Corporate Social Responsibility improves financial results, market reacts skeptical

(1)

Corporate Social Responsibility Improves Financial Results,

Market Reacts Skeptical

Name: Thérèse van Woerkom

Student Number: 10561757

Date: 30th of January, 2017

Bachelor Thesis

BSc Economics and Business, Economics and Finance

Faculty of Economics and Business

University of Amsterdam

(2)

Statement of Originality

This document is written by Thérèse van Woerkom who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is 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)

Abstract

Corporate Social Responsibility (CSR) can be defined as the extent to which a firm includes

economic, environmental, social and corporate governance elements into its business practice. Given the growing CSR activities and demands, the question can be raised whether companies can indeed ‘do well by doing good’. Evidence is found that CSR has a positive effect on the companies’ financial performance, but that the market does not value those higher returns. This study aims to contribute to this area of research by exploring the effect of different elements of CSR on financial performance. The question is what elements of CSR improve financial performance. Panel data analysis for firms listed in the S&P500 over the years 2010 until 2013 is conducted. The Kinder, Leydenberg, Domini (KLD) rating was used to measure CSR and the return on assets (ROA) as well as Tobin’s q were used to measure the companies’ financial performance. The KLD rating was disentangled into the

environment, social and corporate governance dimension. The social dimension was further disaggregated into five categories: community, human rights, employee relations, diversity and product. By doing so, this study finds that firms should focus on social related CSR activities to improve the companies’ financial returns. Community, employee relations and diversity related CSR activities are found to be the main drivers of the positive relationship between CSR and financial returns.

(4)

1. Introduction

Nowadays companies are not only evaluated by their financial performance, but also by the impact they have on social, environmental and economic aspects. Therefore, Corporate Social Responsibility (CSR) has received increased attention over the last decade (Porter and Kramer, 2011). The goal of this paper is to add evidence to the ongoing debate about the effect of CSR on financial performance. A positive and significant correlation is found between CSR and return on assets (ROA) over the next year, which indicates a positive relationship between CSR and financial performance. This study contributes to this area of research by exploring the effect of the different categories of CSR on financial performance. The question is what elements of CSR improve financial performance. Evidence is found that community, employee relations and diversity related CSR activities improve financial results, while product related CSR activities decrease financial results. This study further shows that although CSR activities overall improve financial results; the market reacts skeptical to those socially responsible expenditures.

The implementation of CSR in companies has grown exponentially over the last decade. Companies continuously improve their CSR activities and demonstrate this by publishing assessments and reports (Perrini, Russo, Tencati and Vurro, 2011). A report of the Governance & Accountability Institute shows that nearly 20% of the companies listed in the S&P500 published a Corporate

Sustainability Report in 2011 (Governance & Accountability Institute, 2015). In 2014 this number has grown to 75% of the companies listed in the S&P500. The most recent analysis executed by the Governance & Accountability Institute (2015) finds that 81% of the companies included in the S&P500 published a Corporate Sustainability Report in 2015.

Currently no worldwide uniform definition of CSR exists and across the globe various approaches to CSR emerge, taking into account the interests of customers, employees, suppliers and the community in which the firm operates (McWilliams and Siegel, 2001; Perrini et al., 2012). Porter and Kramer (2006) mention that CSR should not be a stand-alone uncoordinated principle, but that it should be integrated in the firms’ core business operations. Research conducted by McKinsey shows the same trend, finding that US executives realize that CSR should be more strategic (McKinsey, 2012; McKinsey, 2014). The number of respondents citing the goal to align CSR with their business strategy was 43% in 2014 compared to 30% in 2012 (McKinsey, 2012; McKinsey, 2014). The study further shows that in 2014 13% of the CEOs mentioned CSR as their top priority compared to 5% in 2012.

Given the growing CSR demands and activities and the evolvement of CSR into the

mainstream business practice, the question can be raised whether companies can indeed ‘do well by doing good’ and if engaging in CSR activities is advantageous in terms of corporate performance.

(6)

The neoclassical economists’ belief of the position of a firm is that a corporation has no social responsibilities and their actions should merely be focused on maximizing shareholders’ value, and therefore maximize profits (Friedman, 1970; Barnett and Salomon, 2012). Stakeholder theory,

however, emphasizes that a company should extend its attention beyond the traditional commitment to shareholders, taking into account a broad spectrum of stakeholders (Freeman, 1984; Wirl, Feichtinger and Kort, 2013).

CSR is a multidimensional measure and researchers have tried to assemble the different elements into one single measure (Mahoney and Roberts, 2007; Scholtens, 2008). Mahoney and Roberts (2007) summarize the challenge facing CSR related research that folding the different categories of CSR into a one-dimensional measure covers the effects of various categories that are especially important to a company. The sophisticated concept of CSR requests an empirical analysis toward a better understanding of the activities that drive the increase of firms’ financial performance. Although meta-analyses found an overall positive relationship between CSR and financial

performance, the results of several studies examining this relationship are ambiguous. Those mixed results can be caused by the problem of defining and measuring CSR and unsound empirical research (McWilliams and Siegel, 2000).

This thesis contributes to this area of research by unraveling the effect of the different

categories of CSR on financial performance. Panel data analysis is conducted to estimate the effect of environment, social and corporate governance related CSR activities on the financial performance of companies listed in the S&P500 from 2010 until 2013. Performance is measured by ROA as well as Tobin’s q. The Kinder, Leydenberg, Domini (KLD) rating assesses the firms’ level of CSR

engagement. This study provides guidelines for managers to optimize financial performance through CSR engagement.

The remainder of this thesis is organized as follows. Section 2 provides a literature review on the relationship between CSR and firms’ financial performance and clarifies the concept of CSR by explaining the different dimensions. Section 3 describes the hypotheses and section 4 outlines the methodology and data description. Section 5 presents and discusses the results. Lastly, the conclusion is given in section 6.

(7)

2. Literature Review

The definition of CSR has evolved over time and varies across nations, regions and sectors. CSR is a multidimensional measure and the different categories interact with financial performance in various ways. This thesis defines CSR as the extent to which a firm includes economic, environmental, social and corporate governance elements into its business operations. Meta-analyses examining the link between CSR and financial performance show an overall positive effect. Several explanations for the positive effect of CSR on financial performance are given, such as cost reduction, increased firm reputation and creating competitive advantages.

2.1 The definition of CSR

Researchers have difficulties in finding a uniform description of CSR, as it is part subjective and can vary across nations, regions and sectors (Roobeek and De Swart, 2013). Consequently, indicators and measures of CSR vary broadly and capture either one specific element of CSR or broad estimations of CSR as a whole (Margolis, Elfenbein and Walsh, 2007).

A widely used description presented by Carroll (1979) explains the concept using different elements of CSR. The framework contains a pyramid built on four elements: economic-, legal-, ethical-, and discretionary responsibilities. The economic responsibility is fundamental and refers to the companies’ obligation of being profitable. Legal responsibilities refer to a firms’ duty to fulfill its economic goals within the confines of the law. The third building block contains the firms’ ethical responsibilities to go beyond the legal obligations and to prevent harmful influences. The last category, discretionary responsibilities, refers to social responsible initiatives, such as philanthropy, which a company voluntary undertakes even if there are no clear societal expectations (Carroll, 1979).

