA is especially relevant for firms with a strong CSR performance, as they are able to obtain significant returns from the CSR activities. As is for example,

illustrated by the Hyundai and Toyota example.

In contrast, for firms with a weak CSR performance, are most likely not able to reap additional benefits from their CSR activities through CPA, as a low CSR performance does not enhance the firm’s financial performance (Orlitzky et al., 2003; Wang et al., 2016). Mellahi et al. (2016) propose that these firms with weak CSR performances may use their political resources to compensate for their lower CSR performance. Findings of Cho et al. (2006) support these statements as firms with weak environmental performances invest significantly more in political resources than firms with better environmental performances. Moreover, the car manufacturing example from the introduction showed that firms like Ford and General Motors are becoming increasingly politically active, as they are not able to meet the emission standards as set by the Obama administration and are trying to remove these

(25)

25 policies through the Trump administration (Boudette, 2017). Finally, David et al., (2007) state that managers of firms, which do not meet the CSR criteria set by external stakeholders, are moving resources away from their CSR initiatives and invest them in Political activities. The reasoning for this is that is easier to reduce the pressure from external stakeholders through political activities than through improving the CSR initiatives. So managers can see political resources as a means for dealing with pressure from external stakeholders and compensate for their weaker CSR performances.

Comparing both sides, weak and strong CSR performances it becomes clear that there is a difference regarding the interaction with CPA. As previously mentioned theories and results suggest that CPA can strengthen a positive relationship between strong CSR and CFP, while CPA can also be used by firms with weak CSR performances as a means to reduce the negative results of these performances. Therefore, using a total CSR variable would not be logical as there seems to be differences between weak and strong CSR performers with regards to CPA and CFP. In following, a decision was made to independently test the interaction between strong and weak CSR performers and CPA with regards to CFP. As a result hypothesis 3 consists of two partial hypotheses:

H3A: The firm’s Corporate Political Activity positively affects the relationship between the firm’s high Corporate Social Responsibility performance and its Corporate Financial Performance.

H3B: The firm’s Corporate Political Activity positively affects the relationship between the firm’s low Corporate Social Responsibility performance and its Corporate Financial Performance.

Both of these hypotheses suggest that CPA positively affects the relationship between CSR and CFP. However, as argued a high CSR performance should be enhanced by the firm’s

(26)

26 political activities, while if the firm’s CSR performance is low their political activities are used to compensate the negative effects. Therefore, it is expected that the proposed

relationship in Hypothesis 3A is stronger than the proposed relationship of Hypothesis 3B.

3.5 Corporate Environmental Performance, Corporate Political Activity and Corporate Financial Performance

In the CSR section of this study it was stated that CSR is a broad term for a bundle of activities across five dimensions. One of these dimensions refers to the natural environment firms interact with (Dahlsrud, 2008). In hypothesis 2 the relationship between Corporate Environmental Performance and Corporate Financial Performance was extensively discussed. Furthermore in hypotheses 3, on the basis of the RBV the overlap between CSR and CPA in relation to the firm’s financial performance was proposed. However, this included all five dimensions of CSR, while it may be the case that the interaction is different for each of the dimensions with regards to CPA.

Mellahi et al. (2016) suggest that especially the link between Corporate Environmental Performance and CPA with regards to financial performance can be interesting to further explore. The reasoning for this statement is that the interaction with the natural environment and the firm is a recurring theme within both the CPA and CSR literature and is generally seen as one of the stronger areas of overlap (Cho et al., 2006; Dixon-Fowler et al., 2013; Hond et al., 2014; Mellahi et al., 2016; Rasche, 2015; Ambec & Lanoie, 2008). The examples shown in the introduction illustrate this overlap, as green technology car manufacturers are lobbying for stricter emission rules to improve their market position over the competition. On the other hand, the example of Nestlé shows that firms can also use their political activities to overcome their deficiencies in the area of environmental performance, as is supported by their ability to keep producing on expired permits for obtaining the natural resources.

