Socially Responsible Investment Funds and Corporate Tax Avoidance

Transcription

1 Socially Responsible Investment Funds and Corporate Tax Avoidance Thomas Doellman a, Fariz Huseynov b, Sabuhi Sardarli c Draft: January 2016 Abstract We examine whether corporate tax avoidance impacts the investment decisions of socially responsible investment (SRI) funds. After controlling for corporate social responsibility (CSR) constructs, we find that investment by SRI funds is negatively associated with corporate tax avoidance. Further analysis indicates that greater tax avoidance leads to less SRI fund investment particularly in firms with CSR concerns, greater managerial entrenchment, and greater earnings management. Our results suggest that socially responsible investors consider corporate tax avoidance an important aspect of CSR. The results are consistent with recent findings in the literature that corporate tax avoidance is a part of corporate culture, and that corporate insiders view tax avoidance and CSR as substitutes. When corporate culture is likely to promote less responsible activities and greater agency issues in the firm, corporate tax avoidance deters investment by socially responsible investors. However, tax avoidance does not deter investment by social responsible investors in firms with a strong CSR reputation. Key words: tax avoidance, corporate social responsibility, socially responsible funds, mutual funds JEL classification: H26; G11; G23 a Saint Louis University, John Cook School of Business, Davis-Shaughnessy Hall, DS Lindell Blvd. St. Louis, MO tdoellman@slu.edu. Tel: b North Dakota State University, College of Business, P.O. Box 6050, Dept. 2410, R. H. Barry Hall 202, Fargo, ND fariz.huseynov@ndsu.edu. Tel: c Kansas State University, College of Business Administration, 117D Calvin Hall, Manhattan, KS ssardarli@ksu.edu. Tel:

2 I. INTRODUCTION Prior research suggests that corporate tax avoidance may increase firm value by reducing tax costs and increasing shareholders wealth. 1 However, because taxes provide funding for governmental infrastructure and social programs, corporate tax avoidance may benefit shareholders at the expense of society and may be viewed as a socially irresponsible corporate act (Sikka, 2010; Desai, 2012; Huseynov and Klamm, 2012). Despite an increasing number of studies that examine the association between corporate social responsibility and corporate tax avoidance (Huseynov and Klamm, 2012; Lanis and Richardson, 2012; 2015; Hoi et al., 2013; Davis et al., 2015; Watson, 2015), an investigation of how firms outside stakeholders respond to corporate culture that promotes both is still missing (Hoi et al., 2013). This paper fills this void in the literature by investigating whether and how the corporate tax avoidance behavior of a firm affects investment by socially responsible investment mutual funds (henceforth SRI funds). Studies of corporate social responsibility (CSR) associate economic goals of businesses with their social responsibilities, including acting ethically, contributing to development of community, and improving the quality of life of stakeholders (Carroll, 1991; Wood, 1991; Holme and Watts, 2006). If a reduction in tax payments leads to greater firm value as shown in Desai and Dharmapala (2009), then corporate tax avoidance may be associated with reaching economic goals and, hence, may be considered socially responsible. Moreover, socially responsible investors may consider humanitarian CSR as a replacement for the redistribution role of the public sector with a greater outcome than government welfare (Lantos, 2001; Porter and Kramer, 2006; Davis et al., 2015). 1 See for example, Graham and Tucker, 2006; Hanlon and Heitzman, 2010; Hanlon and Slemrod, 2009; Desai and Dharmapala,

3 Alternatively, if paying taxes is a commitment to community, stakeholders may view avoiding corporate taxes as socially irresponsible. Christensen and Murphy (2004) state that tax revenues are the lifeblood of, or the social contract vital to, the development and maintenance of physical infrastructure and to the sustenance of the infrastructure of justice that underpins liberty and the market economy. Erle (2008) and Schön (2008) also argue that avoiding taxes aggressively through tax sheltering should be deemed as socially irresponsible behavior. Thus, given reasonable arguments for either possibility, whether a firm that avoids corporate tax payments is viewed as socially responsible or irresponsible is an empirical question and will depend on the specific views of inside and outside stakeholders, with one group not necessarily agreeing with the other. In terms of inside stakeholders, Desai (2012) shows that while firm managers have increasingly become more embracive of CSR through engagement with community development, corporations have also become more aggressive about minimizing their tax obligations. This finding suggests that corporate insiders view CSR and tax avoidance as either independent of one another or as substitutes. Davis et al. (2015) find that CSR is negatively related to cash effective tax rates and interpret this result as evidence that firm managers view CSR and tax payments as substitutes. On the other hand, Hoi et al. (2013) find that firms with the most excessive irresponsible CSR activities have a significantly higher likelihood of also engaging in tax-sheltering activities and having lower cash effective tax rates. Their findings imply that corporate culture that promotes excessive irresponsible CSR activities also promotes greater tax avoidance. Despite the findings of Hoi et al. (2013) and Davis et al. (2015) regarding inside stakeholders, whether outside stakeholders consider corporate tax avoidance a significant aspect 2

