Do Investors Value a Firm s Commitment to Social Activities?

Transcription

1 J Bus Ethics (2013) 114: DOI /s Do Investors Value a Firm s Commitment to Social Activities? Waymond Rodgers Hiu Lam Choy Andrés Guiral Received: 1 March 2011 / Accepted: 1 October 2012 / Published online: 16 April 2013 Ó Springer Science+Business Media Dordrecht 2013 Abstract Previous empirical research has found mixed results for the impact of corporate social responsibility (CSR) investments on corporate financial performance (CFP). This paper contributes to the literature by exploring in a two stage investor decision-making model the relationship between a firm s innovation effort, CSR, and financial performance. We simultaneously examine the impact of CSR on both accounting-based (financial health) and market-based (Tobin s Q) financial performance measures. From a sample of top corporate citizens, we find that: (1) a firm s social responsibility commitment (CSR) contributes to its financial performance; (2) after controlling for investment in innovation activities, CSR continues to have a positive impact on a firm s financial performance; (3) the customer dimension of CSR has a positive effect on both CFP measures, whereas the employee dimension indicates a significant impact only on financial heath; and (4) the community relation dimension of CSR only affects the market-based CFP measure of firms with high innovation intensity. W. Rodgers (&) University of Hull, Yorkshire, UK w.rodgers@hull.ac.uk W. Rodgers University of Texas, El Paso, TX, USA H. L. Choy Drexel University, Philadelphia, PA, USA A. Guiral Yonsei University School of Business, 50 Yonsei-ro, Seodaemun-gu, Seoul , South Korea Keywords Corporate social responsibility Corporate financial performance Innovation Social dimensions Investor decision making Introduction In recent years, investors, creditors, and financial analysts have started to emphasize the importance of a company s responsibility to other stakeholders rather than focus solely on shareholders. One of the major aims of the guiding conceptual accounting framework depicted in Statement of Financial Accounting Concepts No. 2 (SFAC #2) (FASB 1980) is to help improve users judgments and decisions when confronted with a variety of information sources. Therefore, it becomes increasingly popular to publish corporate social responsibility (CSR) information as part of the president s report to stockholders. This information helps report users to make more informed decisions as highlighted by SFAC #2. As of date, research is quite mixed on the importance of CSR measures and its impact on corporate financial performance (CFP) (McWilliams and Siegel 2000, 2001; Barnett and Salomon 2006; Prior et al. 2008; Hull and Rothenberg 2008; Surroca et al. 2010). The link between CSR and CFP has been intensely examined by the strategic/management/organization literature. The majority of researchers have used the stakeholder theory (Freeman 1984; Sen et al. 2006), and some other related approaches as the resource-based view (Barney 1991) and transaction cost economics (Jones 1995) to explain a positive link from CSR to CFP (Margolis and Walsh 2001). This paper does not try to differentiate which of these theories drives the relationship between CSR and CFP. Rather, we rely on these theories to construct our

2 608 W. Rodgers et al. hypotheses on the relation among CSR, innovation, and CFP. Several reviews (McWilliams et al. 2006; Margolis and Walsh 2001, 2003; Roman et al. 1999; Griffin and Mahon 1997; Pava and Krausz 1996; Wood and Jones 1995; Ullmann 1985), meta-analysis (Orlitzky et al. 2003) and recent special issues in top management and organization journals 1 suggest that CSR efforts improve a firm s CFP. These theories and empirical findings have significantly contributed to our knowledge of why a positive sign of the CSR CFP relation may be expected. At the same time, they have led to various interpretations. Barnett (2007, p. 801) states we now understand the effects of isolated pieces of the overall puzzle, ceteris paribus, but the dots remain unconnected through any theoretical framework. Orlitzky et al. (2003) point out that there is a large amount of unexplained variance across studies. These discrepancies among studies suggest the potential presence of mediator/ moderator variables such as a firm s investments in intangible assets (Barnett 2007; McWilliams and Siegel 2000, 2001; McWilliams et al. 2006; Hull and Rothenberg 2008; Surroca et al. 2010; Guiral 2012; Blanco et al. 2012). 2 The aim of this paper is to extend previous research by exploring the relationship between a firm s commitment to innovative activities, its CSR reputation and financial performance, which SFAC #2 suggests as useful for informed decision makers. Our study of the top 100 corporate citizens complied by KLD Research and Analytics ( provides evidence that a firm s CSR effort has a positive effect on both the accounting-based (i.e., financial health) and market-based (i.e., Tobin s Q) performance measures. 3 CSR has both a direct and an indirect (via the intermediate stage accounting-based performance) effect on Tobin s Q. A firm s commitment to innovate its products or production process also contributes positively to its performance. However, we do not observe any significant interaction effect between 1 See the special issue on Corporate Social Responsibility: Strategic Implications in Journal of Management Studies (volume 43, issue 1, January 2006) and the special topic forum on Corporations as Social Change Agents in Academy of Management Review (volume 32, issue 3, July 2007). Also, see previous special research forum on Stakeholders, Social Responsibility, and Performance in Academy of Management Journal (volume 42, issue 5, October 1999). 2 Moreover, McWilliams et al. (2006), McWilliams and Siegel (2000), and Barnett (2007) argue that the lack of consistency in previous findings may be a consequence of poor definition of the financial performance and CSR constructs, imprecision in research design, model specification biases and deficient analysis of the results. 3 Tobin s Q depicts the market s valuation of a company compared to its assets-in-place (Smith and Watts 1992). CSR and a firm s commitment to innovate on a firm s performance. 