The model presented by Carroll (1979) provides insight into the link between CSR and the firms’ financial performance by examining the exchange between the various layers of the pyramid. The four elements of CSR illuminate the motivations for taken actions within each category. The economic responsibilities are obligatory, the ethical responsibilities are expected and the discretionary responsibilities are desired, meaning that companies should not invest in CSR activities at the expense of financial performance (Carroll and Shabana, 2010). In addition, the model specifies that the firms’ economic responsibility is a factor that has to be considered in CSR, since the firms’ financial performance is not only important for themselves, but also for the society as a whole (Carroll and Shabana, 2010).

CSR can also be seen as a business and investment strategy that includes the task of providing competitive financial results while preserving, retaining and improving the human and natural

resources needed in the future (Artiach and Walker, 2010). Accordingly, CSR can be defined as the extent to which a firm includes economic, environmental, social and corporate governance elements

(8)

into its business operations, and the impact they have on the firm and society (Artiach and Walker, 2010). Since those separate elements are operationalized in the KLD index, this definition of CSR is used for this research.

2.2 CSR and financial performance

A profound comprehension of the effect of CSR on financial performance is essential, because of the increased pressure on firms to invest in CSR activities and due to the incorporation of CSR into core business strategies. Between 1971 and 2001, more than 120 empirical studies have been conducted to measure the relationship between CSR and financial performance (Margolis and Walsh, 2002). Although those studies show ambiguous results, meta-analyses examining the link between CSR and financial performance found an overall positive effect of CSR on the financial performance of firms (Orlitzky, Schmidt and Rynes, 2003; Allouche and Laroche, 2005 and Margolis et al., 2007).

Lo and Sheu (2007) conducted a panel data analysis to the effect of CSR on the market value of non-financial firms listed in the S&P500 over the years 1999 until 2002. Tobin’s q was used as a proxy for market value and the Dow Jones Sustainability Index as a proxy for CSR engagement (Lo and Sheu, 2007). Lo and Sheu (2007) concluded that a significant positive relationship exist between CSR engagement and the firms’ market value and that investors positively react to the incorporation of CSR into business strategies.

Furthermore CSR can influence the firms’ ability to attract capital (Barnett and Salmon, 2012; Ghoul, Guedhami, Chuck, Kwok and Mishra, 2011). Ghoul et al. (2011) examined the effect of CSR on the cost of equity. The study included 12,915 firms in the U.S. over the years 1992 until 2007 and pooled regression analysis was used to study the companies’ cost of equity. Evidence is found that firms with higher levels of CSR engagement exhibit cheaper equity financing (Ghoul et al., 2011). More specific, the results show that investing in employee relations, product and environmental related CSR activities contribute substantially to the reduction of firms’ costs of equity (Ghoul et al., 2011).

Barnett and Salomon (2012) found that the relationship between CSR and financial

performance is U-shaped. They conducted a panel data regression including 1,214 firms over the years 1998 until 2006. This study used the KLD rating as proxy for CSR and the ROA and net income as proxies for financial performance (Barnett and Salomon, 2012). Evidence is found that firms with a low KLD score have higher financial performance than firms with moderate KLD scores, but firms with high KLD scores perform financially the best (Barnett and Salomon, 2012).

Brammer, Brooks and Pavelin (2006), however, found that companies with high CSR scores tend to have lower stock returns. Brammer et al. (2006) investigated 451 companies listed in the FTSE ALL Share Index as of July 2002 at the firm level. In addition, research conducted by Bercchetti, Di

(9)

Giacomo and Pinnacchio (2008) shows a negative link between CSR and financial performance. Bercchetti et al. (2008) investigated this link using a sample of approximately 1,000 companies over a 13 year-period. Evidence is found that CSR increases total sales per employee, while it significantly lowers the return on equity (ROE). Those results corresponds with the neoclassical economists’ view that CSR redirects the focus of firms’ activities from maximizing shareholders’ value to stakeholders’ interests (Ullman, 1985; Hull and Rothenberg, 2008). One possible explanation for these findings is that CSR implies, on the one side, decisions increasing the cost of labour and intermediate output. On the other side, CSR may enhance participation, motivation and identification of the workforce with firm goals, which positively affects productivity (Berchetti at al., MSCI ESG Research, 2015).

2.3 Different categories of CSR

As the description in section 2.1 shows, CSR is a multidimensional concept and firms can take a wide variety of actions to be socially responsible. This study uses the categories operationalized in the KLD rating to clarify the concept of CSR and its relationship with financial performance. The KLD rates the environmental, social and governance (ESG) performance of companies, using the following seven categories: environment, community, human rights, employee relations, diversity, product and corporate governance. Appendix 1 outlines the seven KLD CSR categories and the corresponding strengths and concerns within each category.

2.2.1 Environment

Important environmental issues are: clean energy, pollution prevention, recycling, packaging materials and waste (MSCI ESG Research, 2015). Dow Chemical, for example, reduced their fresh water consumption by one billion gallons, leading to savings of $4 million (Porter and Kramer, 2011). Another example is the decision of a firm to become more energy efficient, having a positive impact on the environment as well as on financial performance due to reduced costs (Bird, Hall, Momentè and Reggiani, 2007).

2.2.2 Social

The social subcategory can be split in the following subjects: community, diversity, employee relations, human rights and product.

Community

Since no company is self-contained, companies’ success depends upon the society around it. Companies should invest in community related CSR activities to create shared value (Porter and Kramer, 2011). A healthy society needs successful companies and successful companies need a healthy society (Porter and Kramer, 2006). Capable local suppliers, for example, improve logistical efficiency (Porter and Kramer, 2011). In addition, a well-established education system expands the

(10)

supply of skilled employees (Porter and Kramer, 2011). The community plays a crucial role in driving productivity, innovations and competitive advantages (Porter and Kramer, 2011).

Diversity

Diversity can improve a company’s legitimacy and reputation, which would in turn positively influence its labour market position. Areas of strengths are superior employee advantages addressing work and family matters, tolerant gay and lesbian policies and initiatives in the employment of hiring disabled people (MSCI ESG Research, 2015). The implementation of those activities increases the firms’ attractiveness as a potential employer and thereby improves the competitive positioning and financial performance by attracting and retaining more talented employees (Perrini et al., 2011).

Employee relations

The presence of strong employees’ support programs helps to create an ethical climate and enhances employees’ willingness to actively participate in organizational activities (Perinni et al., 2011). This can lead to operational benefits, such as increased productivity and lower costs due to reduced employee turnover (Perinni et al., 2011). Johnson and Johnson, for example, implemented several health care programs resulting in savings of $250 million on health care costs (Porter and Kramer, 2011). This corresponds to a return of $2.71 for every dollar spent on those programs between 2002 and 2008 (Porter and Kramer, 2011).

Human Rights

The human rights element includes two areas of strengths: the relation with indigenous people and human rights policies and initiatives (MSCI ESG Research, 2015). The first indicator classifies companies that have formed relations with indigenous people in the area the firm operates in that respect the jurisdiction, culture, land, intellectual property and human rights (MSCI ESG Research, 2015). Human rights policies and initiatives indicate firms that have initiated extraordinary human rights initiatives or have in some way shown to be the industry leader on human rights issues not enclosed in other KLD human rights ratings (MSCI ESG Research, 2015).