4 of CSR is an important yet unanswered empirical question. The aim of our study is to shed light on this question and we argue that SRI fund investment proves an ideal testing ground for this investigation. First, increasing demand from investors who desire socially responsible investments has led to significant growth in the assets under management of SRI funds, from $136 billion in 2001 to over $1.9 trillion in Second, to serve the unique demands of socially responsible investors, SRI fund managers specifically incorporate social and environmental consequences of investments into the financial valuation process when making investment decisions. For instance, Domini Social Equity Fund states in its prospectus that Domini evaluates the Fund s current and potential investments against its social and environmental standards based on the businesses in which they engage, as well as on the quality of their relations with key stakeholders, including communities, customers, ecosystems, employees, investors, and suppliers. 3 Thus, determining whether corporate tax avoidance is associated with SRI fund investment, while controlling for firms CSR reputation, should provide a clear answer to the empirical question being studied. While it is not clear how corporate tax avoidance is viewed by socially responsible investors, anecdotal evidence provides some potential insight. Domini Social Investments states on its website that it views taxes as an investment in society and a source of funds for indispensable services, therefore avoiding tax payments can be deemed as unethical and socially irresponsible. 4 If Domini s view is generally shared by SRI funds, then this evidence would suggest that corporate tax avoidance is in fact an important aspect of CSR and that it may deter 2 For more details, please see the website of US Social Investment Forum: 3 Domini Social Investments provide more details of their stock selection process at their website. 4 For more information, please refer to Making a Difference section at Domini Social Investments website. 3

5 investment by SRI funds. However, again referring to the Domini Social Investments website, Domini states that its Social Equity Fund may determine that a security is eligible for investment even if a corporation s profile reflects a mixture of positive and negative social and environmental characteristics. 5 Thus, even if corporate tax avoidance is seen as a negative social characteristic, tax avoidance may not negatively affect SRI fund investment when firms have CSR strengths, consistent with a substitution effect similar to that found for inside stakeholder by Davis et al. (2015). On the other hand, for firms that have significant CSR concerns, less tax avoidance may still help attract SRI fund investment; whereas, greater tax avoidance may further deter investment. This relationship would be consistent with the findings of Hoi et al. (2013). Studying whether the effect of corporate tax avoidance on SRI fund investment differs based on the strength of a firm s CSR reputation also allows for the identification of a tighter link between tax avoidance and CSR. The finding of a negative relationship between tax avoidance and SRI fund investment while controlling for others aspects of CSR may suggest that SRI fund managers associate tax avoidance with poorer CSR. However, the finding of a negative relationship could also be consistent with SRI managers viewing tax avoidance as being more generally associated with risky corporate behavior and not necessarily dependent on the firm s CSR reputation. Evidence that corporate tax avoidance affects SRI investment for firms with a poor CSR reputation but not necessarily for firms with a strong CSR reputation would be consistent with findings in Davis et al. (2015) and a strong indication that corporate tax avoidance is an important component of CSR. 5 For more details, please refer to 4

6 Following the prior literature (Dyreng et al., 2008, 2010; Hanlon and Heitzman, 2010; Huseynov and Klamm, 2012), we use the term tax avoidance to represent reductions in a firm's taxes relative to pretax accounting income and measure corporate tax avoidance in two ways: GAAP tax expense to pretax accounting income (GAAP ETR) and taxes paid to pretax accounting income (Cash ETR). While Cash ETR has been used as a measure of broad corporate tax avoidance practices by numerous studies (Dyreng et al., 2008; Hoi et al., 2013; Davis et al., 2015), practitioners and tax executives consider GAAP ETR as much as Cash ETR in their investment decisions (Graham et al., 2014). For comprehensiveness, we use both measures. We identify SRI mutual funds using reports published by US Social Investment Forum (SIF) in 2007 and and calculate firms aggregate equity ownership by these funds (Bollen, 2007; Gil-Bazo et al., 2010). We obtain social ratings from KLD Research & Analytics, Inc. Following Huseynov and Klamm (2012) we control for three aspects of CSR, corporate governance, community, and diversity, to capture the marginal impact of tax avoidance, above and beyond the impact of CSR, on a SRI fund manager s decision to invest in a firm. Our main findings indicate that SRI funds are attracted to firms that avoid taxes less and are deterred by firms that avoid taxes more. However, when we separate our sample into two groups of firms, those with the strongest CSR reputation and those with the poorest CSR reputation, we find that corporate tax avoidance negatively affects SRI investment within firms with significant CSR concerns. In contrast, we find no impact of tax avoidance on SRI investment within firms with significant CSR strengths. To test the robustness of these important findings, we also examine whether SRI funds are less likely to invest in tax avoiding firms when management is entrenched and shareholder 6 Reports are available at the website of US Social Investment Forum: 5

