The Impact of Corporate Social Responsibility on Firms’ Financial Performance in South Africa

If Corporate Social Responsibility (CSR) activities are beyond a firm’s legal obligations and potentially require a sacrifice in short-term profits, why do firms promote CSR? This question motivates this investigation of the impact of CSR on a firm’s Corporate Financial Performance (CFP). This relationship is examined for the period from 2004 to 2013 in South Africa. We assess the short-term impact of CSR announcements on financial returns of firms included in or excluded from the Johannesburg Securities Exchange Socially Responsible Investment Index and determine whether there is a difference in the long-term CFP between these two groups for the entire period. The event study methodology shows that investors were rewarded in 2004 and 2012, when firms entered the index, and were penalized in 2013, when firms exited the index. When using regression analysis, the various industries provide mixed results between CSR and CFP for firms over the long term. Based on these results, we find that CSR activities lead to no significant differences in financial performance.

If Corporate Social Responsibility (CSR) activities are beyond a firm's legal obligations and potentially require a sacrifice in short-term profits, why do firms promote CSR? This question motivates this investigation of the impact of CSR on a firm's Corporate Financial Performance (CFP). This relationship is examined for the period from 2004 to 2013 in South Africa. We assess the short-term impact of CSR announcements on financial returns of firms included in or excluded from the Johannesburg Securities Exchange Socially Responsible Investment Index and determine whether there is a difference in the long-term CFP between these two groups for the entire period. The event study methodology shows that investors were rewarded in 2004 and 2012, when firms entered the index, and were penalized in 2013, when firms exited the index. When using regression analysis, the various industries provide mixed results between CSR and CFP for firms over the long term. Based on these results, we find that CSR activities lead to no significant differences in financial performance. Issue 2 2015 193-214 this statement, purporting that sustainability practices are not fully evolved in African economies. Due to the inability of emerging African economies to relate to the CSR standards of the rest of the world, few studies have been conducted on the sustainable practices of African firms. However, Baskin (2006) found that South Africa has not only a significant Socially Responsible Investment (SRI) Index among emerging economies but also the most developed CSR outlook in Africa and the Middle East as a result of the domestic pressures of CSR and the influence of corporate governance (Baskin, 2006).
Studies have been conducted with the aim of explaining the relationship between CSR and financial performance in both developed economies such as the United States (US) and Britain and developing economies such as Brazil, Nigeria, Taiwan, Turkey and Indonesia. However, there exists a gap in the literature pertaining to South Africa. The topic of CSR and CFP has been well documented for developed economies (Balabanis, Phillips & Lyall, 1998;Tsoutsoura, 2004) as well as emerging economies (Aras, Aybars & Kutlu, 2009;Crisóstomo, Freire & de Vasconcellos, 2011). However, little research has been conducted in an African context. This study aims to identify whether there is a relationship between CSR and CFP in a South African context. There are two main objectives of this study: Objective One is to determine the short-term impact of CSR announcements on the financial returns of firms included in or excluded from the Johannesburg Securities Exchange (JSE) Socially Responsible Investment (SRI) Index. The hypotheses pertaining to this objective can be established according to Curran and Moran (2007), as follows: There is a significant difference in financial performance between constituents and non-constituents of the JSE SRI Index. Freeman (1984) suggests that corporate decisions involve a trade-off between shareholder value and other stakeholder benefits. Therefore, some scholars believe CSR initiatives deplete shareholder value due to the decrease in profits. Others, such as Bagnoli and Watts (2003), believe CSR initiatives increase financial performance. The mixed results found by previous researchers in international markets may be puzzling to South African researchers and investors, as no real consensus has been reached regarding CSR and financial performance.
This study on the effect of CSR announcements will help determine the short-term impact on capital market performance of firms' entrance into and exit from the JSE SRI Index -whether firms are penalized, rewarded or remain neutral to such announcements -in terms of market-based indicators. As a result, one can gain insight into the impact of CSR on shareholder value as well as investor preferences (Becchetti et al., 2012).
Thereafter, the extension of the study provides a longterm view on the performance of firms on the JSE SRI Index relative to firms that are not on the JSE SRI Index but are on the overall index (JSE ALSI). This provides insight into the effects of CSR initiatives on financial performance in terms of accounting-based indicators (Heal, 2008). The results of the study will be beneficial not only to academics focusing on CSR but also to business leaders, managers and investors who are concerned with the impact of CSR on firms' financial performance.
The remainder of the study is structured as follows: Chapter 2 outlines the relevant literature on this particular topic; Chapter 3 describes the methodology, specifically the data collection and the overall research design; Chapter 4 presents a discussion of the obtained results; Chapter 5 provides concluding remarks.