4 We contribute to this literature in five ways. First, our research extends previous works by examining how CSR and CFP are linked in a two-stage investor decision-making model. We argue that conflicting results on the relationship between CSR and CFP found in previous studies may have been due to different stages of influence of CSR on CFP. By adopting a process approach, we are able to examine if CSR affects both the intermediary outcome (i.e., financial health) and the final stage (i.e., Tobin s Q)of CFP. Second, we improve the way in which CFP have been traditionally captured. Specifically, we integrate the accounting and market-based performance measures; whereas previous research has merely focused on a single surrogate. Also, we measure a firm s accounting performance from multiple dimensions: profitability, liquidity, and leverage. This provides a more comprehensive measure of a firm s performance. Third, we respond to Barnett (2007), McWilliams and Siegel (2000), and McWilliams et al. (2006) s call for more research on the role played by intangibles (as captured by a firm s investment in innovation activities). Specifically, we examine its relation with a firm s CSR efforts and how they interact to affect a firm s value and performance. Fourth, in addition to the overall evaluation of a firm s CSR performance, we examine the various dimensions employee, customer, and community relations of a firm s CSR performance. We observe that these various dimensions of CSR have different impact on a firm s performance. The customer dimension of CSR has a significant positive effect on both financial health and market value while the effect of the community relation dimension is limited to the market value of firms with significant investment in R&D. That is, while the customer dimension has a broad impact on a firm s financial performance, the effect of CSR investment in community relation is limited to a specific group of firms. This variation in the impact of the different dimensions of CSR on firm performance suggests that when a firm has constrained resources to invest in CSR activities, they should try to identify the ones that can help them get the most bucks out of it. Fifth, we document that the relation between CSR and CFP is dynamic in that a high CSR rating leads to a better CFP, which in turn allows an increase in CSR investment and further enhances investors perception of the firm s social responsibility commitment. In other words, we contribute to the CSR and accounting literature (SFAC #2) by providing evidence that the interaction between CSR 4 The only exception is the community dimension of CSR. The interaction between the community dimension and the investment in innovation has a positive effect on Tobin s Q.

3 Do Investors Value a Firm s Commitment to Social Activities? 609 and CFP is bidirectional with useful information, and not unidirectional. Finally, matching a conceptual model with a covariance structural model could provide new insights regarding the importance of the interactions of CSR and CFP. The next section presents an overview of the theories and empirical research of the CSR CFP paradigm. In the third section, we propose a simultaneous approach through an investor s decision-making model. The development of hypotheses, empirical constructs, and results are reported in the fourth, fifth, and sixth sections, respectively. Additional sensitivity analyses are provided in Additional Analyses section. The section Discussion discusses the implications of our findings and the section Conclusion concludes the paper. The Contribution of CSR Investments to a Firm s Financial Performance Much has been discussed in terms of whether firms should orient themselves toward a shareholder or stakeholder perspective (Sundaram and Inkpen 2004). In an attempt to solve this strategic paradigm, a significant number of research papers have focused on the examination of the CSR CFP link and the presence of potential moderators affecting that relationship (Orlitzky et al. 2003; Margolis and Walsh 2001). The stakeholder theory (Freeman 1984) has been the most important approach in explaining how CSR investment leads to a higher CFP; that is, how a firm s commitment to social activities contributes to its financial wealth. This theory postulates that it is insufficient for managers to focus exclusively on the perceived needs of shareholders (McWilliams et al. 2006). In this regard, firms should cater to the demands of important stakeholders other than their shareholders alone (Ruf et al. 2001). When applied to a firm s commitment to social activities, stakeholder theory supports a firm s investment in CSR in order to enhance its relation with its customers, employees, and shareholders. For instance, Greening and Turban (2000) and Turban and Greening (1997) suggest that people can react to a firm s CSR investment by seeking employment with the firm, instead of just purchasing products from it. Thus, the impact of CSR on a firm s financial performance or its value can be examined from multiple facets. Other researchers have suggested the arguments of the transaction cost economics and resource-based view to illustrate why a firm may pursue the satisfaction of stakeholders demands (Ruf et al. 2001; McWilliams et al. 2006; Jones 1995). From the transaction cost economics it could be argued that firms would try to satisfy stakeholders needs in order to minimize potential transaction costs (Williamson 1985). While shareholders and debt holders have explicit claims on the firm, other stakeholders (e.g., customers, government, and the community) have implicit claims on the firm. 5 When a firm fails to act socially responsible, other stakeholders will have doubt about whether the firm will honor their implicit claims. These stakeholders likely transfer the low-cost implicit contracts into costly explicit claims. Thus, transaction cost economics implicates that firms with good CSR perceptions have low-cost implicit claims whereas those with poor CSR perceptions more likely face high-cost explicit claims (Cornell and Shapiro 1987; Peloza 2006). 6 The resource-based view also suggests a positive impact of CSR on CFP. From this approach firms interpret meeting stakeholder s demands as a strategic investment (Ruf et al. 2001; Russo and Fouts 1997). By investing in such a strategy, organizations develop assets that are valuable, rare, and nonsubstitutable, such as leadership and positive social reputation. These assets in turn lead firms to a competitive advantage and potentially a higher return (Barney 1991; Luo and Bhattacharya 2006). Thus, resource-based view advocates that CSR activities help managers develop better skills and firms develop intangible assets (e.g., brand name) which, in turn, will contribute to better financial performance (Russo and Fouts 1997; Wernerfelt 1984). Two recent reviews developed by Margolis and Walsh (2003) and Orlitzky et al. (2003) adequately summarize the current stage of the CSR CFP empirical research. On one hand, Margolis and Walsh (2003) reported that in their review of the 109 published articles using CSR as an explanatory variable, only 54 studies documented a positive impact of CSR on financial performance while 7 studies showed a negative relation. They also provided a review of 22 studies that employed CFP as a predictor of CSR, in which they found 16 studies pointed to a positive relationship. Margolis and Walsh (2003) found only 4 studies investigating the relationship from both directions. On the other hand, Orlitzky et al. (2003) s meta-analysis of 52 empirical studies shows a general positive correlation between CSR and CFP and suggested that such a relationship tends to be bidirectional. Thus, results of studies on the impact of CSR on a firm s financial performance are at best mixed. Barnett (2007), McWilliams et al. (2006), and McWilliams and Siegel (2001) point out a potential shortcoming in the previous studies: the failure to control for other firm characteristics 5 For example, the firm has an implicit contract with its customers on the quality of products and services. Customers have an implicit claim on the firm when the product does not meet quality standard (e.g., product recalls for safety concern). 6 Peloza (2006) also suggests that corporate social activities, in terms of firm reputation, may play an important role in mitigating the potential consequences of future damaging events.