7 rights are weakly protected leading to inefficiency of corporate tax avoidance and CSR activities (Desai and Dharmapala, 2009; Harjoto and Jo, 2011). Specifically, we find that SRI funds are likely to invest less into firms that avoid taxes more only within firms with high managerial entrenchment based on the Governance Index developed by Gompers et al. (2003). Additionally, we find that greater corporate tax avoidance leads to lower SRI ownership within firms that are most likely to manage their earnings based on discretionary accruals. Similar to our main result, these additional findings suggest SRI funds are less likely to invest in firms that avoid taxes more and generally do not have a corporate culture that is ideal for promoting CSR activity. Our findings are consistent with the substitution effect between CSR and tax avoidance found in Davis et al. (2015) as our results suggest that tax avoidance deters investment by social responsible investors to a significantly lesser degree as a firm strengthens its CSR. However, firms that have significant CSR concerns and avoid taxes aggressively may see a deterring effect of tax avoidance on investment by socially responsible investors. Here, consistent with Hoi et al. (2013), socially responsible investors may view the corporate culture of such firms as particularly engendering corporate activities that are at odds with the spirit of their investment objectives. Again, this is consistent with our finding that SRI fund investment is negatively associated with corporate tax avoidance when the firm has significant CSR concerns, entrenched management, and greater earnings management. Our study contributes to the growing literatures that examine corporate social responsibility and corporate tax avoidance. Specifically, we respond to a call for more research on the cross-section of these two literatures (Hanlon and Heitzman, 2010; Hoi et al., 2013). While Davis et al. (2015) and Hoi et al. (2013) examine inside stakeholders view of the relation between CSR and corporate tax avoidance, our study focuses on the perception of outside 6

8 stakeholders. Our study also differs from Hoi et al. (2013) and Davis et al. (2015) as we analyze the impact of GAAP effective tax rates and find that SRI mutual funds are more sensitive to lower GAAP ETRs than Cash ETRs suggesting that whether a firm avoids corporate tax expenses significantly impacts SRI fund investment. 7 More generally, we also contribute to the literature that investigates determinants of investment decisions by mutual funds. We show that, in response to the needs of their clientele, SRI funds are likely to incorporate corporate tax avoidance behavior of firms in their investment selection criteria and invest in firms with higher effective tax rates. The remainder of the paper is organized as follows. Section II discusses the related literature and develops the hypotheses. Section III presents the sample and research design. Section IV presents the empirical findings of the study, and Section V discusses further robustness checks. Finally, Section VI concludes. II. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT Corporate tax avoidance may enhance firm value if the potential benefits of tax savings outweigh direct and indirect costs. Direct costs include administrative costs, litigation expenses, and penalties imposed by tax authorities (Wilson, 2009), whereas indirect costs include political costs, potential reputation losses, a greater cost of debt capital, and a higher stock price crash risk (Hanlon and Slemrod 2009; Kim et al., 2011; Hasan et al., 2013; Graham et al., 2014). Another possible indirect cost of avoiding corporate tax payments for a firm is being perceived as socially irresponsible or unethical by investors (Hoi et al., 2013). The following reviews the literatures examining corporate tax avoidance and corporate social responsibility and discusses how 7 In a survey of nearly 600 corporate tax executives, Graham et al. (2013) find that 84% of executives care at least as much about the GAAP ETR as they do about cash taxes paid. 7

9 socially responsible investors may perceive corporate tax avoidance based on these literatures findings. Prior studies define corporate social responsibility (CSR) as acting ethically, contributing to economic development and making profit, obeying the law, and improving the quality of life of workers, local communities, and society at large (Holme and Watts, 2006; Carroll, 1991). According to CSR, a firm is responsible not only to shareholders, but also to all internal and external stakeholders (Wood, 1991; Jamali et al., 2008; Sikka, 2010). Among these stakeholders, socially responsible investors take social and environmental risks and governance criteria into consideration when modeling a firm s long-term financial success (Geczy et al., 2005). Institutional investors have satisfied preferences of these clients by offering socially responsible investment (SRI) funds that now account for a $2 trillion segment of the mutual fund industry. For example, the TIAA-CREF Social Choice Equity Fund states in its investment strategy that its evaluation process favors companies that are devoted to serving local communities, committed to higher labor standards,, and those managed in an exemplary or ethical manner. 8 In response to corporate stakeholders growing interest in socially responsible behavior, public corporations have become increasingly willing to issue voluntary CSR reports. For example, the market value of firms issuing such reports constituted over 10 percent of the total U.S. market capitalization in 2007 (Dhaliwal et al., 2011). Because SRI funds target firms that build strong relationships with their stakeholders and are devoted to community development, they may view avoiding corporate tax payments not only as a lack of social investment overall, but also as a behavior that may hurt the value of corporate reputation. Decreasing a firm s tax liabilities may benefit shareholders (Hanlon and 8 Full statement is available at 8