Theoretical background
Over the years, the growing importance of CSR for firms has arisen from the pressure various stakeholders place on these firms to engage in additional CSR Vizja Press&IT www.ce.vizja.pl

195
The Impact of Corporate Social Responsibility on Firms' Financial Performance in South Africa investments (McWilliams & Siegel, 2000). However, managers across various firms do not share the same attitudes towards these concerns about CSR. On one spectrum, there exists managers who believe that in communicating their better social performance with stakeholders such as investors, consumers, suppliers, bankers and employees, the firm's reputation with these stakeholders may improve (Orlitzky, Schmidt & Rynes, 2003). For example, Spicer (1978) found firms' relationships to improve with bankers and investors if they had a higher level of Corporate Social Performance (CSP), thus allowing these firms to access capital and better contractual terms. Bagnolli and Watts (2003) found that firms with high CSP attracted more socially responsible consumers and experienced improved financial performance. Another reason management may address these concerns is because CSR may serve as a strategy to create and maintain a competitive advantage. By adding value to society, firms realize that they can transcend from doing good to doing better in order to survive and compete in the competitive global market (Lin, Yang & Liou, 2009).
In contrast, some managers find that such CSR activities deplete their profits in the long term and try to resist such initiatives because of this perceived tradeoff. Managers who feel that satisfying shareholder interests is a priority and that the primary objective of a business is to make money may render ethical considerations unimportant; therefore, such CSR investments may hurt the financial performance of the business (Friedman, 1970). Ullman (1985) asserts that this negative relationship could result from the inefficient use of resources as firms try to meet the demands of different stakeholders such that the costs incurred from socially responsible actions place these firms at an economic disadvantage.
There is an alternative approach to determining the impact of CSR on a firm's CFP, which is by finding if there is any abnormal financial gain among firms listed on the SRI indices. There has been limited research on the capital market reaction to socially responsible and non-socially responsible activities by corporations. Extant research examines the relationship between positive and negative announcements and changes in stock prices or daily returns. Positive announcements reflect firms that are added to the index, whereas negative announcements reflect firms that have been deleted from the index (Curran & Moran, 2007). Calculating abnormal returns at an announcement date separates the effect of a change in CSR on corporate performance from the reverse causality effect of corporate performance on the CSR of the firm (Becchetti et al., 2012). This is the advantage of investigating the impact of CSR on CFP in financial markets.
From the above, it can be seen that the relationship between CSR and CFP is an important issue for corporate management to consider, and as a result, Preston (1990) suggests that social issues deserve the same attention and rigorous analysis as market factors in determining long-term success.

Measures of CSR
The lack of consensus regarding CSR measurement has been a problem when attempting to determine the impact of CSR on CFP. In some cases, subjective measures such as surveys by faculty members or students (Heinze, 1976;Moskowitz, 1972) or even reputation indices whereby firms are rated and ranked on the basis of various dimensions of social performance (Balabanis et al., 1998;Preston & O'Bannon, 1997) are used. In other cases, researchers such as Babalola (2012) utilize content analysis, which consists of CSR disclosures as stated in annual reports. In countries in which the above methods are not feasible, CSR indices are more commonly used as a dummy variable in terms of regression analysis, whereby a value of 1 is used when firms are included in the Index and a value of 0 is used when they are not (Crisóstomo et al., 2011;McWilliams & Siegel, 2000;Tsoutsoura, 2004).
Each of the above measures is subject to limitations.
Surveys may be subject to bias, as they are based on the researchers' own subjective views. The credibility of the ratings is highly subjective and depends not only on the accuracy of information available to assessors but also on their expertise (Balabanis et al., 1998). Content analysis may be subjective according to the researchers' choice of variables to measure, and firms may not implement their reported actions.