Product

Companies can create products or services with features showing that the firm is involved with social issues (McWilliams and Siegel, 2000). According to McWilliams and Siegel (2000), this strategy helps consumers to believe that the consumption of the product directly or indirectly contributes to society. This enhances firm reputation, which in turn, could increase profitability (McWilliams and Siegel, 2000; Bird et al., 2007). Consequently, companies label their products to show some form of social responsibility. Natural food companies, for instance, specify the use of organic and pesticide-free ingredients (McWilliams and Siegel, 2000). So, areas of product strengths are product and process innovation within the industry (Waddock and Graves, 1997). Areas of concerns are among other things product safety, marketing and contracting controversies (MSCI ESG Research, 2015).

(11)

2.2.3 Corporate Governance

Berk and DeMarzo (2014) define corporate governance as the system of regulations, controls and incentives to prevent fraud. KLD outlines two areas of strength: Corruption & Political Instability and Financial System Instability (MSCI ESG Research, 2015). The first indicator analyses how firms manage the risk of suffering operational disruptions due to property destruction, violence, political instability and corruption and bribery. Companies score higher on this aspect if they help employees with those issues by incorporating programs, guidelines and policies. Financial System Instability analyses how a firm manages its systemic risk within financial markets (MSCI ESG Research, 2015). Good implementation of corporate governance helps to reduce the risk of unexpected financial losses and lowers the costs of operations in regions where corruption and political instability are present (Perrini et al., 2011).

2.4 Gap in the existing literature

The aggregation of the various elements of CSR into one single measure covers the different

interaction patterns between those elements and financial performance (Scholtens, 2008; Mahoney and Roberts, 2007). Scholtens (2008) examined the interaction and direction of causation between the social subcategories of CSR and financial performance using a sample of 289 firms in the U.S. covering the years 1991 until 2004. The study used stock return and financial risk, measured by the standard deviation of the stock price, as proxies of financial performance. CSR engagement is measured by the KLD rating (Scholtens, 2008). Evidence is found that the interaction patterns vary along the different subcategories of the social CSR dimensions (Scholtens, 2008). Scholtens (2008) argues that product related CSR activities precede financial returns, whereas financial returns precede community and product related CSR activities.

In addition, Mahoney and Roberts (2007) examined the effect of CSR on financial

performance by conducting a panel data analysis to publicly held Canadian firms over the years 1996 until 1999. This study used ROA and ROE to measure financial performance and the Canadian Social Investment Database (CSID), similar to the KLD database, to measure CSR (Mahoney and Roberts, 2007). No evidence is found for an overall significant relationship between CSR and financial performance (Mahoney and Roberts, 2007). However, the results show a significant positive

relationship between environmental and international related CSR activities and financial performance (Mahoney and Robert, 2007).

The results of those studies suggest that the categories of CSR are best analysed individually, contributing to the comprehension of the overall effect of CSR on financial performance (Scholtens, 2008; Mahoney and Roberts, 2007). Isolation of the different elements of CSR contributes to management understanding what elements of CSR they can focus on to be more socially responsible

(12)

while also increasing financial performance. It provides guidance for managers to optimize financial performance through CSR engagement.

(13)

3. Hypotheses

Based on the literature described in section 2, four testable hypotheses are formulated. The first null hypothesis is that CSR does not have an effect on the financial performance of firms. Although mixed results to the link between CSR and financial performance are found, three meta-analyses show an overall positive link between CSR and financial performance (Orlitzky et al., 2003;Allouche and Laroche, 2005; Margolis et al., 2007). Therefore, it is expected that a positive relationship can be found for companies listed in the S&P500. Hence, the first alternative hypothesis states that CSR has a positive relationship on financial performance of companies.

Hypothesis 1: CSR has a positive effect on financial performance

This research is extended by an analysis examining whether different categories of CSR affect financial performance in various ways. As described in section 2.4, research has shown that different elements of CSR interact with financial returns in various ways (Scholtens, 2008; Mahoney and Roberts, 2007). Mahoney and Roberts (2007) found that only a significant and positive relationship exists between environmental and international related CSR activities and financial performance for Canadian firms. This thesis builds upon the study by Mahoney and Roberts (2007) by examining the different effect of the various categories of the KLD rating on financial performance for firms listed in the S&P500 using an accounting-based as well as a market-based proxy for financial performance. To examine this, the KLD rating has been disaggregated into the three main categories: environment, social and corporate governance. Based on the existing literature, the direction of the relationship between the three elements and financial performance is unclear. Therefore, the following three hypotheses are formulated:

Hypothesis 2: The environment category of CSR has an effect on financial performance Hypothesis 3: The social category of CSR has an effect on financial performance Hypothesis 4: The corporate governance category of CSR has an effect on financial performance

(14)

4. Research & Methodology

This section outlines the research methodology and explains the variables used in the regression models. An accounting-based measure, the ROA, as well as a market-based measure, Tobin’s q, are used as proxies of financial performance. The KLD rating is used as a proxy of CSR. Since CSR does not immediately affect financial performance, a time lag of one year is used. Firm size, risk, industry and research& development (R&D) are included in the models as control variables. This study examines companies within the S&P500 over the years 2010 until 2013. This data was derived from the COMPUSTAT database.

4.1 Dependent variable: Financial Performance

The dependent variable of this study is financial performance. In the existing literature accounting-based measures (e.g. return on sales or return on assets) as well as market-accounting-based measures (e.g. price per share or Tobin’s q) are widely used measures of financial performance (Orlitzky et al., 2003). Both measures capture different elements of financial performance. Accounting-based measures capture the firms’ internal efficiency and focuses on the historical performance (Scholtens, 2008). However, this evaluation method is subject to biases due to various accounting policies across firms (Orlitzky et al., 2003; Scholtens, 2008). Since market-based measures capture investors’ assessments and expectations of firm performance, those measures are less affected by accounting procedures (Scholtens, 2008). In addition, market-based measures include investors’ expectations of the future financial performance of the company, rather than past performance (Akpinar et al., 2008). However, market-based measures may also be biased due to, for instance, asymmetric information (Scholtens, 2008).

To provide a complete picture, this study includes an accounting-based as well as a market-based measure. Studying both provides more guidance for managers in making decisions about the allocation of their strategic resources, since decisions regarding CSR engagement can depend on the firms’ goal to foster accounting-based performance or to favor the market value of the company.

The accounting-based performance measure most frequently used in the existing literature and showing the strongest link between CSR and financial performance is ROA (Orlitzky et al., 2003). ROA represents the returns of the company with respect to the total set of resources, or assets, under its control (Hull and Rothenberg, 2008). Hull and Rothenberg (2008) mention that ROA provides the most direct information about the allocation of strategic resources. ROA is calculated by dividing the net income (NI) by the total assets (TA) of the company.

In line with the existing literature, Tobin’s q is used as market-based measure in this study. Tobin’s q is defined as the ratio of the market value of a firm to the replacement cost of its assets (Lo and Sheu, 2007). The proxy for the replacement costs of the assets is the book value of total assets, in line with previous studies (Lo and Sheu, 2007; Akpinar 2008). Tobin’s q (TQ) is calculated by

(15)

dividing the sum of the market value (MKVALT) and book value of the liabilities (LT) by the sum of the book value of equity (BKVLPS) and the book value of the liabilities (LT).

4.2 Independent variable: CSR

The KLD rating is used as measure to evaluate the companies’ level of CSR. KLD is the biggest publicly available database containing multidimensional information about CSR and is named as the

de facto standard in research to CSR at this time (Deckop, Merriman and Gupta, 2006; Waddock,

2003).