(27)

27 Several studies have explicitly looked at the relationship between CPA and the natural environment. As Cho et al. (2006) concluded, firms within environmentally sensitive

industries are spending more on political contributions than firms in other industries. In addition, these firms often have a lower environmental performance and results therefore suggest that they are mainly using the additional contributions for compensating this weak environmental performance (Cho et al., 2006).

Moreover, Child & Tsai, (2005), have found empirical evidence that firms are actively influencing environmental policies in emerging economies. As they suggest, firms conduct their political activities with two different goals, protecting their activities against stricter environmental rules or supporting stricter rules for their own benefit. Their results also indicate that CPA can affect CEP (Child & Tsai, 2005).

Furthermore, Peterson & Pfitzer (2009) have looked at several case studies from practise with regards to CPA and CEP. An interesting case they found was Shell that was influencing policymakers to combine their efforts for stopping the erosion of the coastal wetlands in the US. This erosion increased the amount of natural disasters, harming Shell’s operations in the process and decreasing their employee living conditions. Through their political efforts Shell was able to both improve the environmental conditions and their own operations (Peterson & Pfitzer, 2009). These studies show that CPA can positively affect CEP or be used as compensation for the firm’s CEP.

Referring back to the arguments based on the RBV for hypothesis 2, 3a and 3b, it is argued that CEP can also be seen as a bundle of firm specific resources. Building on the theory provided for hypotheses 3A and 3B and the arguments made in this section, there appears to be an overlap with the CPA resources within the firm, as both can be used to strengthen each other. More specifically, CPA can be used to improve a positive

(28)

28 environmental performance as is shown by the examples of Toyota and Shell. On the other hand, the Nestlé example shows that CPA can be used to compensate a weak environmental performance, which otherwise would have negatively impacted the firm’s financial

performance, for example the loss to access of natural resources for Nestlé.

Similarly to hypothesis 3, this hypothesis is also split up in two partial hypotheses, as it appears to be the case that the interaction between CEP Strengths and CEP Concerns and CPA with regards to CFP is different. Therefore the following two partial hypotheses are presented:

H4A: The firm’s Corporate Political Activity positively affects the relationship

between the firm’s high Corporate Environmental Performance and its Corporate Financial Performance.

H4b: The firm’s Corporate Political Activity positively affects the relationship between the firm’s low Corporate Environmental Performance and its Corporate Financial Performance.

Finally, the proposed relationship in H4A should be stronger than the proposed relationship in H4B as CPA should enhance a strong positive relationship between a firm’s high CEP and CFP, while it is theorized to be a compensator for a weak relationship between a firm’s low CEP and CFP. All of these proposed hypotheses are illustrated in the following three figures:

(29)

29 Figure 1. Conceptual model

Figure 2. Conceptual model

(30)

30

4. Methodology

4.1 Sample and data collection

For this study, the database has been built by groups of students over the past few years under specific guidance of the thesis supervisor. This year the database has been expanded by a group of seven students, who all collected data with the same specific methodology. This methodology is based on a very strict and elaborate coding manual provided by the thesis supervisor. Data collection is therefore completed for all areas of research within corporate governance, corporate political activity and corporate social responsibility.

The sample of companies consisted of Fortune 500 publicly listed firms. The Fortune 500 companies had been chosen for several reasons. First, they are obligated by law to make all of their actions publicly available, which enabled us to obtain the required data, for example the financials, corporate governance and managerial incentives. For this study, the financial data was obtained from the Compustat Wharton Databases. Second, CSR data was obtained through the Kinder, Lydenberg, Domini and Co (KLD) databases, which are also based on USA firms. This is an independent organisation, which rates firms based on strengths and concerns with regards to CSR. This methodology is in line with many CSR studies (Cai, Jo, & Pan, 2012; McWilliams & Siegel, 2001; Minor & Morgan, 2011; Orlitzky e.a., 2003; Wang et al., 2016). Finally, for the measurement of CPA, USA based companies are needed, because the company’s political activity is publicly available in the USA. This type of data collection is in line with many CPA studies, which have also used USA based companies (Aggarwal, Meschke, & Wang, 2012; Hadani & Schuler, 2013; Schuler, 1996). These datasets were collected from OpensSecrets, which is a research group concerned with tracking money in U.S. politics (Opensecrets, 2018).