10 Heitzman, 2010), but avoiding payment of fair share of taxes that contribute to societal development is likely perceived as unethical by socially responsible investors given their general views (Erle, 2008; Schön, 2008, Huseynov and Klamm, 2012). Anecdotal evidence in the abovementioned example of Domini Social Equity Fund suggests that the extent of corporate tax avoidance in a firm is taken into consideration by SRI funds when making investment decisions. Based on the above discussions regarding the perception of tax avoidance by SRI funds and corporations, we propose the following hypothesis: Hypothesis 1: Ceteris paribus, investment in a firm by SRI funds is negatively associated with the firm s level of corporate tax avoidance. Despite a growing trend to voluntarily report CSR activities to increase positive publicity and increasing corporate investment in social programs, corporations have also become more aggressive when it comes to minimizing their tax obligations (Desai, 2012). According to the U.S. Government Accountability Office, more than half of American corporations no longer have significant domestic tax obligations (Desai, 2012). Davis et al. (2015) find that socially responsible firms tend to have lower effective tax rates and interpret this as evidence that managers see corporate tax avoidance and CSR as substitutes rather than mutually exclusive. Preuss (2010) and Sikka (2010) also provide qualitative evidence that firms that identify themselves as socially responsible engage in tax avoidance practices. The investors of SRI funds may share these views, especially if they disagree with the view that government is the best allocator of resources and believe that social programs are more efficient when implemented by for-profit corporations. Several studies argue that humanitarian CSR may well replace the redistribution role of the public sector and result in a greater outcome than government welfare (Lantos, 2001; Porter and Kramer, 2006; Davis et al., 2015). Moreover, 9

11 tax payments may reduce future investments of corporations which will not only decrease their profits and limit their capacity to invest in CSR activities, but also decelerate innovation and job growth in the overall economy (de Mooij and Ederveen, 2003; Djankov et al., 2008). Because economic profit is one of the pillars of CSR, socially responsible investors may not perceive a reduction in corporate taxes by a firm as a socially irresponsible activity. Following these studies, we formulate our second hypothesis as follows: Hypothesis 2: Ceteris Paribus, investment in a firm by SRI funds is positively associated with the firm s level of corporate tax avoidance. Prior studies suggest that CSR strengths and concerns carry distinct information and the outcomes of a firm s positive and negative social actions can be different (Mattingly and Berman, 2006; Huseynov and Klamm, 2012). For example, Huseynov and Klamm (2012) find that the interaction of community concerns with tax management fees positively affects both GAAP and Cash ETR, while the interaction of corporate governance strengths and diversity concerns with tax management fees negatively affects Cash ETR. Several empirical studies provide evidence that irresponsible CSR activities have greater explanatory power regarding the underlying CSR construct they intend to capture, while responsible CSR activities are too selfserving and tainted to provide such information (Hoi et al., 2013). CSR may also be evaluated from a broader perspective. For example, Hoi et al. (2013) present CSR as an important dimension of corporate culture that formulates the right course of action that takes into account the economic, social, environmental, and other externalized impacts of the company s activities. Several other studies show that corporate culture affects corporate behavior in various ways by reducing agency problems, increasing communication between executives and shareholders, and improving negotiation channels (Hermalin, 2001; Van 10

12 den Steen, 2010). Hoi et al. (2013) conjecture that CSR should influence corporate practices affecting the firm s various stakeholders. While Davis et al. (2015) show that CSR and tax avoidance are substitutes, by analyzing the association between tax sheltering and irresponsible CSR activities, Hoi et al. (2013) find that firms with excessive irresponsible CSR activities have a higher chance of engaging in tax-sheltering activities. Socially responsible investors may also be more cautious in their decisions to invest in firms with CSR concerns and apply other criteria when selecting their target firms. Specifically, they may consider corporate tax avoidance within the context of corporate culture that affects the effectiveness of CSR and tax avoidance activities of a firm. It is possible that outside stakeholders view corporate tax avoidance as unethical activity particularly when a firm s corporate culture fosters irresponsible CSR activities. Overall, these studies suggest that the effect of corporate tax avoidance on socially responsible investors investment decisions may differ depending on whether a firm is engaged in other socially responsible or irresponsible activities. If corporate tax avoidance is seen as equivalent to a CSR concern, tax avoidance may not negatively affect SRI investment in the presence of CSR strengths. However, consistent with Hoi et al. (2013), firms that avoid corporate taxes and also have CSR concerns may be perceived by SRI funds as companies that have poor corporate culture that engenders activities that are at odds with socially responsible investors beliefs. Thus, tax avoidance in these firms may further deter SRI fund investment. Again, these findings in tandem would be consistent with anecdotal evidence from Domini Social Investments that its funds may invest in firms that have both CSR strengths and weaknesses. They would also be consistent with the finding of a substitution effect between tax avoidance and CSR found in Davis et al. (2015). Our next set of hypotheses follows: 11

13 Hypothesis 3a: Ceteris paribus, investment by SRI funds is not associated with corporate tax avoidance when firms have CSR strengths Hypothesis 3b: Ceteris paribus, investment by SRI funds is negatively associated with corporate tax avoidance when firms have CSR concerns. III. DATA AND RESEARCH DESIGN Sample Selection We identify socially responsible investment (SRI) funds based on reports provided by US SIF (The Forum for Sustainable and Responsible Investment). 9 We match the ticker and fund name information to CRSP Mutual Fund Database portfolio holdings and obtain a firm s ownership by SRI funds. We aggregate holdings at the firm level to estimate the total number of shares held by SRI funds and divide that number by the firm s total number of shares outstanding. This is the main dependent variable in our analysis. Table 1 contains further details about all variables used in this study. Although the portfolio holdings data in CRSP is available starting in 2001, we choose to restrict the time period to due to poor coverage in earlier years. Table 2 reports summary statistics for the variables in this study. As expected, the unconditional mean of SRI fund ownership is low at 0.16% (i.e. firms with zero SRI investment are included). However, for firms that have non-zero SRI fund investment, the conditional mean of SRI ownership is 0.37%. Similar to prior tax research, we define tax avoidance as the reduction in a firm s taxes relative to its pretax accounting income whereby taxes are either the expense or the payments (Dyreng et al, 2008; 2010; Hanlon and Heitzman, 2010; Minnick and Noga, 2010; Huseynov and Klamm, 2012). We use the short- and the long-run values of two effective tax rate measures. We 9 US SIF provides reports on performance of SRI funds. Available at 12