KLD rates the environmental, social and governance (ESG) performance of companies using roughly 70 indicators, containing strengths or concerns, divided within the following seven categories: environment, community, human rights, employee relations, diversity, product and corporate

governance. KLD assesses those indicators using quantitative as well as qualitative data sources, such as surveys, interviews, financial reports, CSR reports, government documents and news from several media channels. Taking all this information into account, a binary scoring model is used to evaluate the indicators within the seven categories:

1 if a company meets the assessment criteria established for an indicator

0 if a company does not meet the assessment criteria established for an indicator

So a company can get a score of 0 or 1 for every indicator within one of the categories. For every firm in each year of the sample within every category, those scores are added leading to an overall rating for the total strengths and total weaknesses within that category. For example: ‘Total Number of Community Strengths’ and ‘Total Number of Community Concerns’. Next, the variable CSR is obtained in the following way. The firm’s total strength score is constructed by adding its number of strengths along all seven categories. The firm’s total concerns score is determined by adding its number of concerns along all seven categories. Lastly, the total number of concerns is subtracted from its total number of strengths of the company to obtain the main independent variable CSR.

In line with the existing literature a one-year lag for CSR and financial performance is used, since changes in CSR do not instantly affect financial performance (Waddock and Graves, 1997; Shahzad and Sharfman, 2015). Instead, the financial performance is dependent on the level of CSR of the previous year. Research on CSR and financial performance conducted by Bird et al. (2007) found that using time lags of more that one year yielded no significantly different outcomes than using a one year lag. The purpose of using the lagged value is to avoid potential endogeneity and to assure that the causal direction runs from CSR and the control variables to financial performance, and not vice versa (Baron, Harjoto and Jo, 2011; Shahzad and Sharfman, 2015).

(16)

4.3 Control Variables

According to the existing literature the variables risk, company size, research & development (R&D) and industry have impact on firm performance as well as CSR (Ullman, 1985; Margolis et al., 2007; Orlitzky, 2001; Wu, 2006; Waddock and Graves, 1997; McWilliams and Siegel, 2000). The control variables are not lagged, since they have an immediate effect on the companies’ financial performance (Waddock and Graves, 1997).

Two arguments to the effect of size on CSR can be mentioned. First, simply the fact that those bigger companies have more financial funds to invest in CSR (Waddock and Graves, 1997). Second, bigger companies receive more public attention increasing the pressure to invest in CSR (Wu, 2006). Furthermore, larger companies are expected to perform financially better than smaller firms do, due to economies of scope, economies of scale and more market power (Berk and DeMarzo, 2014). The meta-analysis conducted by Orlitzky et al. (2003) found that the natural logarithm of total assets is a frequently used measure to control for company size. Hence, the natural logarithm of total assets of the firm is added to the regression model to control for company size.

Risk is another variable affecting both CSR and financial performance (Margolis et al., 2007). Research conducted by Waddock and Graves (1997) shows a significant negative relationship between risk and the firms’ financial performance. Risk also impacts CSR; firms with a low risk profile, measured by the debt to assets ratio, are more likely to invest in CSR activities (Margolis et al, 2007). To control for risk the total long-term debt is divided by the total assets and included in the models.

Furthermore it is argued that R&D expenditures should be included in the regression model to examine the relationship between financial performance and CSR (McWilliams and Siegel, 2000; Van and Gössling, 2008). R&D investments result in knowledge improvement, which leads to more productivity and improved financial performance (McWilliams and Siegel, 2000). In addition, McWilliams and Siegel (2000) argue that CSR investments boost R&D activities. To control for R&D, R&D expenditures are divided by the total assets and included in the model.

Lastly, dummy variables are included in the regression model to control for industry effects (Margolis et al., 2007). The industries are based on the four-digit Standard Industrial Classification (SIC) code. The SIC code divides the market into eleven industries. In this sample, however, there are no observations within the agriculture and public administration industry. Besides, there are only four observations within the non-classifiable establishments. Therefore, those three industries are not included in this study and eight industries remain. Table 1 gives an outline of the distribution of firms across industries within the sample of firms listed in the S&P500.

(17)

Table 1. Allocation of companies across industries

Industry Number of

Observations Mean KLD Mean ROA Mean TQ

Mining 102 1.96 0.042 3.13

Construction 24 -0.91 0.061 2.95

Manufacturing 704 3.45 0.082 4.74

Transportation & Public Health 261 3.15 0.040 2.41

Wholesale Trade 34 1.09 0.073 3.85

Retail Trade 151 1.88 0.102 6.68

Finance, Insurance & Real Estate 371 2.30 0.031 3.00

Services 203 2.54 0.085 6.44

As can be seen in Table 1, the KLD score differs across industries. To further examine the effect of CSR within each industry, appendix 2 presents the correlations between the KLD score and the two measures of financial performance along the eight industries. Only significant correlations are found within the construction, manufacturing and transportation industry. The significant correlation within the construction and manufacturing industry is expected, since firms within those industries have major impact on the environment through energy and resource use, and receive more external pressures to engage in CSR activities (Brammer et al., 2006; Porter and Kramer, 2011). CSR

investments in areas such as waste management, packaging materials and natural resource use reduce the firms’ operational costs (Porter and Kramer, 2011; MSCI ESG Research, 2015).

4.4 Model

The report of the Governance & Accountability Institute (2015) shows notable growth in companies’ CSR engagement from 2010. Moreover, as of writing the KLD rating is available until 2013.

Therefore, the regressions include the panel data for the firms listed in the S&P500 for a period of four years, namely 2010 until 2013. Four panel data regressions are estimated to test the hypotheses.

4.4.1 Empirical model hypothesis 1

Model 1 is estimated to test the effect of CSR on the accounting-based measure ROA.

!"#_$,& = ( + *₊,-._$,&/++ *₁!23,_$,&+ *₄3256_$,&+ *₇!._$,&+ *₈92: + *_;<!#:3 + *₌36!> + *_? !6< + *_@ A2: + *_+B CD"< + *₊₊ E":3 + F_$,&

(18)

Model 2 is estimated to test the effect of CSR on the market-based measure Tobin’s q.

<G$,& = ( + *+,-.$,&/++ *1!23,$,&+ *43256$,&+ *7!.$,&+ *892: + *;<!#:3 + *=36!>

+ *? !6< + *@ A2: + *+B CD"< + *++ E":3 + F$,&

Where

ROA = return on assets

TQ = Tobin’s q

KLD = KLD score

RISK = a proxy for the companies’ level of risk SIZE = a proxy for the companies’ size

RD = a proxy for the companies’ level of R&D

FIN = dummy variable for the finance, insurance & real estate industry TRANS = dummy variable for the transportation & public health industry SERV = dummy variable for the services industry

RET = dummy variable for the retail trade industry MIN = dummy variable for the mining industry

WHOT = dummy variable for the wholesale trade industry CONS = dummy variable for the construction industry

4.4.2 Empirical model hypotheses 2, 3 and 4

Model 3 is estimated to test the second, third and fourth hypotheses to the effect of the different elements of CSR on the accounting-based measure ROA.

!"#_$,& = ( + *₊6:>_$,&/++ *₁3"E_$,&/++ *₄EH">_$,&/++ *₇!23,_$,&+ *₈3256_$,&+ *_;!._$,& + *=92: + *?<!#:3 + *@36!> + *+B !6< + *++ A2: + *+1 CD"<

+ *₊₄ E":3 + F_$,&

Model 4 is estimated to test the second, third and fourth hypotheses to the effect of the different elements of CSR on the market-based measure Tobin’s q.