4 610 W. Rodgers et al. such as investment in R&D. In this study, we address this issue by including a firm s investment in R&D (a proxy for the firm s commitment to innovation). Further, we examine if the effect of CSR depends on this investment in R&D. That is, we investigate if the two have any interaction effect. Although theories and models surrounding the CSR CFP link are abundant, empirical research is still in an early stage (McWilliams and Siegel 2000; Peloza 2009). Harrison and Freeman (1999, pp ) further recommend that researchers should make a significant effort to try to find a way to integrate the economic and the social. Thus, rather than an isolated analysis of the impact of CSR on CFP, we propose an investor s decision-making model by concurrently examining the role played by financial measures, innovation activities, and CSR. By integrating the economic and social investments of a firm in a single model, it allows us to better understand the CSR CFP relationship. A CSR Investor Decision-Making Model A Throughput Model (Rodgers 1997) is presented in this research paper that captures different pathways and stages that can influence a firm s decision. The model proposed here also provides a broad conceptual framework for sequencing our arguments. It incorporates the constructs of perception, available information, financial health status (analysis of information/perception), and decision choice as it applies to firms (Foss and Rodgers 2011). Analysts at KLD Research and Analytics form their perception regarding a firm s performance in different aspects of CSR based on the answers to a set of questions they consider as critical measures or determinants of a firm s commitment to CSR. The experts developed questions related to the different categories of CSR performance measures based on their training, experience, and education. Investors then rely on this expertise perception to form their own perception regarding a firm s performance in CSR. We depict this insight as perception in our model. Perception on CSR efforts together with other information are then used for the second-stage evaluation, an assessment of a firm s financial health in the current case. Information includes the set of financial information reported by firms. We include profitability, liquidity, and leverage as our information (I). In this model, information (I) together with investors perception of the firm s commitment to innovation and CSR measured by customer, employee and community relations (P) affects the financial health judgment (J), i.e., an accounting-based measure of CFP. Similar to the case of CSR efforts, we encode a firm s investment in R&D as perception. Investors form their perception regarding the value of a firm s investment in innovation activities based on the limited information available on the financial statements. Financial health judgment involves a detailed analysis of all the available information and CSR. Decision choice then follows. This final stage represents a culmination of information, perception, and judgment (see Fig. 1). Hypotheses Development Profitability and other financial measures provide useful information on a firm s short-term performance. However, their ties to the long-term well-being of a firm can be limited. In our model, we propose that in addition to accountingbased measure of CFP, investors perception of a firm s CSR efforts and its investment in innovation play a significant role in predicting a firm s long-term performance; and hence, in determining its market value (see Fig. 1). In this section, we first describe the various dimensions of financial information, innovation commitment, and CSR efforts. In addition, we develop our hypotheses regarding their potential impact on both the firm s financial health and its market value. Most of the previous empirical research has focused on the examination of profitability or other isolated measures as individual surrogates of accounting-based CFP. We propose the use of financial health construct (J) as a summary measure of a firm s accounting-based CFP. This construct simultaneously captures a firm s profitability, liquidity, and leverage information (Johnstone and Bedard 2004). It is crucial to capture the various dimensions of a firm s financial health because a profitable firm can suffer from a shortfall of cash flows and not be able to survive if it cannot meet its debt obligations. This going concern issue likely has an adverse effect on the firm s value. In addition, a firm that is highly leveraged can face difficulty in raising additional external funding. All these can constraint the firm s growth and wellbeing in the future. As such, the financial health construct provides a more comprehensive measure of the firm s performance and financial condition than a single profitability measure. While financial information in accounting reports indicates a firm s current CFP status, it does not necessarily reflect the firm s ability to increase or maintain its CFP in the future. Certain studies on intangible assets (Lichtenberg and Siegel 1991; Amir and Lev 1996; Lev and Sarowin 1999; Demers and Lev 2001; Trueman et al. 2000; Xu et al. 2007) suggest that traditional financial reports do not reflect a firm s value in today s high-tech environment. In this information technology environment, a firm s knowledge-based assets become a significant determinant of the firm s value. For firms in such high-tech industries as internet, biotechnology, and computer, one important

5 Do Investors Value a Firm s Commitment to Social Activities? 611 R&D/sales KLD rating score PERCEPTIONS Innovation Corporate Social Responsibility (CSR) KLD measures customer employee community Net income/sales Sales/assets Profitability Quick ratio Cash ratio Liquidity Financial Health Market Value Debt/assets Debt/equity Leverage INFORMATION JUDGMENT Zscore DECISION Tobin s Q Fig. 1 Investor decision-making model: latent variables and indicators determinant of their potential growth is the continued innovativeness of their products. As such, a significant portion of high-tech firms investment is in R&D. These knowledge-based assets constitute a significant part of a firm s economic resources. For instance, Ernst & Young L.L.P. estimates that publicly listed biotechnology firms spent a total of $14.5 billion on R&D in 2003, $15.8 billion in 2004, and $16 billion in 2005 (Ernst and Young 2009). Pharmaceutical Research and Manufacturers of America, which is a trade group that includes major pharmaceutical companies and certain large biotech firms, reported that its members spent an average of 18.8 % of their sales revenues on R&D in This figure has stayed at around 18 % since 1999 (PHRMA 2007). For other biotechnology firms (i.e., excluding major pharmaceutical firms), S&P estimated that about 38 and 40 % of their revenues were spent on R&D in 2003 and 2004, respectively. Firms make such significant amount of investment in R&D in order to continuously improve their existing products or production procedures and explore innovative new products. This continuous product improvement helps to attract new