Measures of financial performance
Similar to CSR measurements, there is a lack of consensus concerning the optimal financial performance measurement instrument to employ. Many researchers utilize accounting measures such as Return on Earnings per Share (EPS) (Cochran & Wood, 1984;Waddock & Graves, 1997). Other researchers such as Vance (1975) use market-based measures of financial performance, such as investor returns; others, such as Balabanis et al. (1998)  Accounting-based measures are said to only capture the historical aspects of a firm's financial performance (McGuire, Schneeweis & Hill, 1986). These measures are subject to bias, as they are affected by managerial manipulation as well as by different accounting procedures (Aras et al., 2009). Market-based measures focus on a firm's future performance as opposed to past performance; thus, they are less susceptible to managerial manipulation and different accounting procedures (Aras et al., 2009). To account for these shortfalls, this study utilizes both accounting-and market-based measures.

The JSE SRI Index
The JSE SRI Index serves as an important mechanism for investors to identify good, socially responsible corporations, as well as a benchmark for firms seeking to improve their CSR. The Index was first launched in May 2004 in response to the growing global demand for SRI and was the first CSR Index to be established and owned by an exchange in an emerging economy (Heese, 2005). The main focus of this index is to promote sustainable business practices. Before the inception of the index, South African firms were already actively involved in CSR activity; the origins of SRI in South Africa can be traced back to the early 1990s, when trade unions refused to invest members' contributions in firms that supported the apartheid regime In order for a firm to be considered for the JSE SRI Index, the JSE SRI Index Advisory Committee annu-ally assesses firms based on an array of criteria. The criteria requires firms to be constituents of the FTSE/ JSE All Share Index, and they also need to comply with the ESG standards through a number of different indicators that measure corporate policy, practice and reporting (JSE, 2010).
The criteria of the JSE SRI Index is constantly evolving, ensuring that South African sustainability practices are aligned to those of the developed world; this process goes against Wilson (2007), Heese (2005), and Jamali and Mirshak (2007), who state that sustainability practices are not fully evolved in emerging economies, as they are unable to withstand the CSR standards used in developed economies.

Empirical findings of previous studies
Evidence from a variety of empirical studies concludes that developed and developing economies demonstrate mixed results on the relationship between CSR and CFP.

The Impact of Corporate Social Responsibility on Firms' Financial Performance in South Africa
One must consider the sample size and different measurement approaches and methods when drawing conclusions from the results. Existing techniques involved samples that were too small, in which fewer than 30 firms were considered (Folger & Nutt, 1975), and time periods that were too short (Alexander & Buchholz, 1978). They also relied on content analysis (Abbott & Monsen, 1979) and reputation indices (Folger & Nutt, 1975), which, as mentioned earlier, involves some drawbacks. Lastly, the small sample of CSR firms was compared to control groups of a similar size. These studies failed to find a direct link between CSR and CFP.
Cochran and Wood (1984) improved these existing techniques by increasing the sample size and by using two time periods, large industry-specific control groups, and a key correlate with CSR -asset age -in their model. The average age of corporate assets was found to be highly correlated with a firm's social responsibility ranking. Firms with older assets tend to have lower CSR rankings. Cochran and Wood (1984) reasoned that 'older' firms may be less flexible in adapting to social change, the types of management that older firms attract may be different to those of 'younger' firms, and regulatory constraints may have been less severe in the time period when firms with older assets constructed plants. Overall weak support for a positive relationship between CFP and CSR was found. The exclusion of asset age could explain why no positive relationship was found in earlier studies. Waddock and Graves (1997) found CSP to be positively related to prior CFP and believed that this finding can be supported by the 'Slack Resource Theory. ' This theory asserts that the availability of financial and other (slack) resources to firms as a result of their better financial performance may result in these firms investing in areas related to social domains, such as employee and community relations and the environment (Jensen, 1986). They also found CSP to be positively related to future CFP, providing 'Good Management Theory' as an explanation. This theory asserts that the strong relationship between CSP and good management practice is a result of management paying more attention to CSP domains, which improves relationships with stakeholders and, thus, financial performance (Freeman, 1984). (2000) argued that the study by Waddock and Graves (1997)