<G$,& = ( + *+6:>$,&/++ *13"E$,&/++ *4EH">$,&/++ *7!23,$,&+ *83256$,&+ *;!.$,&+ *=92:

+ *?<!#:3 + *@36!> + *+B !6< + *++ A2: + *+1 CD"< + *+4 E":3 + F$,&

The analysis is taken one step further by examining the effect of the different subcategories of the social element. Therefore, the social element is disaggregated into the five subcategories of the KLD score: community, human rights, employee relations, diversity and product. This is done for the ROA as well as the TQ, as can be seen in the specifications below.

(19)

!"#_$,& = ( + *₊6:>_$,&/++ *₁.2>_$,&/++ *₄E"A_$,&/++ *₇6AI_$,&/++ *₈DJA_$,&/++ *_;I!"_$,&/+ + *₌EH">_$,&/++ *_?!23,_$,&+ *_@3256_$,&+ *_+B!._$,&+ *₊₊92: + *₊₁<!#:3 + *+436!> + *+7 !6< + *+8 A2: + *+; CD"< + *+= E":3 + F$,&

<G$,& = ( + *+6:>$,&/++ *1.2>$,&/++ *4E"A$,&/++ *76AI$,&/++ *8DJA$,&/++ *;I!"$,&/+

+ *₌EH">_$,&/++ *_?!23,_$,&+ *_@3256_$,&+ *_+B!._$,&+ *₊₊92: + *₊₁<!#:3 + *+436!> + *+7 !6< + *+8 A2: + *+; CD"< + *+= E":3 + F$,&

Where:

ENV = the environment category of CSR SOC = the social category of CSR

CGOV = the corporate governance category of CSR

DIV = the diversity subcategory of CSR

COM = the community subcategory of CSR

EMP = the employee relations subcategory of CSR HUM = the human rights subcategory of CSR PRO = the product subcategory of CSR

(20)

5. Results

5.1 Descriptive statistics

The average KLD score for the sample differs per year. Table 2 shows the trend of the average KLD score for firms listed in the S&P500 over the years 2010 until 2013. As can be seen, the minimum KLD score within this sample is -8 and the maximum score is +19. The minimum KLD score of -8 was only observed for four companies and the maximum KLD score was observed twice. All six companies are active within the manufacturing industry. The size of the companies with the lowest KLD score varies between total assets worth $4,350 to $9,350 million, whereas the companies with the highest KLD score have total assets worth $1,904 million and $71,119 million. This indicates that even small companies can incorporate a high level of CSR.

In addition, it can be seen that the average value of the KLD score increases over the years. The average KLD score in 2013 is 119.8% higher compared to the average KLD score in 2010. This indicates that companies indeed incorporate more CSR activities.

Table 2. The average KLD score for firms listed in the S&P500 over the years 2010-2013

Year Mean KLD score Min KLD Score Max KLD score

2010 1.67 -8 17

2011 2.60 -8 19

2012 3.14 -5 16

2013 3.67 -6 19

2010-2013 2.78 -8 19

Table 3 displays to what extent the different dimensions of CSR contribute to the overall KLD score as shown in Table 2. The largest component of the overall KLD score is the environment

dimension with an average score of 0.94. The corporate governance dimension scores the lowest with an average KLD score of -0.21. A large increase can be seen for the subcategory employee relations. In 2010 this subcategory scored below zero on average: -0.09. Three years later, in 2013, the average score has increased to a positive value of 1.94. This indicates that firms increased their attention to employee related CSR activities over the years. A decrease can be seen for the diversity dimension, a decrease of the KLD score from 1.03 in 2010 to 0.43 in 2013. Table 3 also shows that the community and human rights dimensions stayed constant over time. The product dimension had a negative average KLD score in all years and it does not seem to show noticeable progress.

(21)

Table 3. The average KLD score per dimension for firms listed in the S&P500 over the years 2010-2013

Mean KLD Score Min Max

2010 2011 2012 2013 2010-2013 ENV 1.15 1.28 0.52 0.83 0.94 -5 5 COM 0.49 0.56 0.46 0.45 0.49 -1 4 HUM -0.02 0.02 0.04 0.06 0.03 -2 2 EMP -0.09 0.11 1.67 1.94 0.93 -4 8 DIV 1.03 1.06 0.60 0.43 0.77 -3 7 PROD -0.52 -0.01 -0.12 -0.04 -0.17 -4 2 CGOV -0.36 -0.42 -0.03 -0.04 -0.21 -4 2

Note: ENV=Environment; COM=Community; HUM=Human Rights; EMP = Employee Relations; DIV=Diversity; PROD=Product; CGOV=Corporate Governance

Appendix 3 displays the correlation matrices. The correlation matrix shows that the correlation between the KLD score and ROA is positive (0.069) and significant at a 1% level. This indicates a positive relationship between the companies’ KLD score and their ROA over the next year. The control variables size and risk are significantly negatively correlated with the financial performance. The variables have an opposite effect on the KLD score. Size has a positive correlation with the KLD score, whereas risk has a negative correlation with the proxy for CSR. This is in line with the existing literature arguing that bigger companies invest more in CSR activities and that firms with more risk engage less in CSR activities. However, the negative correlation between size and ROA is surprising.

To test for possible multicollinearity the variance inflation factors (VIF) were estimated for all independent variables in the models. The results of the VIF tests, outlined in appendix 4, show no proof of disturbing multicollinearity (Field, 2014). Further, a Breusch-Pagan test was used to analyze whether heteroskedasticity of the residuals was present. The results of the Breusch-Pagan test are shown in appendix 5. All tests estimate a p-value of 0.000 and therefore the null hypothesis of

constant variance of residuals over the observations is rejected in all four cases. So, the Breusch-Pagan test shows heteroskedasticity of the residuals in all four models. Therefore, robust standard errors are used in the regression models, as the length of the time period measured is limited.

5.2 Regression analysis

Appendix 6 presents the regression results to the effect of CSR on firms’ financial performance. The results are generated by running four regression models using panel data analysis for firms listed in the S&P500 over a four-year period, from 2010 until 2013. Model 1 and model 3 include ROA as measure of financial performance and model 2 and 4 include Tobin’s q. The first two models examine the overall effect of CSR on financial performance and the last two models examine the effect of the different categories of CSR.

(22)

Since no companies within the construction industry contain data about R&D expenditures, this dummy variable is omitted when controlling for R&D. In addition, a scan of the results shows that controlling for R&D expenditures lowers the explanatory variable of the models estimating the effect of CSR on ROA and that this control variable is insignificant in these models. Therefore, the

regressions without the R&D expenditures are used to analyze the effect of CSR on the firms’ ROA. R&D expenditures, however, do have a significant effect in the models estimating the effect of CSR on Tobin’s q. R&D expenditures send a signal to the market that the firm focuses on technological knowledge improvement. This information increases market participants’ expectations about the firms’ future economic performance; positively affecting Tobin’s q. Consequently, the regressions including R&D expenditures are used to analyze the effect of CSR on Tobin’s q.

Model 1 demonstrates that CSR has a significant positive effect on the firms’ ROA (0.0094; p<0.01). Model 2 shows that CSR has a significant negative effect on Tobin’s q (-0.054; p<0.05). The disaggregation of the CSR in model 3 indicates that only the social element of CSR has a significant effect on ROA. Specifically, employee relations (0.0021; p<0.05), diversity (0.0031; p<0.01) and to a lesser extent community (0.0025; p<0.1) related CSR activities positively affect ROA, whereas product (-0.0060; p<0.01) related CSR activities have a negative effect on ROA. Lastly, the results of model 4 show that only environmental related CSR activities have a significant impact on Tobin’s q and that this effect in negative (-0.2604; p<0.05).