14 refer to the short-run variables as SCash ETR and SGAAP ETR. SCash ETR is the income taxes paid (Compustat data item TXPD) divided by pretax income (data item PI) minus special items (data item SPI). SGAAP ETR is tax expense (data item TXT) divided by pretax income. The short-run ETRs allows us to capture the impact of short-term fluctuations in tax variables on the decisions of SRI to invest in a firm. Our long-run ETR measures (LCash ETR and LGAAP ETR) are more extensively used in prior literature. We calculate LCash ETR and LGAAP ETR similarly to the short-run measures except the centered moving sum of the input variables over three years are used in the calculation instead. 10 Despite accounting differences between Cash ETR and GAAP ETR, we use both ETR measures as they are the result of a firm s corporate tax avoidance strategy. This also allows us to compare and contrast the impact of a reduction in a firm s tax payment and/or expense. Dyreng et al. (2008) argue that Cash ETR is a better measure of corporate tax avoidance, however, recent studies show that GAAP ETR is essential to tax management and other corporate decisions (Armstrong et al, 2011; Graham et al., 2013). 11 Summary statistics indicate that mean SCash ETR and LCash ETR levels are 22% and 24%, respectively, while both GAAP ETR measures are slightly higher at 26% (Table 2). To isolate the effect of corporate tax avoidance, we control for other firm specific factors that may affect SRI fund investment in a firm. First, we control for a set of variables that have been found to affect mutual fund investment in a firm (Del Guercio, 1996). These variables include firm size, firm stock volatility, dividend yield, book-to-market ratio, leverage, tax-loss carry-forwards and dummy variables for firms with positive earnings and regulated utilities sector. We collect data from CRSP and Compustat for these firm control variables. Second, we 10 We use the centered moving sum of variables to incorporate forward-looking analysis of tax variables by potential investors. 11 Graham et al. (2013) survey responses from nearly 600 corporate tax executives and find that 84% of care at least as much about the GAAP ETR as they do about cash taxes paid. 13

15 control for a firm s social rating information provided by KLD Social Rating Database. Following Huseynov and Klamm (2012) we focus on corporate governance, community, and diversity aspects of CSR. 12 For each category, KLD adds the number of strengths and concerns separately. Items in the governance category include ownership, compensation, transparency, and political accountability. The community category includes support for community, charitable giving, and volunteering. Finally, the diversity category rates the representation of minority and diverse employee groups in the organizational structure of a firm and a firm s involvement in affirmative action controversy. For a given firm and year, CGOV_STR, COM_STR, and DIV_STR are the sum of corporate governance (maximum is 3), community (maximum is 5), and diversity strengths (maximum is 7), respectively (see Panel A of Table 2). Similarly, for a given firm and year, CGOV_CON, COM_CON, and DIV_CON are the sum of corporate governance (maximum is 4), community (maximum is 3), and diversity concerns (maximum is 3), respectively. The correlation matrix presented in Panel B of Table 2 indicates that SRI fund ownership is positively correlated with effective tax rates. Thus, SRI fund managers are more likely to own firms that avoid taxes less. The correlation coefficients between SRI fund ownership and KLD variables provide univariate evidence that SRI funds are more likely to invest in firms with CSR strengths and are less likely to invest in firms with diversity concerns. Interestingly, we find a positive correlation between governance concerns and community concerns, and SRI fund ownership. It is consistent with anecdotal evidence discussed in the introduction that SRI funds may invest in firms that have both CSR concerns and strengths. 12 Our main results are similar when we also control for the strengths and concerns based on employee relations and environmental factors. 14

16 We also note that all effective tax rates are highly correlated with one another suggesting that firms that engage in corporate tax avoidance are likely to reduce both tax payments and expenses in the short- and the long-run. We do not observe a clear pattern of correlation between tax avoidance and KLD variables; however, we note that less diverse firms (firms with more diversity concerns) are likely to have lower ETRs, and hence avoid taxes more. Methodology Our main research question is whether corporate tax avoidance affects a firm s ownership by SRI mutual funds. We test this relationship using the following regression model: SRI Ownership i,t = α 0 + β 1 ETR i,t + θ k Controls k + ε i,t (1) where ETR represents our four measures of tax avoidance, the short- and the long-run Cash and GAAP ETR, regressed separately, and SRI Ownership is the aggregate proportion of a firm s shares held by mutual funds that are identified as social responsible investment funds. Control variables include firm stock liquidity (LogStockLiquidity), a dummy variable for whether the firm is a component of the S&P500 Index (SPIndex), firm size (LogMCAP), stock volatility (Inverse_Trisk), dividend yield (Div_Yield), book-to-market ratio (LogBTM), and tax-loss carryforward (TaxLossCF). Dummy variables for whether a firm has announced positive earnings (Pos_Earnings) and whether the firm is a regulated utility firm (Reg_Util_Dum) are also included. We also include CSR variables to control for firms social responsibility aspects that are key for the investment decisions by SRI funds. We include the total number of CSR strengths and total number of CSR concerns for the three KLD categories mentioned above. We also control for each of the three CSR categories separately to avoid loss of information during aggregation (Huseynov and Klamm, 2012). We expect that ETR will have a significant impact on 15