Sampling and data collection
Firms that appear to be most sensitive to social responsibility issues are the firms that are JSE SRI Index compliant. To compare the financial performance of these firms to less socially responsible firms, firms that are not JSE SRI Index compliant but listed on the JSE ALSI will be selected to represent less socially responsible firms; this does not imply that these firms do not engage in CSR. Data for the JSE SRI Index is only available from 2004, which was the year in which the index was launched; hence, the period of the study will run from 2004 to 2013 inclusive, a total of 10 years.
CFP will be measured using accounting-and market-based indicators. The relevant accounting data for each firm considered will be obtained from the The long-term study focuses on the Financials, Industrials, Consumer Services and Basic Materials industries. The Consumer Goods, Healthcare, Telecommunications and Technology industries will not be represented, as these industries comprise too few firms. The total sample of firms that will be considered is 42.

Description of overall research design
Regarding objective one, to assess the effect that CSR has on CFP in the short-term, short-term performance will be determined in financial markets by the stock return performance over time, where the event will be the annual announcement of JSE SRI Index constituents. The sample period will be October 2004 to December 2013. A few assumptions must be made before employing the event study methodology (Gladysek & Chipeta, 2012): • Market agents do not anticipate the announcement.
• No confounding effects during the event window are analyzed.
Daily abnormal share returns will be a proxy for shortterm CFP, and the SRI Index will proxy for CSR. The event window will consist of 20 days prior to the announcement date (time T 0 ) and 20 days following the announcement date. Therefore, the event window will consist of 41 days. This event window is sufficient enough to capture any leakage of information prior to T 0 (Becchetti et al., 2012). It also gives investors enough time to react to the news, as some investors may be latecomers to the announcement (Gladysek & Chipeta, 2012). Logarithmic share returns will be calculated, with expected returns proxied by the JSE SRI Index return.
Returns will be computed from 20 trading days prior to the announcement date to 20 trading days after the announcement date (-20, +20). Abnormal returns will then be defined as the difference between the share return and index return. Should the announcement of the SRI constituents have an effect on the market, abnormal returns will change to reflect this new information. For the purposes of this study, CSP will be a proxy for CSR of a particular firm, which is the independent variable. CSP is defined as a dummy variable (0, 1), as no quantifiable measure of the SRI Index exists, nor is information on the ranking of the constituents publically available in South Africa. CSP is 1 if a firm is included in the JSE SRI Index for a given year and 0 if it is not included but listed on the JSE ALSI (McWilliams & Siegel, 2000).
CFP it represents the corporate financial performance of firm i in the long term, which is the dependent variable that will be captured by the following effective accounting metrics: Return on Assets (ROA), Return on Equity (ROE) and Earnings per share (EPS). These profitability ratios have been used in numerous empirical studies of this nature to capture corporate financial performance.
Comparisons across firms will be more accurate when using these accounting ratios, as financial leverage and the influences of risk on CFP will be included The Impact of Corporate Social Responsibility on Firms' Financial Performance in South Africa in the model (Cochran & Wood, 1984). Each financial performance measure will be separately regressed on the independent and additional explanatory variables, such as firm size and risk. These firm-and industryspecific differences must be included, as they have been shown to be related to CFP and CSR in prior empirical research (Ullman 1985). Burke et al. (1986) explains that larger firms attract more attention from external constituents and need to respond more openly to stakeholder demands, which is why they disclose their CSR behavior more often than smaller firms. Perrini et al. (2007) provide evidence that larger firms have a tendency to participate more in environmental management practices than small to medium-sized firms, as large firms are more likely to identify relevant stakeholders and meet their specific requirements through various CSR strategies.
To control for size, the natural logarithm of market capitalization will be used (Aras et al., 2009).
In order to account for the influence of risk, the ratio of long-term debt to total assets will be used. Previous empirical works have found a significant negative relationship between firm performance and the level of debt. Waddock and Graves (1997) reason that risk must be explained, as management's risk tolerance influences its attitude toward activities that have the potential to elicit savings, incur future and present costs, and build or destroy markets. R&D intensity must be included to avoid a misspecified model, as it captures the effect of innovative activity on a firm's performance, considering that the various dimensions of CSR can create process and product innovation (McWilliams & Siegel, 2000). Waddock and Graves (1994) find that the differences in R&D intensity across industries have an impact on the performance of these industries. Separating the firms into industries as mentioned above will take these differences in R&D intensity into account. It will also account for differences due to industry-level factors, such as economies of scale and competitive intensity, which must be explained (McWilliams & Siegel, 2000). Industries will be determined in accordance with the Industry Classification Benchmark (ICB), which is the global standard for industry sector analysis that the JSE uses. This study will divide firms into different industries and thereafter run regressions for the firms within each industry. The model will be estimated using E-views. Table 1