5.2.1 Hypothesis 1

Evidence is found that CSR has a significant and positive effect on the firms’ ROA over the next year. However, CSR negatively affects Tobin’s q. This finding contradicts the conclusion of Lo and Sheu (2007) that investors positively react to the incorporation of CSR into business strategies. Lo and Sheu (2007), however, used the Dow Jones Sustainability Index as proxy of CSR. Setting up a dummy variable for CSR engagement causes loss of information and might bias the results. Using the KLD rating, which shows more variance, should therefore achieve more reliable results.

The results suggest that firms taking a broad spectrum of stakeholders into account create an organizational resource causing higher return on assets. However, the market does not evaluate those higher financial returns. One possible argument for those results is that the market might be slow to respond to the benefits associated with CSR activities. If this is true, a positive relationship might be identified when using a time lag longer than one year. Therefore, model 1 and model 2 were estimated again using a time lag of two years. However the results, presented in appendix 7, show no significant relationship between CSR and Tobin’s q. A positive, but less significant, result is found for the relationship between CSR and ROA in this situation. The findings suggest that the market might be more neoclassical than expected. Although proof exists that CSR has a positive effect in terms of financial returns of a company, the market does not expect those higher returns.

(23)

In conclusion, evidence is found to reject the first null hypothesis that CSR does not have an effect on the financial performance of firms.

5.2.2 Hypotheses 2, 3 and 4

Model 3 and 4 were estimated to examine the effect of the different categories of CSR on financial performance. What might be regarded surprising is that environment related CSR activities have a significant negative effect on Tobin’s q, which indicates that the market punishes companies investing in the environment. Bird et al. (2007) found in their study that environmental strengths as well as environmental concerns have a negative effect on market valuation. Those results in combination with this study’s findings suggest that the market acknowledges the necessity to meet environmental regulatory requirements, but punishes firms that extend strategic resources beyond the minimum requirements within this area (Bird et al., 2007). Since environment related CSR activities do not affect the ROA, management should focus on the signal it sends to market participants.

Evidence is found that community, employee relations and diversity related CSR activities have a positive effect on the ROA over the next year. Activities related to these three dimensions are not so costly, and pertain to mostly policy. Therefore, the results can have an instant effect on performance. These activities havea positive effect on labor productivity. The employment of minorities, for example, is at no added cost to recruitment, nevertheless, it sends an immediate signal of socially responsibility to employees and increases the firms’ attractiveness as a potential employer. Attracting and retaining more talented employees improves the competitive positioning and thereby the returns of the company (Perrini et al., 2011; Porter and Kramer, 2011).

Product related CSR activities, however, have a negative effect on the ROA over the next year. Product related CSR activities are expensive and the effort can only be seen long after

investments have been made. Product innovation, for instance, can take several years and increased net income can only happen once the product in launched.

Corporate governance related CSR activities are not significantly related to the ROA and Tobin’s q. This can be explained by the fact that activities within this dimension are more concerned with avoiding future negative outcomes and that these initiatives are fundamental to the company. Therefore, those activities are not rewarded by improved financial performance and do not cause market participants to increase their expectations about the firms’ performance. This is in line with research conducted by Margolis et al. (2007), finding that corporate policies have no significant impact on financial performance. Accordingly, Orlitzky et al. (2003) argue that in future studies corporate governance should not count towards CSR. This study supports that claim.

Except for environmental, no significant effect of the different elements of CSR on Tobin’s q is found. This could be explained by the fact that market participants have difficulties in evaluating the

(24)

added value of these mostly policy based categories. The base for a consideration of costs and benefits is lacking, so the potential value is hard to determine. By contrast market participants see investments in environment related issues as negative for market performance.

In conclusion, no evidence is found to accept the second hypothesis that the environment category of CSR has an effect on firms’ financial performance measured by ROA. However, the second null hypothesis is rejected when Tobin’s q is used as dependent variable. The third null hypothesis to the effect of the social category of CSR on financial performance, measured by ROA, is rejected. No evidence is found to reject the third null hypothesis when Tobin’s q is used to measure financial performance. Lastly, no evidence is found to reject the fourth hypothesis to the effect of the corporate governance category of CSR on financial performance.

(25)

6. Discussion and conclusion

This study finds that CSR drives financial performance of companies, but negatively affects market performance. Collapsing the different categories of CSR into a one-dimensional measure covers the effects of the various categories that are especially important to a company. The findings of this study provide directions for companies to optimize financial performance through CSR engagement and managers should take the results reported above into account when taking decisions about the allocation of their strategic resources. The results show that the relationship is complicated, since different elements of CSR drive financial performance in various ways. However, practical conclusions can be drawn.

Managers should focus on social related CSR activities to improve the companies’ financial returns. Particular attention should be given to the community, employee relations and diversity related CSR activities. Meantime, managers should realize that product related CSR activities are likely to decrease the companies’ financial performance over the next year. However, product innovation can be crucial to keep up with competition in the longer run. Managers should further realize that although CSR improves companies’ financial returns overall, CSR investments do not boost market participants’ expectations. Management should be particularly careful with investments in environment related CSR activities. Firms should show their environmental concerns, without overinvesting.

Several limitations of this study should be considered. The main limitation of this thesis is related to the construction of the KLD score. Although the KLD rating is mentioned as the de facto standard in research to CSR at the time, the construction of this rating also received criticism. The KLD rating fails to account for any differential levels of importance that may be attached to a specific indicator. If the assessment criterion for an indicator is established, the firm receives a score of 1. Companies that perform even better than this benchmark do not receive a higher score for the particular indicator. Therefore the KLD indicator does not vary enough, which means a lot of information is lost.

Moreover, the reader should bear in mind that this study is based on firms listed in the S&P500. As the S&P500 covers the largest public companies in America, the generalizability of this thesis is limited. Smaller firms and private companies operate under different circumstances. Private companies’ shares are not traded on a public exchange market and their financial disclosure is not included in databases. Future research can increase the scope of the analysis to other countries and contexts using alternative data sources on CSR.

A potential direction is the role of managerial intentions. Based on the current data it was not possible to determine whether companies intended to use CSR as a signal of social responsibility, as a

(26)

mechanism to improve financial performance or both. Future research could focus on examining the underlying motivations for engaging in CSR activities. These could mainly be explored in qualitative research.

To address the increased societal expectations companies face nowadays, firms are increasing their CSR activities. This study finds that firms can indeed ‘do well by doing good’. Nevertheless, the market has yet to fully appreciate their efforts.

(27)

References

Artiach, T., & Walker, J. (2010). The determinants of corporate sustainability performance.

Accounting and Finance, 50, 31-51.

Barnett, M. L., & Salomon, R. M. (2012). Does it pay to be really good? Addressing the shape of the relationship between social and financial performance. Strategic Management Journal,

33(11), 1304-1320.

Baron, D. P., Harjoto, A. M., & Jo, H. (2011). The economics and politics of corporate social performance. Business and Politics, 13(2).

Bercchetti, L., Giacomo, S. Di, & Pinnacchio, D. (2008). Corporate Social Responsibility and

Corporate Performance: Evidence from a Panel of US Listed Companies. Applied Economics,

40(5), 541-567.