17 SRI ownership even after controlling for CSR variables. Because a lower ETR indicates greater corporate tax avoidance, a positive (negative) coefficient on ETR would suggest that SRI fund ownership is positively (negatively) associated with effective tax rates and hence negatively (positively) associated with corporate tax avoidance. IV. EMPIRICAL RESULTS Tax avoidance and SRI mutual fund ownership Before conducting a multivariate regression analysis, Table 3 reports univariate statistics for SRI investment across firm groups based on the level of their CSR. First, for each CSR category, we separate our sample into two groups of firms with strengths only and those with concerns only, and compare mean SRI ownership across these groups. The results in Panel A of Table 3 indicate that firms with CSR strengths only tend to have greater SRI investment. The differences vary between 0.07 to 0.17 percentage points, depending on the CSR category. Thus socially responsible investors are likely to invest more (less) in firms with CSR strengths (concerns). Next, within each group of firms formed based on the level of CSR, we further divide firms into two groups based on their effective tax rates low ETR and high ETR using the median level of the ETR measure. We do this for each CSR category and each ETR measure. The results in Panel B of Table 3 indicate that within the firms with CSR concerns only, SRI fund ownership is lower in firms that avoid taxes most (low ETR). The differences are significant at the 1% confidence level in most cases. However, SRI fund ownership is not different based on the ETR level for firms with CSR strengths only. In fact, within the firms with 16

18 community strengths, mean SRI ownership is significantly higher in firms that avoid tax expenses (GAAP ETR) most. Overall, the univariate results suggest that SRI fund investment is higher in firms with CSR strengths only, and greater tax avoidance within this subsample of firms does not deter SRI fund investment. These univariate results are consistent with CSR and tax avoidance being viewed as substitutes by SRI fund managers, similar to the view of firm managers (Davis et al., 2015). On the other hand, SRI fund investment is lower in firms with CSR concerns only, and greater tax avoidance within this subsample of firms deters SRI fund investment even more. These findings support the results of Hoi et al. (2013) as tax avoidance deters investment by SRI funds when firms exhibit signs of an overall corporate culture that promotes corporate activities that are at odds with the spirit of socially responsible investors investment objectives. Our main multivariate results for the effect of corporate tax avoidance on SRI fund investments are presented in Table 4. For each ETR measure, we present our results first controlling for the total number of concerns and strengths across the three CSR categories studied (governance, community, and diversity), and then controlling for the number of concerns and strengths by each of the three CSR categories separately. 13 Within the community category, we control for tax disputes separately with a dummy variable equal to one if the firm has concerns related to tax disputes. This is done to control for tax disputes directly rather than potentially losing it in the sum of community concerns. We want to ensure that our ETR measures capture the effect of corporate tax avoidance above and beyond the effect of aspects of CSR, including specifically concerns related to tax disputes. 13 It is important to note that our results are robust to including other categories of CSR, such as environment, employee relations and product available in the KLD database. 17

19 In Column 1 through Column 4 of Panel A, the coefficient of SCash ETR is positive in each specification, however it is statistically significant (at the 10 percent confidence level) only when we control for the impact of community strengths and concerns. However, in Column 5 through 8 we find that the coefficients of SGAAP ETR are positive and strongly significant even after controlling for a firm s CSR aspects. This finding indicates that higher SRI fund ownership is associated with lower corporate tax avoidance. In other words, SRI funds are likely to invest more in firms with higher GAAP effective tax rates. More specifically, a one standard deviation increase in SGAAP ETR at the mean (from 26% to 42%) increases SRI fund ownership by 2.2 to 2.5 percentage points at the mean. Interestingly, such an increase in ETR would bring a firm s effective tax rate closer to the statutory tax rate. We also note that SRI funds tend to invest more in larger firms, in firms with lower stock liquidity, and in firms with lower stock volatility. They also favor firms that are components of the S&P 500 Index and firms that are in the regulated utilities sector. When using the long-run tax avoidance measures, LCash ETR and LGAAP ETR (Panel B), we find that greater tax avoidance again leads to lower SRI investment, and the effect is statistically significant across all specifications. Overall, the results in Table 4 provide evidence supporting H1 rather than H2 as effective tax rates have positive and statistically significant effect on SRI fund ownership controlling for firm s CSR reputation. In other words, consistent with H1, SRI funds are likely to invest more in a firm that avoids taxes less (i.e. has larger tax payments and tax expenses). We also find that SRI funds tend to invest in firms with strengths in governance, community, and diversity, consistent with their mission statements and investment selection criteria. CSR strengths and concerns 18