After collecting all the data, the datasets were merged into one file. Firms that either merged or were acquired, were removed from the data. The rationale for this is that the data

(31)

31 from these merged or acquired firms could not be compared anymore to the previous years. In addition, non-profits firms were also excluded from the sample as they are not comparable to the other firms within the sample. In addition, I had to exclude several firms, which had missing financial, CPA or CSR data points, which will be further discussed in the variables section. Furthermore, the average values of the variables were taken for the years 2014-2016, in order to improve the reliability of the sample. As a results of these adjustments, a final sample of 393 publicly listed USA firms for the years 2014-2016 was created.

4.2 Dependent Variable

Corporate Financial Performance (CFP)

In this study a decision had to be made for either using accounting measures or market measures as a proxy for Corporate Financial Performance. Within the CSR literature there is no wide acceptance for either adopting accounting measures or market measures. Scholars argue that both have their advantages and disadvantages (Orlitzky et al., 2003; Tsoutsoura, 2004; Wang et al., 2016). Financial results based on accounting measures are reflected by historical returns, are a result of the firm’s internal decision making and are often short-term orientated (Orlitzky et al., 2003). Furthermore, accounting measures are prone to being manipulated by managers through the different accounting processes (Tsoutsoura, 2004). On the other hand, market measures are based on the external environment, especially how shareholders value the firm’s performance. These measures are also more long-term

orientated as they take into account how shareholders see the future performance of the firm. However, this measurement assumes that shareholders appropriately value the firm’s

(32)

32 In this study the decision was made to use a market measurement for Corporate

Financial Performance, as it is a more long-term oriented measurement. Tobin’s q is utilized as a proxy this market measure. Tobin’s q is the ratio of the market value of the firm’s assets, which are measured by the value of its outstanding stock and debt, divided by the replacement cost of the company’s assets (Chung & Pruitt, 1994). This proxy for market measures is widely used and accepted by scholars for testing CFP (Cai et al., 2012; Chung & Pruitt, 1994; Lang & Stulz, 1994; Orlitzky et al., 2003; Wahba, 2008; Wang et al., 2016; Wernerfelt & Montgomery, 1988).

4.3 Independent Variable

Corporate social responsibility (CSR)

Corporate Social Responsibility is an independent variable in this study. First, CSR was measured through the KLD database ratings. In this database the strengths and concerns regarding six issues of CSR are given for the years 2014 till 2016. Based on previous CSR studies like Cai et al. (2012), a decision was made to exclude Corporate Governance indicators from this sample, as these do not overlap with other CSR indicators. As a result CSR consists of five indicators; Environment, Community, Human Rights, Employees and Product (diversity). Within each of these categories and corresponding year, several indicators of strengths and weaknesses are measured by KLD’s independent analysists with a binary scale of 0 and 1. A 1 rating indicates that the strength or concern is present, if a 0 rating is given this means that the strength or concern is not present. In following of this process a summation for each category can be made of both the strengths and concerns.

Furthermore, based on the study of Chin, Hambrick, & Treviño (2013) a decision has been made to subtract the total concern scores from the total strength scores. The reasoning for this is that strengths are seen as a positive indicator, while concerns as a negative indicator

(33)

33 for a company’s CSR score. By subtracting the concern scores from the strength scores a total CSR score is found per issue and year, in which the height of the CSR score indicates the quality of a company’s CSR Performance with respect to the specific year and issue (Chin et al., 2013). This enabled me to compare the companies within this dataset with each other in terms of total CSR scores per issue and year.

An issue with the KLD dataset was that a lot of data for specific indicators was missing. Therefore, this study implemented a threshold of at least 40 percent available data for an indicator, in order to create a sample that would be large enough for an analyses. Unfortunately, as a result of these missing data points the total predictors available for this were 23 out of the 57 could be used. Notably, no predictors for Product strengths and Human strengths were available. On the other hand, most of the predictors for concerns were

available for this dataset. This is consistent with other CSR studies that use the KLD databases for the basis of CSR data as Wang et al. (2016) concluded in their meta-analysis.