customers/retain current customers, and improve production efficiency, which in turn contributes to the firm s profit and growth. As such, we expect the investment in R&D has a positive impact on a firm s value. We incorporate this potential value-creating investment in our model. Because the investment in R&D data are not available to investors until after the year-end, we expect investors form their perception regarding the firm s commitment to innovation based on the firm s prior year investment in R&D. In addition to its innovation commitment, the social perception of a firm can also be an invaluable asset to the firm. We evaluate a firm s CSR efforts as perceived by its three stakeholder groups (i.e., employee, customer, and community relations) on the firm value. To this end, our model conceptualizes that, in addition to the financial viability of a firm, investors incorporate their perception of the firm s commitment to meet CSR in their valuation decisions. We focus only on the three stakeholder groups since data on other stakeholder groups are not available for the early years of our sample period. In order to provide a consistent measure over time, we include only these three stakeholder groups in our analyses. Employee Perception With the significant growth in high-tech firms, the competition for talents in the labor market has been more intense than ever (Collins and Smith 2006; Ballou et al. 2003). Firms use various long-term compensation schemes, such as stock options and restricted stock grants with several years of vesting period, to retain their employees. Still other firms

6 612 W. Rodgers et al. tried more innovative ways to attract employees. A positive firm image created by CSR efforts can attract potential employees because individuals tend to identify with firms that are socially responsible and are more likely to seek employment from these firms (Sen et al. 2006; Greening and Turban 2000; Turban and Greening 1997). Hein (1981) also reports that a firm s current employees view corporate social reporting as an important mean of communication between the firm and employees. An increase in current employees goodwill can improve productivity and financial outcomes (Davis 1973; McGuire et al. 1988; Waddock and Graves 1997). Unfortunately, this important human capital asset of a firm is not recorded anywhere on a firm s balance sheet. However, with its significant impact on a firm s long-term well-being, we anticipate investors account for this employee satisfaction factor and the positive image of a firm in their investment decision. Hence, this human capital asset is likely to be factored into investors pricing decision. Customer Perception In addition to well-adjusted employees, it is equally important that a firm s customers are satisfied with its products and/or services (Luo and Bhattacharya 2006; Maiga and Jacobs 2005). A well-maintained customer relation helps to retain old customers for repetitive businesses, build up customer loyalty, and attract new businesses by the word of mouth. Since the costs of keeping existing customers are likely to be lower than those of acquiring new customers (Voss and Gruber 2005), a wellmaintained customer relation can contribute to a firm s financial performance by reducing marketing and other customer support costs. Retention of existing customers is particularly important for a company operating in a mature industry (Reichheld 2003). While these customer-relation efforts have both shortand long-term benefits, only the short-term benefits are recognized as sales on financial statements. On the other hand, all such costs incurred in improving the customer relation and promoting the products in the current period are recorded as expenses on the current income statement. Without accounting for the customer relation or brand name on the financial reports, these financial statements again understate the assets of a firm. If the market realizes the value of brand name and/or customer relation to a firm, investors will incorporate this information in their pricing. Community Relation Besides improving relation with current employees and customers, it also helps a company to cultivate its relation with potential employees and customers. One way to achieve this is to have a good community relation. Maintaining a good community relation promotes the image of a company. A good company image contributes to a good brand name. 7 By actively participating in community programs, either through charitable donation, sponsorship, and/or setting up policies to encourage employees to volunteer, firms can promote their brand name. We expect this long-term relationship with the community improves the firm s image and helps to attract not only customers, but also potential employees and investors, which leads to an increase in its market value. This positive image can also provide insurance on shareholder value. These arguments lead us to our first hypothesis: Hypothesis 1 Employee, customer and community satisfaction with firm CSR will be positively related to the market valuation of the firm. Investment in CSR and R&D are expected to be positively correlated (McWilliams et al. 2006; McWilliams and Siegel 2000; Barnett 2007). McWilliams and Siegel (2000) advocate that many firms that actively engage in social activities are also pursuing a differentiation strategy, involving complementary strategic investments in intangible resources. One such intangible is the investment in R&D to continuously improve a firm s products. As a result, it would be very difficult to isolate the impact of CSR on CFP without simultaneously controlling for the investment in innovation activities (proxied by R&D). Also, these authors argue that R&D investment and CSR are positively correlated, since many aspects of CSR lead to either a product innovation, a process innovation, or both (McWilliams and Siegel 2000). Thus, it is important for us to control for a firm s investment in R&D in our analysis of the impact of CSR on both accounting (financial health) and market-based (investor decision) CFP measures. McWilliams and Siegel (2000) suggested that excluding innovation intensity from the CSR CFP relation analysis might bias the empirical results. Indeed, McWilliams and Siegel (2000) found that when innovation was taken into account, CSR had a neutral impact on CFP over the period More recently, Hull and Rothenberg (2008) also argue that when innovation is essential to immediate survival, CSR may not have much impact on CFP. However, when the firm is not forced to invest in innovation, CSR investments may have some effect on CFP. In this regard, these authors found that for a sample of firms in the 7 For example, Ford Motor was selected Latina Style magazine s 2004 Fifty Best Companies based on, among other criteria, its mentoring program and its relationship with the Hispanic community. This publicity can help Ford to attract better prospective employees (Greening and Turban 2000; Turban and Greening 1997), increase customer-company identification (Sen and Bhattacharya 2001) and affect customers product attitude (Berens et al. 2005).