20 In this sub-section, we examine whether the impact of corporate tax avoidance on SRI fund ownership is affected by the level of CSR in a firm. For this analysis, we separate firms into one of two groups, Good CSR or Bad CSR, using three different selection criteria described below. We then run the regression model described in equation (1) for each group separately and report the results in Table 5. Because our groupings are done based on CSR variables, we drop them from our control variables group. Panel A of Table 5 reports our results for CSR groups defined based on a firm s number of strengths versus its number of concerns. Specifically, Good CSR (Bad CSR) includes firms for which the sum of CSR strengths (concerns) is greater than the sum of CSR concerns (strengths) for the three KLD categories studied. We find that all ETR measures have a positive and statistically significant effect on SRI fund ownership for Bad CSR firms. For instance, a one standard deviation increase in SGAAP ETR (16 percentage points) at the mean increases SRI fund ownership by 2.5 to 4.0 percentage points at the mean. In contrast, the impact of ETRs on SRI fund ownership is positive but statistically insignificant for Good CSR firms, with the exception of LCASH ETR which is significant at just under the 10% confidence level. In Panel B, Good CSR includes firms with a sum of CSR strengths that is greater than that of the firm s industry median and a sum of CSR concerns that is equal to or below that of the firm s industry median. Similarly, Bad CSR includes firms with a sum of CSR concerns that is greater than the firm s industry median and a sum of CSR strengths that is equal to or below the firm s industry median. We find similar results to Panel A as the effect of each of the four ETR measures on SRI ownership is positive and statistically significant for Bad CSR firms. The effect of ETR on SRI ownership is statistically insignificant for Good CSR firms in each of the four specifications. 19

21 Finally, in Panel C we present our results where the level of CSR is defined based on whether a firm has CSR strengths only or concerns only. Specifically, Good CSR includes firms with a sum of CSR strengths greater than zero and a sum of CSR concerns equal to zero for the three KLD categories studied; whereas, Bad CSR includes firms that have a sum of CSR strengths equal to zero and a sum of CSR concerns greater than zero for the three KLD categories studied. The results in Panel C are again very similar to those in Panels A and B as tax avoidance is associated with lower SRI investment for only firms in the Bad CSR subsample. We interpret the findings in Table 5 as evidence to support our hypothesis (H3b) that SRI mutual funds respond more to firms corporate tax avoidance policies when firms have social responsibility concerns. Specifically, the results are consistent with the findings of Hoi et al. (2013) that corporate tax avoidance is influenced by corporate culture and SRI funds view corporate tax avoidance negatively when a firm also has irresponsible CSR activities. On the other hand, SRI funds seem to be more forgiving for firms with strong social ratings when it comes to corporate tax avoidance. This may be due to SRI investors belief that investment in humanitarian CSR by socially responsible firms leads to significantly positive outcomes that outweigh any potential negative social and environmental effects of corporate tax avoidance. Overall, the results in Table 5 are also consistent with the finding by Davis et al. (2015) that socially responsible firms have lower cash effective tax rates suggesting a substitution effect between CSR and tax payments. Their results suggest that socially irresponsible firms tend to have higher effective tax rates; but, our results suggest that, all else equal, socially irresponsible firms with less tax avoidance will experience significantly higher investment by SRI funds. It is also important to note that, although Hoi et al. (2013) and Davis et al. (2015) do not include GAAP ETRs in their analysis, our GAAP ETR results also complement their findings that the 20

22 socially sensitive external stakeholders represented by SRI funds are concerned by corporate tax avoidance particularly in firms with CSR concerns. Evidence from Other Characteristics of Corporate Culture We examine the impact of two additional aspects of corporate culture on the association between corporate tax avoidance and SRI investment. First, we examine the role of managerial entrenchment. Van den Steen (2005, 2010) and Hoi et al. (2013) argue that strong corporate culture reduces agency problems because it produces shared beliefs and/or conventions of business practices that persist over time. Corporate culture may also align goals of managers with those of shareholders which is essential for value efficiency of corporate tax avoidance (Graham and Tucker, 2006). Prior tax research places tax avoidance in a principal-agent framework and examines whether the existence of greater agency problems exacerbates the negative outcomes of tax avoidance such as managerial diversion and the waste of scarce resources. 14 Desai and Dharmapala (2006) models the relationship between governance structures and tax avoidance and find that when management is less entrenched properly designed incentives can limit aggressive tax avoidance practices of managers. Similarly, Minnick and Noga (2010) find that stronger corporate governance is associated with an increase in tax efficiency. Prior CSR literature envisages an overlap between CSR and corporate governance (Jamali et al., 2008; Huseynov and Klamm, 2012) such that some models define corporate governance as a foundation for CSR or CSR as a dimension of corporate governance. Corporate governance structure ensures that firms are held responsible not only to internal stakeholders, but 14 For example, Chen and Chu, 2005; Crocker and Slemrod, 2005; Desai and Dharmapala, 2006, 2009; Minnick and Noga, 2010; Hanlon and Heitzman, 2010; Chen et al., 2010; Armstrong et al.,