Continuing based on this proposed methodology, summations have been made for total concern scores and total strength scores per CSR issue and corresponding year. This will result in a total CSR Performance score per year and issue for each company within the dataset. These total scores per issue and year are then compiled into a total score CSR

Performance Score. Afterwards, the average of the CSR Performance scores over the 2014 till 2016 time period were taken for suitable companies within this dataset. As a result, each company within the dataset will get a total CSR rating based on an average of a three year period.

High and Low Corporate Social Responsibility

As argued in the theoretical framework, this study also tests the differences between a high and a low level of CSR performance. More specifically, as Orlitzky et al. (2003) and Wang et

(34)

34 al. (2016) conclude in their meta analyses, there are disagreements on how to measure CSR based on the KLD databases. Some studies only use KLD strengths for measuring CSR (Cai et al., 2012; Jo & Na, 2012) , while others subtract the concerns from the strengths for a total CSR rating (Harjoto & Jo, 2011; Chin et al., 2013). Moreover, Mattingly & Berman (2006) provide empirical evidence that the KLD strengths and concerns should be separately used for analysing CSR effects. In general, KLD strength indicators are positively impacting CSR performance, while the KLD concern indicators are negatively impacting a firms CSR performance (Cai et al., 2012; Harjoto & Jo, 2011; Jo & Na, 2012; Mattingly & Berman, 2006; Orlitzky et al., 2003; Wang et al., 2016). Therefore, this study used KLD strength indicators for a high CSR performance and KLD concern indicators for a low CSR performance. As a result, two variables were created CSR Strengths and CSR Concerns.

Corporate Environmental Performance (CEP)

The same methodology for this variable was used as for CSR. However, only the

environmental indicators of KLD were included, as previously used by Cho et al. (2006). Unfortunately, only four of the fifteen environmental strength indicators were available for this sample. In contrast, seven of the nine environmental concern indicators. Similarly to CSR, a variable for total Corporate Environmental Performance was created by subtracting the environmental concern indicators from the environmental strength indicators. Afterwards, two variables were created for High and Low Corporate Environmental Performance, based on the KLD environmental strength indicators and KLD environmental concern indicators. As a result two variables were created, CEP Strengths and CEP Concerns.

(35)

35 4.4 Moderating Variable

Corporate Political Activity (CPA)

Within the literature there are two main proxies for testing the firm’s political activity; PAC contributions and Lobbying expenses and both will be independently used as moderators within this study (Aggarwal et al., 2012; Cho et al., 2006; Hadani & Schuler, 2013; Mathur et al., 2013).

First, the PAC contributions, this data was obtained from Opensecrets (2018), which is a non-profit organisation that registers all PAC contributions from individuals and firms to officials running for office (F.E. Senate and Congress) and the two political parties. This data was coupled to the specific donating firm, which enabled the comparison with similar firms in the sample. This is the main methodology for analysing PAC data in the CPA field of research (Aggarwal et al., 2012; Cho et al., 2006; Hadani & Schuler, 2013; Mathur et al., 2013). Based on this data the variable CPA PAC contributions was created.

Second, lobbying expenses were also obtained from Opensecrets (2018) and coupled to the appropriate firms. This methodology was the same as for PAC contributions and also enabled the comparison with the other firms in the sample. As a result the variable CPA Lobbying was created.

Final, we were able to obtain additional Corporate Political Activity data, namely the amount of issues firms lobby for and specifically the environmental issues. This data was collected for hypotheses 4a and 4b as it focuses on the natural environment. There are three issue categories with regards to the natural environment; Clear Air & Water, Environmental/ Superfund and Waste. This method as a proxy for CPA was not previously used within the CPA field of studies. Therefore, I had to create the proxy based on logical reasoning, however due to the novelty of the method the theoretical rigour is low. In order to create a testable

(36)

36 variable, the issues regarding the natural environment (C&A, S/W and W) were summed up and divided by the total issues a specific firm was lobbying for. This created a ratio of environmental issues with regards to the total amount of issues. As a result, the variable Environmental Issues was created. In addition, this variable was multiplied by the total

lobbying expenditures of a specific firm, in order to approximate the lobbying expenditure per environmental issue. This resulted in the variable Lobby Expenditures Environmental Issues. As previously stated, this variable lacks theoretical rigour, however it interesting to test if lobbying on environmental issues affects Corporate Environmental Performance with regards to Corporate Financial Performance.