7 Do Investors Value a Firm s Commitment to Social Activities? period, CSR investment most strongly affected CFP in low-innovation firms. 8 Surroca et al. (2010) also suggest that there is no direct relationship between corporate responsibility and financial performance merely an indirect relationship that relies on the mediating effect of a firm s intangible resources. Therefore, in addition to the positive correlation between R&D investment and CSR, we expect that CSR and R&D investments interact to affect a firm s financial health and market value. Hypothesis 2 As a firm s innovation intensity increases, the effect of stakeholder satisfaction with its CSR on firm performance will become greater. Empirical Constructs CSR Measure We use the employee and customer satisfaction, the community relation indices, and the overall social responsibility score provided on Business-ethics.com ( ) to capture a firm s investment in CSR. Each year, the website generates a list of top 100 corporate citizens based on the data provided by KLD Research and Analytics. Our sample period starts in 2000 because this is the first year Business-ethics.com started generating the list and ends in 2006, the last year the list is available when we first conducted our analyses. We have a total of 497 observations for the sample period. This top 100 corporate citizens list is constructed based on a firm s scores in the welfare measures of four stakeholder groups: the customer, employee, community, and stockholders. Over the years, the criteria have gradually expanded to include the welfare of seven stakeholder groups. However, in order to provide a consistent comparison over years, we include only the initial four stakeholder groups in our analyses. For each stakeholder group, KLD Research and Analytics identified the strengths and concerns of a firm in that area. The number of concerns was then subtracted from the number of strengths to arrive at a net score. This net score was then standardized by the mean net score of each stakeholder category. That is, KLD used the number of standard deviations a firm s scores was from the mean of each stakeholder category as a measure of the firm s points received in that category. An overall score for a firm is then computed using the average of points received in all stakeholder categories. This constitutes the overall social responsibility standing of a firm (CSR). 8 Both McWilliams and Siegel (2000) and Hull and Rothenberg (2008) based their conclusion on a single accounting-based measure: profitability (proxied by return on assets ratio). For the customer category, KLD examines whether the firm has a quality management program and its quality, any quality awards the firm has won, customer satisfaction measure, any customer lawsuit pending, etc. In the employee relation category, KLD considers, among other criteria, employees wages relative to the industry, benefits, employee empowerment, family-friendly policies, and other policies the firm has in place to accommodate employees needs. The factors considered in the community relation category includes whether the company has any foundation, community service projects, employee volunteer programs, etc. We took the scores a firm receives in the customer, employee, and community categories from the website and rank them relative to other firms that made the top 100 list in the year ( KLD measures in Fig. 1). 9 We use the relative ranking in each category as a proxy for the quality of its customer, employee, and community relations. In addition, we use the overall social responsibility score a firm receives as a proxy for the general perception investors have regarding the firm or the image a firm projects to the general public ( CSR in Fig. 1). Financial Health and Firm Value We use a firm s financial health status as a proxy for the firm s accounting-based CFP. This serves as a benchmark against which we can measure the contribution of the perception (both social and innovation) to the overall market value of a firm relative to the financial viability. We use -1 times the Zmijewski score (Zmijewski 1984) asa proxy (i.e., -1 9 Zmijewski score) for financial health. 10 This financial health score measures a firm s financial viability. The higher the score, the higher the probability a firm will stay financially healthy. The Zmijewski score is constructed based on a firm s profitability, liquidity, and leverage ratios as follows: ZFC ¼ 4:336 4:513 ðroaþþ5:679 ðfinlþ þ 0:004 ðliqþ where ROA is the return on assets, FINL is the financial leverage, and LIQ is the liquidity measure. The financial health measure is computed as -ZFC. We consider three sets of financial measures that are likely to be significant determinants of a firm s financial health: 9 Indicators for our latent variables are shown in Fig Numerous accounting studies have used Zmijewski score for the recognition of financial distressed firms (e.g., Pava and Krausz, 1996; Ruiz-Barbadillo et al., 2004; Johnstone and Bedard, 2004; Carcello and Nagy, 2004). Since we are interested in the effect of CSR on a firm s performance, we use the negative of Zmijewski score as a proxy for the firm s financial health. We have also repeated our analyses using the inverse of Zmijewski score as a proxy and the tenor of the results does not change.