23 also to external stakeholders and society in general. Therefore, socially responsible investors may view managerial entrenchment as a result of poor corporate governance and potentially leading to irresponsible CSR activities. Moreover, socially responsible investors may consider corporate tax avoidance by entrenched management as likely to benefit the managers and to divert firm s resources, rather than enhancing firm value. Overall, these studies suggest that CSR investors perception of corporate tax avoidance may be affected by the level of managerial power in a firm. Indeed, Hanlon and Slemrod (2009) discusses anecdotal evidence that institutional investors are generally concerned with the potential of substantial loss of shareholder rights when a firm engages in tax sheltering. 15 It is possible that the SRI funds may invest less in a tax avoiding firm particularly when the firm has greater levels of management entrenchment stemming from a broader corporate culture that fosters irresponsible CSR activities. While one of the KLD categories focused on to this point is meant to capture CSR related to corporate governance, G Index (Gompers et al., 2003) is an alternative measure of corporate governance that has been used extensively in prior studies, including as a proxy for management entrenchment when studying the interplay between CSR and governance (Harjoto and Jo, 2011). G Index reflects managerial power, the rights of shareholders, and the level of investor protection. Specifically, G Index is simply the number of antitakeover provisions adopted by a given firm with lower values indicating a lower degree of insulation of incumbent managers (Desai and Dharmapala, 2006). Thus, G Index provides an alternative measure of governance, 15 For example, Hanlon and Slemrod (2009) provide an article at CalSTRS (California State Teachers Retirement System) Web site about institutional investors taking out a full page ad in the USA Today calling for Ingersoll Rand (which had moved its headquarters to Bermuda for tax reasons) to move back to the U.S. citing a substantial loss of shareholder rights that goes with off-shore incorporation. 22

24 and specifically a measure of managerial entrenchment, that proves an ideal robustness test of our main results in Table 5. Similar to Desai and Dharmapala (2006), we use G Index in 2006 as our measure of agency problems to obtain the widest coverage of firms. 16 We separate firms in our sample into two groups: low management entrenchment firms and high management entrenchment firms based on the median level of G Index. We run the regression model in equation (1) for each subsample separately and report the results in Table 6. We find that all ETR variables are positive and statistically significant in firms with high management entrenchment whereas neither ETR variable has a significant impact on SRI ownership in firms with low management entrenchment. Although not reported, it is also important to note that the results are very similar when we control for each CSR category separately as in Table 4. Overall, the results from Table 6 are in line with the notion that corporate culture affects the perception of corporate tax avoidance such that SRI ownership is negatively associated with corporate tax avoidance particularly when managerial entrenchment is high and shareholder rights are not protected well. We interpret this as evidence that SRI funds are concerned with corporate tax avoidance particularly when a firm s agency problems are severe. Next, we examine the impact of corporate tax avoidance on SRI investment when considering firms earnings quality. Li et al. (2013) find that culture has a significant impact on corporate risk-taking, including the use of aggressive accounting practices such as earnings management. Because the manipulation of earnings hurts firm value in the long-run, CSR investors may perceive earnings management as unethical behavior. Indeed, Hong and Andersen (2011) find that firms with stronger CSR have higher financial reporting quality in the form of 16 We obtain the index values for each firm from Andrew Metrick s website. 23

25 higher quality accruals and less activity-based earnings management. Several studies also link corporate tax avoidance to earning management and suggest that firm managers can avoid taxes or use certain tax accounts, such as the valuation allowance, contingency reserve, and the foreign tax benefits, as a part of broader earnings management agenda to meet or beat financial analyst forecasts (Dhaliwal et al., 2004; Frank and Rego, 2006; Blouin et al, 2010; Hanlon and Heitzman, 2010). Consequently, corporate tax avoidance as part of an earnings management scheme should worsen the information environment and deter potential investors. For example, Kim et al. (2011) argue that managers tend to hoard bad news (including earnings manipulation and tax avoidance) for an extended period and increase the probability of future stock price crashes. Thus, it is possible that the impact of corporate tax avoidance on social investors may be significant when a firm s accounting quality is low. To examine whether the impact of corporate tax avoidance on SRI fund ownership is affected by the firm s level of earnings quality, we divide our sample into two groups - high earnings management and low earnings management - based on the median value of absolute discretionary accruals. Following prior literature (Yu, 2008; Hong et al., 2014), we use the modified cross-sectional Jones (1991) model, as described by Dechow et al. (1995) and modified by Kothari et al. (2005), to estimate discretionary accruals. The absolute value of discretionary accruals is our measure of earnings management. We then run our main regression model in equation (1) for each group separately. Results are presented in Table 7. We find that the level of earnings management indeed plays a significant role for the perception of corporate tax avoidance by SRI funds. Although the short-run ETR coefficients are not statistically significant, we find some evidence that SRI fund ownership is positively associated with long-run Cash ETR and GAAP ETR in firms with high earnings management. 24