Berk, J., & DeMarzo, P. (2014). Corporate Finance (3rd ed.). Harlow, United Kingdom: Pearson Education.

Beurden, P., & Goessling, T. (2008). The worth of values - A literature review on the relation between corporate social and financial performance. Journal of Business Ethics, 82(2), 407-424. Bird, R., Hall, A. D., Momentè, F., & Reggiani, F. (2007). What Corporate Social Responsibility

activities are valued by the market? Journal of Business Ethics, 76(2), 189-206.

Brammer, S., Brooks, C., & Pavelin, S. (2006). Corporate Social Performance and Stock Returns: UK evidence from disaggregate measures. Financial Management, 35(3), 97-116.

Carroll, A. B. (1979). A Three-dimensional conceptual model of corporate performance. Academy of

Management Review, 4(4), 497-505.

Carroll, A. B., & Shabana, M. (2010). The Business Case for Corporate Social Responsibility: A Review of Concepts, Research and Practice. International Journal of Management Reviews,

12(1), 85-105.

Deckop, J. R., Merriman, K. K., & Gupta, S. (2006). The effects of CEO pay structure on corporate social performance. Journal of Management, 32(3), 329-342.

Field, A. (2014). Discovering Statistics Using IBM SPSS Statistics. London, United Kingdom: Sage Publications.

Freeman, R. E. (1984). Strategic Management: A stakeholder approach.

Friedman, M. (1970). The social responsibility of business is to increase profits. New York, 122-124. Ghoul, S., Guedhami, O., Kwok, C. Y., & Mishra, D. (2011). Does Corporate Social Responsibility

affect the cost of capital? Journal of Banking & Finance, 35(9), 2388-2406.

Governance & Accountability Institute. (2015). Flash Report. Retrieved from

http://www.ga- institute.com/nc/issue-master-system/news-details/article/flash-report-eighty-one-percent-81-of-the-sp-500-index-companies-published-corporate-sustainabi.html

Hull, C., & Rothenberg, S. (2008). Firm performance: The Interactions of corporate social

performance with innovation and industry differentiation. Strategic Management Journal, 29, 781-789.

(28)

Lo, S. F., & Sheu, H. J. (2007). Is Corporate Sustainability a Value-Increasing Strategy for Business?

Corporate Governance: An International Review, 15(2), 345-358.

Margolis, J. D., Elfenbein, H. A., & Walsh, J. P. (2007). Does is pay to be good? A meta-analysis and redirection of research on the relationship between corporate social and financial performance.

Working Paper.

Margolis, J. D., & Walsh, J. P. (2002). Misery loves companies: Whither social initiatives by business? Working Paper, Harvard University.

McKinsey. (2012). The business of sustainability: McKinsey Global Survey Results. Retrieved from http://www.mckinsey.com/business-functions/sustainability-and-resource-productivity/our-insights/the-business-of-sustainability-mckinsey-global-survey-results

McKinsey. (2014). The business of sustainability: McKinsey Global Survey Results. Retrieved from http://www.mckinsey.com/business-functions/sustainability-and-resource-productivity/our-insights/sustainabilitys-strategic-worth-mckinsey-global-survey-results#0

McWilliams, A., & Siegel, D. (2000). Corporate social responsibility and financial performance: correlation or misspecification? Strategic Management Journal, 21, 603-609.

McWilliams, A., & Siegel, D. (2001). Corporate social responsibility: A theory of the firm perspective. Academy of management review, 26(1), 117-127.

MSCI ESG Research. (2015). MSCI ESG KLD STATA: 1991-2014 [Methodology]. Retrieved from https://wrds-web.wharton.upenn.edu/wrds/ds/kld//index.cfm

Orlitzky, M. (2001). Does firm size comfound the relationship between corporate social performance and firm financial performance? Journal of Business Ethics, 33(2), 167-180.

Orlitzky, M., Schmidt, F. L., & Rynes, S. L. (2003). Corporate social and financial performance: meta-analysis. Organization Studies, 24(3), 403-441.

Perrini, F., Russo, A., Tencati, A., & Vurro, C. (2011). Deconstructing the Relationship Between Corporate Social and Financial Performance. Journal of Business Ethics, 102, 59-76.

Porter, M. E., & Kramer, M. R. (2006). Strategy and society: the link between competitive advantage and corporate social responsibility. Harvard Business Review, 84(12), 78-92.

Porter, M. E., & Kramer, M. R. (2011). Creating Shared Value. Harvard Business Review, 89(1), 62-77.

Roobeek, A., & De Swart, J. (2013). Sustainable Business Modeling. Amsterdam, The Netherlands: Academic Service.

Shahzad, A. M., & Sharfman, M. P. (2015). Corporate Social Performance and Financial Performance Sample Selection Issues. Business & Society, 1-30.

Strauss, J. R. (2015). Challenging Corporate Social Responsibility: Lessons for public relations from

the casino industry. Abingdon, United Kingdom: Routledge.

Ullmann, A. A. (1985). Data in search of a theory: A critical examination of the relationship among social performance, social disclosure, and economic performance of US firms. Academy of

Management Review, 10(3), 540-557.

Waddock, S. A., & Graves, S. B. (1997). The corporate social performance-financial performance link. Strategic Management Journal, 18(4), 303-319.

(29)

Waddock, S. (2003). Myths and realities of social investing. Organization & Environment, 16(3), 369. Wirl, F., Feichtinger, G., & Kort, P. M. (2013). Individual firm and market dynamics of CSR

activities. Journal of Economic Behavior & Organization, 86, 169-182.

Wu, M. L. (2006). Corporate social performance, corporate financial performance, and firm size: A meta-analysis. Journal of American Academy of Business, 8(1), 163-171.

(30)

Appendix 1

KLD Research & Analytics, Inc. Ratings Criteria

Strengths Concerns

Environment Environment Opportunities; Waste Management;

Packaging Material and Waste; Climate Change;

Environmental Management Systems; Natural Resource Use;

Other Strengths

Regulatory Compliance; Toxic Emissions and Waste; Energy and Climate Change; Impact of Product and Services; Biodiversity and Land Use; Operational Waste;

Supply Chain Management; Water Stress;

Other Concerns

Community Community Engagement;

Innovative Giving; Support for Education; Other Strengths

Community Impact; Investment Controversies; Tax Disputes;

Other Concerns

Diversity Board of Directors;

Woman and Minority Contracting; Employment of Underrepresented Groups;

Gay and Lesbian Policies; Work-Life Benefits; Other Strengths

Workforce Diversity; Board of Directors;

Employee Relations Union Relations; Cash Profit Sharing; Labor Management;

Human Capital Development; Employee Involvement; Compensation and Benefits Professional Development; Employee Health and Safety; Other Strengths

Union Relations Concerns; Employee Health and Safety; Supply Chain;

Child Labor;

Labor Management Relations; Other Concerns

Human Rights Indigenous People Relations Strength; Human Rights Policies & Initiatives

Support for Controversial Regimes Freedom of Expression and Censorship; Human Rights Violations;

Other Concerns

Product Product Safety and Quality;

Social Opportunities; R&D Innovation; Other Strengths

Product Quality and Safety; Marketing and Advertising; Anticompetitive Practices; Other Concerns

Corporate Governance

Corruption & Political Instability; Financial System Instability;

Governance Structures Controversies; Reporting Quality;