8 614 W. Rodgers et al. profitability, liquidity, and leverage. We use the profit margin (net income to sales) and assets turnover (sales to total assets) ratios to measure a firm s profitability. A firm s liquidity is also measured by two indicators: the quick ratio (the ratio of sum of cash, marketable securities, and receivables to current liabilities) and the ratio of cash to current liabilities. Leverage is measured by the ratio of total debt to total assets and the ratio of total debt to equity. Finally, the market s valuation of a firm is measured by Tobin s Q, computed as the sum of the market value of equity and the book value of debt, scaled by the book value of assets. It has been widely used in prior research as a measure of a firm s performance (Smith and Watts 1992; Kaplan and Zingales 1997; Pava and Krausz 1996; Garcia-Castro et al. 2010). Tobin s Q measures the market s valuation of a firm relative to its assets-in-place. It captures the market s expectation about the firm s potential growth and hence its future prospect. If investors perceive CSR (commitment to innovation) as a value-enhancing investment and can help a firm to grow, this positive perception should be reflected in its market value and hence improves its Tobin s Q. Results We first use partial least squares (PLS) to analyze our research models in order to understand the overall relation among our constructs. As our model involves multiple latent constructs with multiple indicators (see Fig. 1), PLS has proven to be particularly useful in depicting the overall relationship in this case. Also, PLS imposes minimal restrictions on measurement scales, sample size, and residual distributions. Table 1 provides descriptive statistics on the indicators and Table 2 presents their correlations. The average KLD score of our sample is with a standard deviation of The three dimensions of CSR employee, customer, and community are positively correlated with one another. Also, their correlation with the overall KLD score ranges from to Measurement Validation The measurement model evaluates the relationship between indicators and latent constructs by assessing the reliability and validity of the scales measures. Each measure s reliability is assessed by examining the loading of the indicators on the corresponding construct. All measures have a loading level above In addition, measurement residuals are small. All loadings have the expected signs (i.e., non-negative) and are significant at the level (one-tailed). Further, all constructs present a composite reliability (see, Fornell and Larcker 1981) above 0.70, the benchmark level suggested by Nunnally (1978). Table 1 Descriptive statistics Indicator Mean Standard deviation Also, convergent and discriminant validity can be evaluated within the PLS model. According to Chin (1998) and Chin et al. (2003), convergent and discriminant validity is inferred when the square root of each construct s Average variance extracted (AVE) is larger than its correlations with other constructs (the average variance shared between the construct and its indicators is larger than the variance shared between the construct and other constructs). Our models show satisfactory discriminant validity since the square root of the AVE of each construct is much larger than any correlation between the construct and any other constructs. The CFA procedure in PLS is also performed. Convergent validity of a construct is measured by the ratio of the variance of its indicators captured by the construct to the total amount of variance ( average variance extracted o vc ). The total amount of variance includes the variance due to measurement error. As a rule of thumb, a ratio of less than 0.50 implies the convergent validity assumption is violated because more variance is explained by the error than the construct. In all our models, average variance extracted (o vc ) ranges between 0.65 and 1.00, indicating satisfactory convergent validity for the constructs. Another meaningful indicator of the fit of a model with respect to its measurement is the overall communality coefficient (in our models from 0.49 to 0.847). This coefficient exceeds Falk s (1987) recommendation that it should be greater than Thus, it can be concluded that the constructs are measured with sufficient precision, that is, the model is both reliable and valid. CSR Leading to Higher CFP Maximum Minimum Net income to sales Sales to assets Quick ratio Cash ratio Total debt on assets Total debt on equity R&D to sales Community Employee Customer KLD score Z score Tobin s Q The PLS path coefficients for our Models 1, 2, and 3 are shown in Table 3. Overall, our results suggest that investors integrate both financial and non-financial measures in their valuation of

9 Do Investors Value a Firm s Commitment to Social Activities? 615 Table 2 Correlation matrix Indicator (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) Net income to sales (1) 1 Sales to assets (2) Quick ratio (3).100* -.235** 1 Cash ratio (4).105* -.235**.996** 1 Total debt on assets (5) -.092* ** -.395** 1 Total debt on equity (6) ** -.124** -.104*.579** 1 R&D to sales (7) **.436**.406** -.336** -.148** 1 Community (8).118** ** -.190**.167**.136** -.178** 1 Employee (9) *.247** 1 Customer (10) ** **.531** 1 KLD score (11) * -.113*.119**.108* **.710**.744** 1 Z-Score (12).295**.240**.255**.239** -.804** -.544**.223** Tobin s Q (13).164**.142**.147**.141** -.173** -.147**.199** **.169**.421** 1 * Significant at p \.05; ** significant at p \.01 afirm.inmodel1(seetable 3), we show the effect of CSR on financial performance without controlling for investments in R&D. We observe that a firm s effort in satisfying the needs of its various stakeholders employees, customers, and the community has a significant influence on the market s perception of the firm s commitment to meet its social responsibility (b 5 = 0.920, p \ 0.01; R 2 = 0.847), which in turn, has a significantly positive effect on the firm s financial health (b 6 = 0.044, p \ 0.05; R 2 = 0.763) and market valuation (b 7 = 0.179, p \ 0.01; R 2 = 0.209). These results support our Hypothesis 1. Consistent with prior research, the financial health latent variable has a significantly positive effect on the market value of companies (b 4 = 0.425, p \ 0.01). In other words, companies with sound financial health are those with higher market value. Further, companies with better financial health are those with higher profitability (b 1 = 0.309, p \ 0.01), higher liquidity (b 2 = 0.059, p \ 0.01) and lower leverage (b 3 =-0.819, p \ 0.01). In Model 2 (see Fig. 2; Table3), we incorporate both investors perception of a firm s commitment to innovation and CSR in our analysis. We observe that both the CSR and investment in R&D have significant positive impact on a firm s financial health (b 6 = 0.041, p \ 0.10; b 8 = 0.139, p \ 0.01; R 2 = 0.775) and its market valuation (b 7 = 0.187, p \ 0.01; b 9 = 0.127, p \ 0.01; R 2 = 0.224), respectively. Similar to Model 1, firms with better financial health are those with higher profitability (b 1 = 0.351, p \ 0.01), higher liquidity (b 2 = 0.119, p \ 0.01) and lower leverage (b 3 = , p \ 0.01). Further, the firms commitment to their stakeholders helps to explain variation in the CSR latent variable (b 5 = 0.921, p \ 0.01; R 2 = 0.847). In Model 3 (see Fig. 3; Table3), we further expand our specification by allowing CSR and investment in R&D to interact. While CSR continues to have a significantly positive effect on both the financial health (b 6 =.043, p \.05; R 2 =.775) and the market valuation (b 7 =.120, p \.05; R 2 =.233) measures, the effect of investment in R&D is limited to the accounting-based performance measure (b 8 =.148, p \.01). Its effect on Tobin s Q, while still positive, is insignificant (b 9 =.029, p [.10). More importantly, the interaction term between CSR and R&D does not have any significant impact on either the financial health (b 10 =.011, p [.10) or Tobin s Q (b 11 =.148, p [.10). These results do not support our Hypothesis 2 regarding the interaction effect of the two perceptions on a firm s performance. 11 We have also included the correlations among our constructs in Model 3 (see Fig. 3). Apparently, our results do not support McWilliams and Siegel (2000) s expectation 11 Once again in model 3, firms with a better financial health had higher profitability (b 1 =.351, p \.01), higher liquidity (b 2 =.118, p \.01), and less leverage (b 3 = -.799, p \.01). Also, the KLD measures latent variable explained most of the variance of the CSR construct (b 5 =.920, p \.01; R 2 =.847).

10 616 W. Rodgers et al. Table 3 CSR leading to higher CFP model parameter estimates Pathways (regression weights) Model 1 Model 2 Model 3 Only CSR CSR and innovation CSR, innovation and interactions Profitability? financial health (b 1 ).309***.351***.351*** Liquidity? financial health (b 2 ).059***.119***.118*** Leverage? financial health (b 3 ) -.819*** -.801*** -.799*** Financial health? investor decision (b 4 ).425***.398***.388*** KLD measures? CSR (b 5 ).920***.921***.920*** CSR? financial health (b 6 ).044**.041*.043** CSR? investor decision (b 7 ).179***.187***.120*** Innovation? financial health (b 8 ).139***.148*** Innovation? investor decision (b 9 ).127***.029 CSR * innovation? financial health (b 10 ).011 CSR * innovation? investor decision (b 11 ).148 Multiple R 2 (explained variance) CSR Financial health Investor decision * Significant at p \.10; ** significant at p \.05; *** significant at p \.01 regarding a positive relationship between R&D investment and CSR. We found a negative but not significant correlation among these constructs (r 1 =-0.066; p [ 0.10). Even though these results do not support McWilliams and Siegel (2000) s expectation, they are similar to the negative correlation found by Hull and Rothenberg (2008). 12 Further, we found that a firm s lagged investment in R&D is negatively correlated with its current period profitability (r 2 =-0.324; p \ 0.01) and leverage (r 4 =-.315; p \.01), but positively correlated with its liquidity (r 3 = 0.520; p \ 0.01). The correlation between CSR and profitability is positive (r 5 = 0.09; p \ 0.05), its correlation with leverage is positive (r 7 =.124; p \.01), and its correlation with liquidity is negative (r 6 =-.089; p \.05). Overall, our results suggest that the social perception of a firm s commitment to meet CSR contributes both directly to the firm s market value and indirectly via the financial status of the firm, which supports our Hypothesis 1. However, its impact on the performance is not dependent on the firm s investment in innovation activities (Hypothesis 2). Additional Analyses Various Dimensions of CSR In a recent paper, Barnett and Salomon (2006) recommend researchers of the CSR CFP link to go beyond the debate 12 While McWilliams and Siegel (2000) reported a correlation between CSR and innovation of (p \ 0.01) over the period of , Hull and Rothenberg (2008) found a correlation of (p \ 0.05) over the period of on the financial merits of being socially responsible and explore the merits of the different social dimensions instead. They observe that for a sample of socially responsible investing (SRI) funds, while the community relation dimension had a positive impact on CFP, other types of social dimensions, such as employee and environmental, had a negative effect on the market-based CFP. 13 Barnett and Salomon (2006) pointed out that a limitation of their study is the use of one single marketbased measure of CFP, instead of an accounting-based one, or both. In Table 4 we show the path coefficients of Models 4, 5, and 6, which examine the impact of the employee, customer, and community relation dimensions of CSR, respectively. Models 4, 5, and 6 are similar to Model 3 but instead of the overall KLD measure of CSR, we examine the individual dimensions of CSR. That is, Model 4 shows the impact of the employee social commitment of CSR, Model 5 demonstrates that of the customer, and Model 6 illustrates the effect of the community relation on CFP. We also include innovation and its interaction with the individual dimensions of CSR in each model specification. The impact of profitability, liquidity, and leverage on a firm s financial health and the subsequent effect on investor decision (Tobin s Q) is similar in all the models. Similarly, innovation has a significant positive impact on the accounting-based measures of CFP in all models. However, its effect on the market-based performance measure, 13 In their study, Barnett and Salomon (2006) employed risk-adjusted financial performance as the dependent variable, measured as the monthly percentage change in a fund s market value adjusted by the fund s specific beta.