201
The Impact of Corporate Social Responsibility on Firms' Financial Performance in South Africa   AARs, which were greater than 2.5% and less than -2.5% (MacKinlay, 1997). The annual announcement of JSE SRI constituents will be perceived as good news should the daily AARs exceed 2.5%. When AARs are less than -2.5%, the announcement will be perceived as bad news. Therefore, from the figures above, it can be seen that 2008 and 2013 carried both good and bad news.

Exit announcements of the JSE SRI Index
As shown in Table 3  For firms that were excluded from the index, Table 4 shows that the AARs for seven out of eight years have insignificant p-values, implying that these AARs are equal to zero. The year 2013 is significant at the 90%   The Impact of Corporate Social Responsibility on Firms' Financial Performance in South Africa confidence interval, suggesting that investors perceive the exclusion of firms from the index as negative information. The direction of the AARs for five of the eight years was negative (2006, 2007, 2008, 2012 and 2013), and although four years showed insignificant AARs, this result nevertheless suggests that the announcements of firms excluded from the JSE SRI Index may be associated with negative share price changes for those firms.

Objective two
Three financial ratios were used to analyze the financial performance of firms in relation to their social responsibility measures.
The purpose of the regression analysis is to test whether CSR is significantly associated with superior or inferior financial performance within industry groups. The finding of a significant positive relationship may suggest that CSR activities by firms should be increased to produce better implications for firm performance.
A leading CSR firm that derives the following benefits highlights why these activities may be characterized by higher financial performance: relationships with stakeholders improve, management of brand and reputation may be more effective, attraction and retention of highcaliber employees may be promoted, and the organization may receive recognition for displaying a high degree of ethical standards (Freeman, 1984;Orlitzky et al., 2003;Waddock & Graves, 1997).
A negative relationship may be explained by the additional costs of undertaking CSR activities that do not enhance or contribute to shareholder value. As a firm becomes more socially responsible, the more difficult it becomes for the firm to increase its economic profits, as it cannot readily engage in projects without assessing their implications for ESG frontiers.
Two possible reasons for why there may be weak support (a neutral effect) for a link between CSP and CFP include the reliability of accounting figures, as some firms manipulate financial statements, and the notion that non-constituents cannot systematically experience weaker performance, as Lee, Faff and Smith (2009) highlight that these firms stand a chance of being liquidated or acquired by market leaders or adjusting their CSR practices in order to maximize their economic profit. Lee et al., (2009)

Industrials Industries
The results for the Industrials industry show that size is highly positively correlated to EPS and CSR.

Conclusion
The results of this study add to the body of literature on CSR and CFP. Very few studies have considered both entry and exit announcements using the JSE SRI Index. This study employs an event study methodology for the short-term and regression analysis for the longterm. The analysis in this study examines short-term shareholder wealth returns around announcement dates and makes a long-term comparison of the op-  Overall, it can de deduced that as the years progressed, the announcement event was considered as good and bad news entering the market.
The analysis over the long term also provides mixed results for CSR and CFP within the different industries. Within a particular industry, it is possible to find significant positive, negative and neutral differences in CFP between constituents and non-constituents of the JSE SRI Index; thus, the relationship between CSP and CFP may be sensitive to the type of CFP measure used. Therefore, it is uncertain whether JSE SRI Index compliance leads to significant differences in financial performance.
Future research could improve the long-term analysis by increasing the sample size of firms under consideration by looking at year-to-year changes as opposed to analyzing firms listed on the indices for the entire ten-year period. Another avenue to consider would be to include asset age in the regression analysis due to its importance, as previously mentioned. If the JSE were more explicit about the relative ranking of the firms on the JSE SRI Index, investors would be provided with a better understanding of CSR and why firms are included in and excluded from the index. This information would allow potential investors to include these considerations in their own investment portfolios.