4.5 Control Variables

Firm Size

Firm size is often measured in relationship to CSR and is established as a factor that

significantly affects CSR and CFP (Orlitzky et al., 2003; Wang et al., 2016). As Udayasankar (2008) study shows, larger firms have more visibility, resource access and scale operations. These differences in firm size can therefore significantly affect the CSR performances of firms. Therefore, this study controls for Firm Size by using logged numbers of employee, as logged values are normally distributed.

Industry

Industry is also an established factor that affects both CSR and CFP and should be controlled for (McWilliams & Siegel, 2001; Orlitzky et al., 2003; Tsoutsoura, 2004). In this study nine dummy variables were created based on the SIC codes of the firms within the sample, as shown below:

(37)

37

Dummy SIC Industry

1 0100-0999 Agriculture, Forestry and Fishing 2 1000-1499 Mining

3 1500-1799 Construction 4 1800-1999 not used 5 2000-3999 Manufacturing

6 4000-4999 Transportation, Communications, Electric, Gas and Sanitary service 7 5000-5199 Wholesale Trade

8 5200-5999 Retail Trade

9 6000-6799 Finance, Insurance and Real Estate 7000-8999 Services

(38)

38

5. Results

5.1 Descriptive statistics and correlation analysis

Before conducting the multiple regression analyses, as proposed in the hypotheses, the variables had to be checked for measurements mistakes or missing data. Four cases for the control variable Firm Size Log Employees were missing and thus excluded from the analysis. Next descriptive statistics were run to analyse the variables and check their kurtosis and skewness. The variables; Corporate Social Responsibility Strengths, Corporate

Environmental Performance Strengths, Firm Size (Log Employees), Ratio of Environmental Issues, displayed no deviations of the acceptable skewness and kurtosis ranges of -2.58 to 2.58, since the sample is larger than 200 (Ghasemi & Zahediasl, 2012). In contrast, the other variables in this study did show problems with skewness and kurtosis. Creating boxplots of these variables showed that extreme outliers were causing these problems. However, these outliers are not a result of measurement mistakes, but a result of individual firm behaviour. For example, some firms spend a lot more on lobbying than other firms. Therefore, these are valuable data points for this study and should not be removed. Furthermore, most of these variables have a lot of 0 values within the sample, which makes it not possible to create natural logarithms of these variables. The suggested solution for these problems is explained by Field (2013), through winsorizing the extreme outliers to a value of the variable mean plus three standard deviations. This methodology was applied on the remaining variables with extreme outliers and did not surpass the acceptable 5% threshold for winsorizing (Field, 2013).

First, the descriptive statistics of Tobin’s q showed that it had a heavy tail to the right (skewness = 1.583, kurtosis = 2.998). This was the result of extreme outliers, winsorizing the top 1,2 percent of the data improved the distribution within acceptable ranges (skewness = 1.358, kurtosis = 1.742). Second, the distribution of CSR Concerns also showed a heavy tail

(39)

39 to the right (skewness = 1.885, kurtosis = 4.388). Similarly, this was the result of several extreme outliers. Winsorizing the top 1.2 percent of the data resulted in a well-shaped distribution (skewness = 1.885, kurtosis 1.487). Third, the descriptive statics of CEP Concerns displayed a heavy tailed distribution to the right (skewness = 2.386, kurtosis = 6.354). By winsorizing the extreme outliers to acceptable values (2.5 percent of the data), a normal distribution was created (skewness = 1.448, kurtosis = 1,487). Fourth, Lobbying Expenditure, showed a heavy tailed distribution to the right (skewness 2,541; kurtosis 7,146, which was also a result of extreme outliers and a high amount zero values. Winsorizing the top 3.5 percent of the data resulted in a slight heavy tailed distribution to the right (skewness = 1.849, kurtosis 2.679). However, this is mainly a result of the high amount of zeros within this variable and therefore the slightly heavy tailed Lobby Expenditure distribution variable will be used in this study. Finally, descriptive statistics of the variable PAC Expenditure showed slight positive skewness and a strongly heavy tailed distribution (skewness = 3.036, kurtsosis = 11,144). Similarly to the other variables, a boxplot showed some extreme outliers. By winsorizing the top 3.5 percent of the data, a normal distribution was formed (skewness = 1.728, kurtosis = 2.226. After trimming data of the variables CSR Concerns and CEP