11 Do Investors Value a Firm s Commitment to Social Activities? 617 PERCEPTIONS Innnovation CSR R 2 = =.921*** KLD measures 6 =.041* 7 =.187*** 8 =.139*** 9 =.127*** Profitability 1 =.351*** Liquidity Leverage 2 =.119*** 3 = -.801*** Financial Health R 2 =.775 JUDGMENT 4 =.398*** Market Value R 2 =.224 DECISION INFORMATION N= 497 observations Period Fig. 2 PLS results for Model 2 (both CSR and innovation) r 1 = PERCEPTIONS Innovation CSR R 2 = =.920*** Lagged KLD measures CSR*Innovation 6 =.043** 10 = =.120*** r 2 = -.324*** r 5 =.090** 8 =.148*** 11 = =.029 Profitability r r 3 =.520*** 6 = -.089** Liquidity r 4 = -.315*** r 7 =.124*** Leverage 1 =.351*** 2 =.118*** 3 = -.799*** Financial Health R 2 =.775 JUDGMENT 4 =.388*** Market Value R 2 =.233 DECISION INFORMATION N= 497 observations Period Fig. 3 PLS results for Model 3 (CSR, innovation, and interactions)

12 618 W. Rodgers et al. Table 4 Social dimension types leading to CFP model parameter estimates Pathways (regression weights) Model 4 Model 5 Model 6 Employee Customer Community Profitability? financial health (b 1 ).358***.351***.350*** Liquidity? financial health (b 2 ).122***.125***.112*** Leverage? financial health (b 3 ) -.798*** -.801*** -.798*** Financial health? investor decision (b 4 ).391***.397***.387*** CSR? financial health (b 6 ).141***.046**.036 CSR? investor decision (b 7 ) ***.040 Innovation? financial health (b 8 ).141***.141***.144*** Innovation? investor decision (b 9 ) ***.082 CSR * innovation? financial health (b 10 ) CSR * innovation? investor decision (b 11 ) ** Multiple R 2 (explained variance) Financial health Investor decision * Significant at p \.10; ** significant at p \.05; *** significant at p \.01 Tobin s Q, is significant only in Model 5 where the customer dimension of CSR is included. The various dimensions of CSR also have different impact on CFP. The customer dimension of CSR has a significant positive effect on both the financial health and Tobin s Q CFP measures. The employee dimension, on the other hand, only has a positive affect on the financial health CFP measure (i.e., the accounting measure) but has no significant impact on Tobin s Q. CSR investment in community does not have any significant effect on either of the CFP measures. However, its interaction with innovation has a significant positive impact on Tobin s Q, suggesting that CSR investment in community relations only has a positive effect on investor decision for firms with high research and development intensity. That is, investors value a firm s investment in community relation only when it also invests in product innovation activities. Since the community relation is the least likely, among the three dimensions of CSR we track, to have any direct impact on a firm s performance, investors view the investment as value-enhancing only if the firm does not sacrifice product quality (as reflected in its commitment to innovation) for the investment in community relation programs. CFP Leading to Higher CSR and Innovation Investments In the previous sections, we have analyzed the effect of CSR investment on a firm s performance. One unsolved question in the CSR literature is the causality relationship between CSR and firm performance (Surroca et al. 2010). Preston and O Bannon (1997), Waddock and Graves (1997), McGuire et al. (1988), and Prior et al. (2008) have used the arguments of the slack resource theory to justify the opposite causality of CSR and CFP. Similar to the stakeholder theory, slack resource theory proposes a positive relation between CSR and CFP but in the opposite direction of causality. The main assumption of the slack resource theory is the availability of discretionary funds. Under this theory, high prior period CFP provides the surplus fund necessary for a firm to engage in CSR (McGuire et al. 1988, Ullmann 1985; Waddock and Graves 1997; Prior et al. 2008; Garcia-Castro et al. 2010). In other words, this theory implies that even if a firm wants to be a good citizen, its investment in CSR depends on its previous financial well-being (Waddock and Graves 1997). That is, a firm s CFP determines the resources a firm can invest in CSR activities and its reputation as a socially responsible citizen. While our results thus far support the hypothesis that CSR has a positive contribution towards a firm s financial well-being, it does not negate the possibility that a firm s CSR investment depends on its previous financial performance. In this section, we investigate whether a firm s performance in prior period affects its CSR commitment and social perception in the current period by reversing the direction of impact in our model. Specifically, according to the slack resource theory, we examine in Model 7 (see Fig. 4) the impact of lagged financial performance on a firm s CSR investment and hence the investors perception of the firm s social responsibility. Table 5 and Fig. 4 present the results of PLS analysis of this model and the correlations among the various constructs. Our results suggest that lagged financial health