Controversial Investments; Bribery and Fraud;

(31)

Appendix 2

Correlation across industries

Industry Correlation with KLD

ROA TQ

Mining -0.0086 -0.1095

Construction 0.5047** 0.3859**

Manufacturing 0.0760** -0.0329

Transportation & Public Health -0.0553 -0.1263**

Wholesale Trade 0.1335 -0.0201

Retail Trade -0.0787 -0.0824

Finance, Insurance & Real Estate

-0.0464 -0.0054

(32)

Appendix 3

Correlation Matrix

KLD ROA TQ ENV SOCIAL CGOV RISK SIZE RD

KLD 1.000 ROA _0.069*** _1.000 TQ -0.023 0.419*** 1.000 ENV 0.659*** 0.067** -0.030 1.000 SOCIAL 0.877*** 0.046** -0.057** 0.420*** 1.000 CGOV 0.524*** 0.032** -0.015 0.238*** 0.590** 1.000 RISK -0.116*** -0.234*** -0.326*** -0.056** -0.105** -0.009 1.000 SIZE 0.223*** -0.311*** -0.351*** 0.081*** 0.234** 0.026 -0.080*** 1.000 RD 0.162*** 0.033** 0.229*** 0.115*** 0.053 -0.006 -0.261*** -0.145*** 1.000 Note: ** Significant at p < 0.05 *** Significant at p < 0.01

(33)

Correlation Matrix

KLD ROA TQ ENV COM HUM EMP DIV PROD CGOV RISK SIZE RD

KLD 1.00 ROA 0.069*** 1.00 TQ -0.023 0.419*** 1.00 ENV 0.659*** 0.067** -0.030 1.00 COM 0.620*** 0.067** -0.018 0.390*** 1.00 HUM 0.248*** 0.006 0.039 0.120*** 0.102*** 1.00 EMP 0.558*** 0.043* 0.067*** 0.140*** 0.171*** 0.101*** 1.00 DIV 0.634*** 0.011 -0.134*** 0.301*** 0.392*** -0.031 0.120*** 1.00 PROD 0.232*** 0.023 0.106*** 0.041 0.011 0.174*** 0.094*** -0.196*** 1.00 CGOV 0.524*** 0.032** -0.015 0.238*** 0.240*** 0.150*** 0.154*** 0.177*** 0.166*** 1.00 RISK -0.116*** -0.234*** -0.326*** -0.056** -0.106*** 0.017 -0.075*** -0.125*** 0.022** -0.009 1.00 SIZE 0.223*** -0.311*** -0.351*** 0.081*** 0.201*** -0.126*** 0.114*** 0.456*** -0.342*** 0.026 -0.080*** 1.00 RD 0.162*** 0.033** 0.229*** 0.115*** 0.122*** 0.070** 0.254*** 0.013 0.015 -0.006 -0.261*** -0.145*** 1.00 Note: ** Significant at p < 0.05 *** Significant at p < 0.01

(34)

Appendix 4

VIF Values Model 1

VIF 1/VIF KLD 1.26 0.795827 SIZE 1.18 0.845377 RISK 1.32 0.757885 YEAR2011 1.58 0.634094 YEAR2012 1.63 0.614239 YEAR2013 1.67 0.600149 FIN 1.29 0.772235 TRANS 1.03 0.969474 SERV 1.07 0.935897 RET 1.21 0.827006 MIN 1.06 0.942717 WHOT 1.06 0.939561 RD 1.36 0.736413 Mean VIF 1.29

VIF Values Model 2

VIF 1/VIF KLD 1.25 0.797217 SIZE 1.19 0.842519 RISK RD 1.26 1.32 0.793675 0.760153 YEAR2011 1.56 0.641461 YEAR2012 1.61 0.620417 YEAR2013 1.58 0.631580 FIN 1.21 0.828171 TRANS 1.03 0.968283 SERV 1.06 0.947009 RET 1.19 0.839168 MIN 1.06 0.942229 WHOT 1.07 0.938436 Mean VIF 1.26

(35)

VIF Values Model 3

VIF 1/VIF ENV 1.74 0.574350 COM 1.59 0.629857 HUM REL 1.15 0.867405 EMP REL 1.53 0.655321 DIV 1.90 0.527282 PROD 1.28 0.779895 CGOV 1.41 0.711090 SIZE 1.78 0.560719 RISK 1.32 0.755106 YEAR2011 1.82 0.548309 YEAR2012 1.94 0.516063 YEAR2013 2.19 0.457209 FIN 1.38 0.724140 TRANS 1.04 0.957533 SERV 1.14 0.876755 RET 1.24 0.876755 MIN 1.11 0.902258 WHOT 1.08 0.922544 Mean VIF 1.43

VIF Values Model 4

VIF 1/VIF ENV 1.74 0.575263 COM 1.59 0.630607 HUM REL 1.15 0.866958 EMP REL 1.51 0.662932 DIV 1.87 0.533626 PROD 1.27 0.787506 CGOV 1.42 0.705139 SIZE 1.77 0.563525 RISK 1.27 0.788378 YEAR2011 1.81 0.553798 YEAR2012 1.92 0.521415 YEAR2013 2.05 0.487230 FIN 1.26 0.792072 TRANS 1.05 0.955889 SERV 1.13 0.886623 RET 1.22 0.820252 MIN 1.11 0.901797 WHOT 1.08 0.921894 Mean VIF 1.43

(36)

Appendix 5

Breusch-Pagan Test Model 1

H0: Constant variance

Chi2(13) = 921.34

P-value = 0.0000

Breusch-Pagan Test Model 2

Chi2(13) = 1145.84

P-value = 0.0000

Breusch-Pagan Test Model 3

Chi2(19) = 919.58

P-value = 0.0000

Breusch-Pagan Test Model 4

H0: Constant variance Chi2(19) = 1212.05 P-value = 0.0000

(37)

Appendix 6

Regression Output Model 1

Dependent Variable ROA

(1) (2) (3) KLD 0.00093*** (0.00032) 0.00115*** (0.00043) 0.00936*** (0.000310) RISK -0.07783*** (0.01621) -0.07763*** (0.00877) SIZE -0.00810*** (0.00147) -0.01098*** (0.00109) RD -0.12076 (0.12073) MIN -0.01718* (0.01220) -0.02714*** (0.00563) CONS -0.03637** (0.01438) TRANS -0.01777 (0.01497) -0.02699*** (0.00342) WHOT -0.020576** (0.00873) -0.01167** (0.00569) RET 0.01224** (0.00602) 0.01592* (0.00491) FIN -0.03650*** (0.00533) -0.04061*** (0.00324) SERV 0.00307 (0.00603) -0.00027 (0.00464) YEAR2011 0.00173 (0.00506) 0.00222 (0.00352) YEAR2012 -0.00417 (0.00549) -0.00261 (0.00376) YEAR2013 -0.00135 (0.00507) 0.00184 (0.00356) α 0.06344*** (0.00156) 0.17903*** (0.01466) 0.20127*** (0.01517) Observations F-statistic Prob(F-statistic) R-squared 1828 8.75 0.0031 0.0043 1024 27.09 0.000 0.1502 1820 67.83 0.0000 0.2687

Note: The individual coefficient is statistically significant at the *10% level, the **5% or the ***1% significance level. Heteroskedasticity robust standard errors are given in parentheses under the coefficients. KLD is lagged with one year.

Corporate Social Responsibility improves financial results, market reacts skeptical