Concerns, a new variable for total CSR (strengths minus concerns) and total CEP (strengths minus concerns) had to be created as the extreme values were not taken into account for CSR Concerns and CEP Concerns. These new variables did not deviate from the allowed skewness and kurtosis ranges and were not violating any normality tests.

Furthermore, after analysing the variable Lobby Expenditure Environmental Issues, it became clear that it was strongly skewed to the right and had a very heavy tailed distribution (skewness = 4.939, kurtosis = 30.367). In addition, the sample size for this variable was smaller than that of all of the other variables (N=280), thus removing all of the extreme outliers would violate the 5% trimming rule (Field, 2013). Therefore, the decision

(40)

40 was made to exclude this variable from the analysis, as the data was not meeting the criteria and could significantly affect the results.

In table 1 the descriptive statics and correlations are presented. The industry control variable has been excluded from the correlations, as it consists of 9 dummy variables representing 10 industry groups, which would create a very large correlation matrix.

However, in the analyses this variable is controlled for and described as included, within the tables (Walls, Berrone, & Phan, 2012). An interesting descriptive statistic is that the mean 1.795 for CSR Strengths is higher than the mean 0.922 for CSR Concerns. This suggests that on average firms within this sample are performing well with regards to concerning CSR indicators, while also obtaining positive results for CSR Strength indicators.

First, analysing the correlations of Tobin’s q gave several expected and unexpected results. As expected, Tobin’s q is positively correlated to CEP (r = .204, p<0.01), suggesting that when Tobin’s q is higher the firms also have a higher CEP. Moreover, Tobin’s q

positively correlates with both CEP strengths (r = 0.173, p<0.01) and CSR strengths (r = 0.150, p<0.01). In contrast, CSR does not correlate with Tobin’s q (r= 0.098, n.s.), this can be explained by the fact that CSR Concerns (r = 0.069, n.s.) does not correlate with Tobin’s q and as CSR is a summation of CSR Strengths minus CSR Concerns. These results indicate that Tobin’s q is neither higher nor lower for firms with regards to CSR. Second, by observing the correlations between CSR Strengths and CSR Concerns with regards to the CPA variables strong correlations were found. As expected both CPA PAC (r = 0.280, p<0.01) and CPA Lobbying (r=0.336, p<0.01) are both positively correlated to CSR Strengths, suggesting that firms with higher CSR Strengths have higher lobbying and PAC expenditures. Moreover, as expected the positive correlation between CPA PAC (r = 0.448, p<0.01), CPA Lobbying (r = 0.554, p<0.01) and CSR Concerns is stronger than that for the CSR Strengths. Indicating that firms with higher CSR Concerns have higher lobbying and PAC expenditures. Third, as CEP

(41)

41 Strengths and CSR Concerns are both part of CSR Strengths and CSR Concerns and thus have similar correlation results with the CPA variables PAC and Lobbying, as is illustrated in table 1.

However, for CEP Strengths and CEP Concerns an additional variable was added, the ratio of lobbying on Environmental Issues. Environmental issues are as expected strongly correlated with CEP Concerns (r=0.450, p<0.01), suggesting that firms with higher CEP Concerns have higher ratios on Environmental Issues. On the other hand, no correlation between Environmental Issues and CEP Strengths was found (r= -0.046, n.s.), which argues that higher CEP Strengths does not mean firms have higher ratios on Environmental Issues. Finally, the control variable Firm Size (Log Employees) was analysed, this variables

correlates with all of the other variables (as shown in table 1) except CSR (r= 0.069, n.s.) and CEP Concerns (r=-0.074, n.s.).

Corporate social responsibility, corporate political activity and corporate financial performance : an empirical study on the environmental dimension of CSR

MSc. Thesis: