The Effects of Institutional Corporate Social Responsibility on Bank Loans

Abstract

The authors study the impact of institutional corporate social responsibility (CSR)—defined as CSR targeted at a borrowing firm’s secondary stakeholders—on bank loans. Findings suggest that higher levels of institutional CSR are associated with lower levels of interest rates and loan spreads. In addition, institutional CSR also tempers the positive impact of loan maturity and firm leverage on interest rates and loan spread. These effects were strongest among firms that demonstrated sustained performance, rather than among firms that showed mixed performance in terms of their secondary stakeholder-related activities. This study indicates institutional CSR is valued by stakeholders for its risk mitigating and transaction cost reducing effects independent of technical CSR, defined as CSR targeted at primary stakeholders.

Keywords

bank loans corporate social responsibility CSR institutional CSR stakeholders technical CSR

How does corporate social responsibility (CSR) affect a firm and its stakeholders? This fundamental question has stimulated an extensive body of literature examining the relationship between CSR and financial performance (see reviews in Margolis & Walsh, 2003; Orlitzky, Schmidt, & Rynes, 2003). This body of work, however, reports conflicting evidence, and it is unclear whether in fact engaging in CSR provides any economic benefits to a firm.

Given the inconclusive nature of the evidence, more recently scholars have turned attention to culling out the underlying mechanisms by which social responsibility confers and reduces value (Jayachandran, Kalaignanam, & Eilert, 2013; McWilliams & Siegel, 2010). In this regard, there has been rich progress, with studies highlighting that CSR play a critical role in fostering the development of various intangible resources (Gardberg & Fombrun, 2006; Hull & Rothenberg, 2008; Surroca, Tribò, & Waddock, 2010), such as human capital and innovation. Others have noted that CSR enables a firm to differentiate itself in product markets (Hull & Rothenberg, 2008) while enhancing customer (Lev, Petrovits, & Radhakrishnan, 2010; Servaes & Tamayo, 2013) and employee satisfaction (Vitell & Davis, 2004).

A discernible shift in these latter studies is that they examine the mechanisms by which CSR adds value not just to the firm and its shareholders but also to various other stakeholders, such as customers and employees. The underlying logic is that if CSR benefits various stakeholders, eventually it would also enhance firm performance as these stakeholders pass on the benefits in the form of more favorable implicit and explicit terms of exchange. Continuing with this line of thinking, the authors study the impact of CSR on another group of stakeholders that has received comparatively less attention, namely, banks. From a theoretical standpoint, banks are important because as stakeholders they are specifically concerned with the default risk inherent in a firm at the time of the lending transaction, and its ability to pay back a loan. As a result, banks are likely to reward the CSR activities of a firm with lower interest rates to the extent that these activities defray default risk and preserve value (Godfrey, Merrill, & Hansen, 2009; Wang & Bansal, 2012). Equally importantly, banks also incur substantial transaction costs when administering a loan to a firm. These costs include various ex-ante costs, such as assessing creditworthiness, and ex-post costs such as monitoring to ensure proper progress is being made toward paying back their loans. Thus, studying bank loans also offers a rich opportunity to examine the important question of whether CSR confers value on a firm and its stakeholders by reducing transaction costs.

This study focuses on theoretically and empirically demonstrating these dual effects of CSR. When examining these effects, we make a critical distinction between CSR directed at primary stakeholders (or technical CSR) and secondary stakeholders (or institutional CSR). Goss and Roberts (2011) study the impact of CSR on bank loans, but they do not distinguish between the two dimensions. This lack of distinction makes it unclear whether the effects of CSR directly stem from activities that are intertwined with profits rather than from investments directed at secondary stakeholders, which exemplify the spirit of CSR. Thus, while Goss and Roberts (2011) made an important advance, a comprehensive understanding of the effects of CSR on bank loans (and concomitantly on reducing risk and transaction costs) remains as yet an unresolved issue.

This study’s main findings are as follows. First, the authors find that higher levels of institutional CSR, or CSR directed at secondary stakeholders, are associated with lower levels of interest rates and loan spreads¹ on bank loans made to firms engaging in these activities. These effects persist even when controlling for the effects of technical CSR directed at primary stakeholders. Moreover, we find that institutional CSR also reduces the positive effect of two critical factors that have been consistently found to increase loan spread in prior literature: firm leverage and loan maturity. We do, however, find that institutional CSR is effective in reducing loan spread only as long as the firm shows sustained and coherent performance in addressing secondary stakeholder concerns. Where the performance was erratic and the firm registered negative concerns or events, institutional CSR activities were no longer impactful in reducing loan spread.

The authors interpret this evidence as indicating that institutional CSR has effects independent of technical CSR in reducing default risk and transaction costs in the loan market. Prior literature has been less than clear-cut with regard to the effects of institutional CSR, with studies finding both positive (Godfrey et al., 2009) and negative effects (Hillman & Keim, 2001). This study provides critical support for the former view, while also highlighting the boundary conditions beyond which institutional CSR loses its beneficial effects.

The remainder of this article has five additional sections. The “Theoretical Background” section provides the theoretical background. The “Institutional CSR and Its Impact on Bank Loans” section presents the hypothesis development. The “Data and Method” section describes the data and methodology. The empirical results are described in the “Results” section; and the final section presents conclusions and direction for future research.

Theoretical Background

CSR can be conceptualized as a broad spectrum of strategies and investments used by firms to facilitate stakeholder relationships. Examples of these investments include pollution control (King & Lenox, 2002; Kock, Santaló, & Diestre, 2012), controls and practices to reduce discrimination, workplace safety measures, and investments in local communities. Stakeholder Management Theory suggests that managers have a fiduciary responsibility to all stakeholders, and not just to shareholders (Freeman, 1994). CSR investments constitute an avenue for firms to develop and attain legitimacy with their various stakeholders by demonstrating that they are not just responsive to shareholder concerns (Basu & Palazzo, 2008).

While stakeholder theorists have long recognized the distinction between primary stakeholders and secondary stakeholders in understanding the effects of CSR, more recently a new theoretical perspective has been brought to this distinction by Mattingly and Berman (2006) and Godfrey et al. (2009). These authors relate the two stakeholder groups to the technical and institutional environments of the firm, respectively (Scott & Meyer, 1983). These two environments are fundamentally distinct in the sense that technical environments comprise the exchanges between the firm and the immediate providers of goods and services, with the underlying logic of these exchanges rooted chiefly in effectiveness and efficiency. In contrast, institutional environments comprise the broader rules, regulations, and norms that are inherent in a society. In these environments, rather than effectiveness and efficiency, the goal is to build legitimacy and gain acceptance with various institutional and societal actors. These two distinctions have played a fundamental role in how institutional theorists have conceptualized the behaviors of organizations and firms. For our purposes, they also point to distinct mechanisms by which CSR can create and confer value—that is, by building relations with immediate exchange partners when a firm engages in technical CSR, and by building relations and legitimacy with a wider range of societal members and actors when a firm engages in institutional CSR.

A central premise of institutional theory is that that isomorphism, or conformity with institutional norms and values, buffers the firm from turbulence and improves its odds of survival (DiMaggio & Powell, 1983; Meyer & Rowan, 1977). When there is isomorphism, the continuity of resource flows to the firm is assured from various societal members (such as government regulators, opinion leaders, etc.) as the firm embeds itself within the web of social relations. As a consequence, the firm’s survival becomes less sensitive to technical performance, as various actors including investors become willing to extend support to it under a wide range of environmental circumstances. These arguments directly speak to the effects of institutional CSR in terms of reducing the risk inherent in a firm and preserving value (see Campbell, 2007, for a much more comprehensive treatment of CSR and institutional theory). Institutional CSR essentially acts as a form of linkage (Baum & Oliver, 1991) through which firms can build ties with various societal actors, while strengthening legitimacy (Bansal & Roth, 2000; Mitchell, Agle, & Wood, 1997). As a mechanism for building legitimacy, it is particularly powerful because unlike mandated rules and regulations, institutional CSR reflects attempts at self-regulation to conform to norms of society (Brammer, Jackson, & Matten, 2012). Thus, apart from these activities directly benefiting various institutional actors, the nature of the investments suggests that the firm has instituted strong “societal governance” (as opposed to just corporate governance; Brammer et al., 2012) in its activities, making institutional CSR a particularly effective means to build credibility and attain legitimacy.

In return for these societal governance-related activities and investments, institutional actors and secondary stakeholders may extend support to the firm and buffer it in various ways, such as for example through support for its products and services, and by ensuring support and continuity of resources both through good and bad times (Ulmer, 2001). While these arguments suggest a direct link between institutional CSR and the firm’s intrinsic riskiness and ability to sustain itself, a second impact that institutional CSR could have on stakeholders is in terms of the costs of transacting with the firm. Williamson (1985) argues that contracts governing transactions with primary stakeholders are fundamentally incomplete. The incompleteness arises because it is impossible to anticipate all future states of the environment, and hence to specify the parties’ responsibilities under each state ex-ante. Given that contracts are fundamentally incomplete, ceteris paribus, firms that engage in institutional CSR are likely to be viewed as more attractive transacting partners by primary stakeholders than non-CSR firms. Institutional CSR investments can act as a particularly credible signal to primary stakeholders that the firm is willing to uphold norms of fairness and reciprocity in its transactions (Margolis & Walsh, 2003; Matten & Moon, 2008). As a consequence, stakeholders may perceive firms engaging in institutional CSR as being potentially less opportunistic, and a more reliable and fair transacting partner than firms that do not engage in such activities (Aguilera, Rupp, Williams, & Ganapathi, 2007).

The consequence of this perception is that stakeholders are likely to incur lower transaction costs when engaging in various types of exchanges with the firm. For example, stakeholders may be more willing to leave various terms implicit and “to be decided” later as circumstances emerge, resulting in lower contract writing costs ex-ante (Cornell & Shapiro, 1987). Transacting with firms engaging in institutional CSR may carry the assurance that the implicit terms governing the exchange between parties are more likely to be honored under different states of the environment ex-post (McGuire, Sundgren, & Schneeweiss, 1988), leading once gain to lower contracting costs. Furthermore, by providing a signal of fairness and trustworthiness, institutional CSR may also reduce the search costs for stakeholders as they decide who to transact with in various intermediate markets. As a consequence, stakeholders may also be willing to pass on some of the benefits to the firm in the form of lower prices and/or more favorable terms of exchange.

While institutional CSR can provide benefits through these various mechanisms, prior literature points out that it is also fragile in the sense that its benefits can evaporate rapidly. Expectancy violation theory suggests people react more strongly to actions that violate their previous perceptions of how the other party is likely to behave (Mishina, Block, & Mannor, 2012; Rhee & Haunschild, 2006). The higher the expectation along a particular dimension, the greater the negative reaction to violations along that dimension (Pfarrer, Pollock, & Rindova, 2010; Rhee & Haunschild, 2006; Servaes & Tamayo, 2013). Moreover, the negative reactions are likely to be particularly pronounced when the expectation surrounds an intangible, character-based dimension (Rindova & Fombrun, 1999). In the case of such dimensions, positive cues are considered the baseline and conforming to expectations, whereas negative cues are considered much more indicative of the unobservable, underlying character of the firm (Pfarrer et al., 2010).

In the context of institutional CSR, this suggests negative events (such as governmental actions and employee lawsuits) and acts of irresponsibility (e.g., violations of laws and regulations) can have particularly detrimental effects and neutralize its risk-reducing and transaction cost benefits rapidly. These effects may occur as stakeholders reevaluate their perceptions and no longer view the firm as a trustworthy transacting partner who practices norms of fairness and reciprocity. Indeed, the skepticism may not just cause a neutralization of benefits but also lead to institutional CSR being transformed into a liability as stakeholders feel let down and start viewing the firm negatively. Such sentiments may also result in the overall revaluation of the organization, including any previous CSR acts, being reassessed in a skeptical light. We build on these arguments to study the impact of institutional CSR on bank loans.

Institutional CSR and Its Impact on Bank Loans

The Bank Loan Market

In the bank loan market, the central price charged by lenders is the loan spread, or interest rate in excess of London Interbank Offered Rate (LIBOR). The higher the loan spread, the higher the interest rate, and the more costly the loan. A primary determinant of loan spread is the borrower’s default risk. This risk ultimately reflects the bank’s level of exposure by virtue of contracting with a borrower (Graham, Li, & Qiu, 2008). During the process of drafting a loan, banks evaluate a potential borrowers’ default risk based on various firm and loan characteristics. To make an accurate evaluation, the bank usually has to overcome significant information asymmetry because creditworthiness and the risk of default depend not only on tangible capital but also on intangible capital (Strahan, 1999). The banks’ evaluation of credit risk is ultimately incorporated into the price of the loan, with greater levels of risk being associated with a higher spread. Alternatively, banks may also share or pass on the lower costs of assessing creditworthiness to a firm (Bharath, Dahiya, Saunders, & Srinivasan, 2011) thereby resulting in lower loan spreads.

In addition to ex-ante costs of assessing creditworthiness, banks also incur significant costs in terms of ex-post monitoring of loans. Banks typically include various covenants in their loan contracts which necessitate that borrowing firms fulfill conditions, such as maintaining a minimum level of capital and net worth, and regular reporting of financial information to the bank. The reason is because banks are exposed to downside risks whereas, unlike shareholders, they do not capture upside gains. When these loan covenants are violated, substantial control is shifted to the bank which can then choose to take the appropriate action that best results in the pay back of the loan (Chava & Roberts, 2008). The significant ex-post contractual ability to monitor firms and the attendant costs that are incurred also set banks apart from other stakeholders, such as suppliers.

Given these various factors, the loan market provides a particularly useful context to study the effect of CSR in reducing risk as well as in terms of its effects in reducing transaction costs.² Besides these advantages, the loan market is important in its own right. Bank loans are the primary source of external corporate financing for U.S. corporations. As of 2007, nonfinancial U.S. businesses received nearly US$1 trillion in new loans each year. Thus, research in the all-important loan market also has the potential to highlight the impact of CSR on the firm’s cost of capital.

Hypotheses

Our first hypothesis concerns the direct effect of institutional CSR on loan price and loan spread. By enabling the firm to build legitimacy with various actors, institutional CSR ensures the continued flow of resources to the firm. This buffers the firm from environmental uncertainties and fluctuations while increasing its odds of survival. Furthermore, by creating expectations of fairness and reciprocity, institutional CSR also fosters trust between the firm and various stakeholders. In the specific context of the loan market, banks can take greater confidence and trust that managers of firms engaging in institutional CSR will not fund investments with greater ex-post risks than what the loans were initially contracted for. This is a significant agency cost of debt which can be imposed on lenders (Jensen & Meckling, 1976) that is often difficult to observe. To the extent that there is lower likelihood of such behavior among firms that engage in institutional CSR, it would imply lower informational requirements for the bank not only ex-ante when assessing creditworthiness but also ex-post when administering and monitoring the loan.

In addition, we posit that when undertaking loan transactions, there may be an implicit transfer in the “burden of proof” to the firm which may further mitigate credit assessment and information gathering costs. When a firm with a socially responsible reputation seeks a loan, there may be a presumption of responsible borrowing on the part of the firm by the bank. This presumption may be justified not only on the grounds of prior reputation but also on the grounds that any default or violation of contractual terms may be particularly costly for such firms. Together, these arguments lead to the following overarching hypothesis:

Hypothesis 1: Ceteris paribus, higher levels of institutional CSR will be associated with lower levels of interest rates and loan spreads.

Hypothesis 1 captures the overall effects of institutional CSR in reducing loan spread and creating value. In addition to these overall effects, bank loans also provide multi dimensional information about the loan transaction both at the firm level and loan level which enable us to comprehensively test effects of institutional CSR, especially in terms of reducing transaction costs. Of these various factors, two central factors that have been extensively studied in the banking literature include firm leverage and loan maturity.

Strahan (1999) points out that one important feature by which banks control their amount of exposure is by specifying maturity of the loan.³ The typical yield curve suggests that as the maturity of a loan increases, the spread associated with the loan increases nonmonotonically. That is because the longer the maturity of the loan, the higher the risk of default due to unanticipated events. In the case of bank loans, the relationship tends to be complicated by the fact that longer maturity loans are usually associated with more extensive information gathering ex-ante by a bank regarding the uses of the loan and managerial capabilities (Barclay & Smith, 1995). Hence, typically they are issued to high quality firms, suggesting the term of the loan could also be negatively correlated with loan spread as the lower transaction costs outweigh default risk considerations and the savings are passed on by banks to these firms.

In contrast with these opposing predictions, we posit that firms engaging in institutional CSR are likely to obtain a lower spread than non-CSR firms for the same loan maturity due to the following reasons. Ceteris paribus the longer the maturity, the higher the monitoring costs incurred by the bank as it engages in information gathering over a longer time frame. In the case of firms engaging in institutional CSR, the need for information gathering may be reduced not just due to the relative stability likely to be enjoyed by these firms over a given time period, but also because of the expectation that these firms will borrow prudently commensurate with their capacity and their risk/return profile ex-ante. Moreover, for a given maturity, the costs of monitoring and ensuring that progress is being made toward repayment (e.g., by specifying various liquidity ratios, and ensuring they are satisfied, etc.) is also likely to be lower for firms engaging in institutional CSR. Hence, we suggest the following:

Hypothesis 2: Ceteris paribus, higher levels of institutional CSR will be associated with lower levels of interest rates and loan spread for a given loan maturity—CSR will reduce (strengthen) the positive (negative) impact of loan maturity on loan spread.

Another factor that plays a critical role in affecting loan spread at the firm level (as opposed to the loan level) is leverage. All else equal, higher leverage increases the probability of default and bankruptcy (Graham et al., 2008). As Fisher (1959) observes, when leverage is low (e.g., 5%) the value of the firm has to fall substantially for the firm to default on its debt (below 5%). Conversely, when leverage is high, default on debt can occur at higher values of the firm. Moreover, as leverage increases, the agency costs of debt are also likely to rise (Jensen & Meckling, 1976). When there is higher leverage, there is a greater probability that managers will transfer risks from equity shareholders to debtors, and that debtors will end up funding investments with greater risks than what they initially contracted. This suggests higher ex-post costs of monitoring with respect to these loans.

We argue that for a given level of leverage, firms engaging in institutional CSR are likely to be associated with lower default risks because of the stability they are likely to enjoy through stakeholder support. Furthermore, once again we expect that banks are also likely to incur lower ex-ante information gathering costs and ex-post monitoring costs with regard to these firms, and some of these benefits may eventually be passed on to the firm resulting in a lower price and interest rates. Hence,

Hypothesis 3: Ceteris paribus, higher levels of institutional CSR will be associated with lower levels of interest rates and loan spread for a given level of firm leverage-institutional CSR will weaken the impact of firm leverage in increasing loan spread.

Our hypotheses so far mainly point to the benefits conferred by institutional CSR in the loan market. However, as noted earlier, these effects can also dissipate rapidly if there are any negative events or acts of irresponsibility associated with the firm. The issue is complicated by the fact that developing institutional CSR as a valuable resource poses many unique managerial challenges.

Like other forms of reputation, the reputation generated by institutional CSR has a distinct path dependency component (Tang, Hull, & Rothenberg, 2012) wherein stakeholders are constantly evaluating new information regarding socially responsible activities against their prior assessments and expectations (Rhee & Haunschild, 2006). Consequently, to build an image of social responsibility, a firm needs to engage in investments in a sustained and coherent manner over time in various activities that matter to secondary stakeholders (Brammer & Millington, 2008; Roberts & Dowling, 2002). In recognition of the path dependency, Barnett (2007) introduces the notion of “stakeholder influence capacity.” The basic premise is that effective institutional CSR requires that a firm carefully develop the capability to identify and act on opportunities that improve stakeholder relations at various points in time (Tang et al., 2012). The difficulty lies in the fact that societies often tend to promulgate conflicting norms and values (e.g., pro choice and pro life; Meyer & Rowan, 1977). This makes it particularly challenging to find CSR opportunities that appeal to a wide group of secondary stakeholders without alienating any one group. In support of this point, Sen and Bhattacharya (2001) show that CSR can actually hurt consumer perceptions when there is lack of congruence between the issues that are important to consumers and the issues pursued by the firm. Conversely, Bhattacharya and Sen (2004) show that some firms have positioned themselves very effectively on the CSR platform compared with others, by carefully identifying the appropriate issues to focus on.

In addition to finding the issues that are of concern to stakeholders, institutional CSR also involves various organizational challenges. In particular, it entails a systematic inculcation of values throughout the hierarchy (Basu & Palazzo, 2008; Hillman & Keim, 2001) so that employees at all levels understand the importance of safeguarding the interests of multiple stakeholders. Moreover, engaging in institutional CSR has to be preemptive rather than reactive if the firm intends to build legitimacy with social actors and secondary stakeholders (Aguinis & Glavas, 2012). Investing in CSR after the occurrence of a negative event is unlikely to generate benefits, and at best may serve as damage control to prevent negative perceptions from spreading further, rather than providing added advantages.

The implication of these various arguments is that institutional CSR is more than simply a matter of engaging in activities that speak to secondary stakeholder concerns. It is a complex managerial capability which requires sustained effort before stakeholders accord any value to it and factor it into their transactions. Accordingly, firms that merely engage in these activities in the short run with the hope that they will capture benefits such as a lower spread are unlikely to enjoy any advantages. Equally important, firms that do not display sustained and coherent performance (i.e., high levels of corporate social performance [CSP]) in these activities and show evidence of negative events may not necessarily instill enough confidence in stakeholders to capture significant gains.

The literature has extensively discussed the effects of institutional CSR in mitigating the adverse consequences of a negative event (e.g., Godfrey et al., 2009). While we do not dispute these arguments, the point we wish to make is that such buffering effects may protect the firm from downside risks in the immediate vicinity of the event itself. However, in the longer term, these buffering effects will not necessarily mitigate the transaction costs incurred by stakeholders in subsequent transactions as they feel a renewed need to incur various search and monitoring costs. Hence,

Hypothesis 4: Ceteris paribus, the effects of institutional CSR in reducing loan spread will be strongest among firms that show sustained and coherent performance along this dimension, rather than among firms that display sporadic and varying institutional CSR performance.

Data and Method

Data Sources

To test the above hypotheses, we combined data from three sources: (a) DealScan, a Loan Pricing Corporation (LPC) database, which provides information on loan spreads and various loan characteristics; (b) the Kinder, Lydenburg, Domini, and Company (KLD) database, which provides information on social responsibility activities; and (c) Compustat, which provides information on various firm characteristics.

The time period for our study spans the years 1996 to 2005. This period saw a pronounced increase in CSR activities and disclosure by firms (Dhaliwal, Oliver Zhen, Tsang, & Yong George, 2011), as well as an increased awareness by the public of the importance of CSR, giving firms that have already been engaging in these activities a potential advantage. Corresponding to the time frame, the authors retrieved data on all commercial loans obtained by corporations listed in the DealScan database. The data in DealScan are gathered by LPC from the Securities and Exchange Commission (SEC) filings and from direct contacts with borrowers and lenders (Chava & Roberts, 2008). These data are particularly appropriate to study firms with public equity and debt (Drucker & Puri, 2005).

The unit of empirical analysis in our study is a loan. In DealScan, each individual loan is referred to as a “Facility.” For each Facility, DealScan contains a variable, All-In-Spread-Drawn, which is the interest rate for the loan. All-In-Spread-Drawn is the spread over LIBOR in basis points (100 basis points equal 1%) and is the variable usually used by studies examining the spread of bank loans (Drucker & Puri, 2005). In addition to loan spread, DealScan also lists other loan attributes including loan size, maturity, and the number of lenders. Although each loan has only one borrower, it may have multiple lenders due to syndicated lending.

DealScan does not provide data on borrowing firms’ financial and operating characteristics, and consequently we retrieve these data from Compustat. The specific data items and variables obtained are discussed below. Banks and other financial institutions are removed from the database, along with firms with missing information in Compustat. After retrieving data related to loans from DealScan, and after merging the data with Compustat and KLD social responsibility ratings (for the year prior to the loan issuance year), we were left with 2,480 observations spanning the period 1996-2005. The 2,480 loans corresponded to 1,785 unique firms, with the maximum number of loans taken out by a firm during this time period being six.

Dependent and Independent Variables

The dependent variable in our study is Loan Spread, which is the number of basis points over LIBOR for a particular loan. Information on the variable is drawn from DealScan.

In terms of independent variables, our main interest lies in constructing an appropriate measure of institutional CSR which pertains to a firm’s social responsibility activities specifically directed at secondary stakeholders. In addition, we are also interested in calculating a measure of a firm’s technical CSR directed at a firm’s primary stakeholders, and stakeholders operating in the technical environment. Measuring technical CSR is important for us as one of our primary objectives is to show that independent of this dimension, institutional CSR confers value by reducing risks and mitigating transaction costs.

We derive our measures of these two types of CSR activities based on the social responsibility ratings provided in KLD. KLD provides ratings along six main dimensions: Community Activities, Diversity, Employee Relations, Environmental Record, Products, and Corporate Governance. Along each of these dimensions, there are specifically defined components which constitute strengths and weaknesses. For example, on the community dimension, strengths are measured in terms of generous giving, innovative giving, support for housing, and support for education. Correspondingly, weaknesses are measured as activities that are associated with investment controversies, negative economic impact, tax disputes, and “other community concerns.” Mattingly and Berman (2006) conduct an exploratory factor analysis of these six dimensions including both the strengths and weaknesses (as opposed to studies such as Goss & Roberts, 2011, which separately conduct a factor analysis for strengths and weaknesses). Their analysis reveals that the six strengths and weaknesses in turn can be reduced to four latent factors. Based on the loadings of the individual dimensions, they argue that the four latent factors conceptually correspond to strengths and weaknesses related to institutional and technical CSR, respectively. Following their study, we measured Institutional Strengths as the sum of the strengths on community relations and diversity. Institutional Weakness was measured as the sum of the weaknesses along the environmental and community relations dimensions. Technical Strengths was measured as the sum of the strengths along employee relations, product and governance, while Technical Weakness was measured as the sum of the remaining weaknesses. Institutional CSR was measured as Institutional Strength minus Institutional Weakness. Technical CSR was measured as Technical Strengths minus Technical Weakness. All measures were constructed based on KLD data for the firm in the year prior to the issuance of the loan. Strictly speaking strengths and weaknesses cannot be subtracted from each other and treated as equivalent. We return to this point in our analyses below. The Mattingly and Berman (2006) and Godfrey et al. (2009) approaches provide a reliable way to disaggregate the KLD data in measuring primary and secondary stakeholder orientation, given the manner in which the KLD data were collected and its intrinsic structure.³

Hypothesis 1 predicts higher levels of Institutional CSR will be associated with lower levels of interest rates and loan spreads. Support for Hypothesis 1 implies that the coefficient of Institutional CSR will be negative and significant in the Loan Spread regression. Furthermore, we expect this negative relationship to hold even after controlling for the effects of Technical CSR. To test Hypothesis 2, we construct the variable Log Loan Maturity measured as the natural log of the duration of the loan in years. Hypothesis 2 proposes that institutional CSR will temper the positive effect, or conversely strengthen the negative effect of loan maturity on Loan Spread. Support for Hypothesis 2 implies that the interaction term Log Loan Maturity × Institutional CSR will be negatively signed and significant. To test Hypothesis 3, we construct the variable Leverage, defined as the ratio of long-term debt to total assets reported by the firm in the previous fiscal year. Hypothesis 3 posits that ceteris paribus while leverage will have a positive impact, institutional CSR will weaken this positive impact of leverage on loan spread. Thus, our expectation is that the interaction term Leverage × Institutional CSR will be negative and significant.

Hypothesis 4 posits that the risk-reducing benefits of Institutional CSR will accrue mainly to firms that show sustained superior performance along this dimension, rather than to firms that show mixed and sporadic performance. To test the hypothesis, we draw on prior literature which makes a distinction between firms that take a proactive approach toward social responsibility, versus firms that take a defensive approach and do what is required (Carroll, 1979; Cochran & Wood, 1984). Drawing on this idea, for analytical convenience, we divide our sample into four groups: Superior Performers, Conformers, Mixed Performers, and Poor Performers. Superior Performers are the subset of firms which have institutional strengths but at the same time no institutional weaknesses in KLD (i.e., Institutional Strengths is greater than 0, Institutional Weakness is equal to 0). These are firms that arguably went beyond expectations of institutional actors and secondary stakeholders in terms of fulfilling norms related to social responsibility (McGuire, Dow, & Argheyd, 2003) while instituting strong societal governance. We define Conformers as firms that have neither strengths nor weaknesses. This group can be viewed as taking a defensive approach (Basu & Palazzo, 2008) and doing what is expected, no more and no less. The third group, Mixed Performers, comprises of firms that have both strengths and weaknesses. These are firms that have invested in institutional CSR and have at the same time also experienced negative events. Thus, they can be viewed as firms engaging in CSR but showing mixed and sporadic CSP. Finally Poor Performers are the ones that have only institutional weaknesses and no institutional strengths. The classification scheme takes into account the idea that strengths and weaknesses are distinct, and are not simply mirror images of each other by examining the distribution of strengths and weaknesses. Hypothesis 4 implies that the risk-reducing and value-preserving effects of CSR would be relatively strongly evidenced among Superior Performers when compared with Mixed Performers given that Mixed Performers have negative events and acts of irresponsibility associated with them in KLD. Accordingly, we expect that Hypotheses 1, 2, and 3 would be strongly supported in the subgroup Superior Performers when compared with the group Mixed Performers.

Control Variables

We include a variety of controls in our models at the loan, firm, and industry level. These controls are used to account for the impact of nonprice features of the loan and various firm and industry characteristics independent of CSR on loan spread. At the loan level, we include Log Loan Size as a control, measured as the log of million dollars lent to the firm. Ceteris paribus, larger loans can potentially be more risky and be associated with a higher risk of default. Conversely, larger loans are also associated with lower administration costs for banks, which can be passed on to borrowers in the form of lower spread. The second control we use at the loan level is the Number of Lenders to capture the number of banks that are in the syndicate for a particular loan, as risky loans tend to have fewer members in the syndicate.

For the firm level controls, we include firm size measured as Log Sales as larger firms are likely to experience lower bankruptcy risk. In addition, we include Tangibility, measured as the ratio of property, plant, and equipment to total assets. Following previous studies (e.g., Bharath et al., 2011), Tangibility is used to control for a lower level of opaqueness and information asymmetry of a firm on loan price. Profitability, the ratio of earnings before interest, taxes, depreciation, and amortization (EBITDA) to total assets, is used to control for the impact that better accounting performance has on loan spread. Current Ratio, measured as the ratio of current assets to current liabilities, is used to control for higher short term liquidity, which can lower loan spreads. Next, we also included a dummy, Previous Relationship, to account for the presence of a prior loan between a focal firm and a lead bank. Evidence suggests prior relationships can lead to a lower spread due to lower monitoring costs and the bank’s familiarity with the firm (Bharath et al., 2011). Finally, we also included a control for the senior debt rating of the firm by S&P. If the senior debt rating is BBB and above it is considered investment grade, and the firm is considered to be a high quality borrower (Bharath et al., 2011). We therefore include the dummy Investment Grade to control for any determinants of risk that are not captured by our loan and firm characteristics.

At the industry level, we control for firms that are in industries that are prone to litigation (Dhaliwal et al., 2011). The dummy variable Litigation Risk is coded as 1 if the firm’s primary Standard Industrial Classification (SIC) reported in Compustat belongs to SIC codes 2833-2836, 3570-3577, 3600-3674, 5200-5961, and 7370. The SIC codes pertain to firms falling in medicinal chemicals, pharmaceuticals, biological products such as vaccines, computers, electronics and electrical equipment, the retail sector, and services such as programming and data processing. High litigation risk industries are associated with a high cost of capital, including higher loan spreads. We also include controls for the industry of the firm, and year dummies for each year between 1996 and 2005.

Results

Summary Statistics

Table 1 presents the means and correlations of the variables used in our analysis. With the exception of the institutional and technical CSR variables and the variables for which we took logs, all other variables are winsorized at the 0.025 level. The mean spread of loans taken out by the firms in our sample is 98.2 basis points. The average Institutional CSR score for the 2,480 loans taken out by the 1,785 firms is 0.44. The Institutional CSR score ranged from −6 to a maximum of +10. The average Technical CSR score on the contrary was negative at −0.72. Table 1 further shows that the average of Institutional Strengths in our sample was 1.04, while the average of Institutional Weakness was .61.

Table 1.

Descriptive Statistics and Intercorrelations.

Variable	M	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1. Spread	98.20	93.24	1.00
2. Institutional CSR	0.44	1.85	−.18	1.00
3. Technical CSR	−0.72	1.59	−.06	.07	1.00
4. Institutional Strength	1.04	1.54	−.23	.79	.00	1.00
5. Institutional Weakness	0.61	1.14	−.02	−.55	−.11	.07	1.00
6. Log Loan Maturity	0.86	0.81	.20	−.05	−.01	−.14	−.12	1.00
7. Log Loan Size	19.87	1.21	−.28	.07	−.15	.27	.24	−.05	1.00
8. Number of Lenders	35.45	28.25	.06	.01	.01	−.02	−.04	.05	−.02	1.00
9. Leverage	0.28	0.16	.19	−.10	−.07	−.02	.14	−.05	.22	.02	1.00
10. Log Sales	8.10	1.38	−.39	.14	−.17	.41	.33	−.27	.64	−.09	.07	1.00
11. Market to Book	1.96	1.00	−.25	.32	.04	.23	−.21	−.09	.01	−.01	−.21	.05	1.00
12. Tangibility	0.35	0.21	−.08	−.22	.03	−.08	.25	−.04	.09	.01	.22	.04	−.23	1.00
13. Profitability	0.15	0.07	−.34	.21	.08	.16	−.13	−.02	.07	.03	−.13	.09	.61	.04	1.00
14. Current Ratio	1.62	0.92	.18	−.01	.02	−.16	−.20	.19	−.32	.04	−.32	−.38	.05	−.35	−.03	1.00
15. Previous Relationship	0.34	0.47	−.12	.04	−.05	.12	.10	−.16	.23	−.02	.10	.28	.04	.04	.02	−.13	1.00
16. Litigation	0.23	0.42	−.02	.21	−.07	.12	−.17	.00	.00	.04	−.18	.18	.17	−.06	.10	.13	.04	11.00
17. Investment Grade	0.35	0.48	−.39	.12	−.09	.22	.10	−.19	.28	−.08	−.04	.41	.22	.02	.17	−.19	.17	−.01

Note. Observations = 2,480 loans which appear on the KLD list. CSR = corporate social responsibility; KLD = Kinder, Lydenburg, Domini, and Company.

In terms of loan characteristics, the average loan size in our sample was US$535 million, while the average maturity was 37.4 months. As a percentage of assets, the average loan was about 16.6% of assets with a median of 10.5%. These figures indicate that the loans were significant from the firms’ standpoint and from the bank’s standpoint as well. Thus, there were strong incentives for banks to thoroughly assess creditworthiness before administering them as well as to monitor them ex-post, while potentially incurring significant transaction costs.

In terms of firm characteristics, firms in our sample had sales of US$7.8 billion indicating they were fairly large. Nearly 34% of the loans are transactions where there is a prior relationship between the borrower and the lead bank, indicating a fair prevalence of relationship banking. Table 1 also shows the Pearson–product moment correlations between the different variables. Of particular interest is the correlation between Institutional CSR and Loan Spread. The correlation is significant and negative at −0.18 indicating preliminary support for Hypothesis 1. The correlation coefficient between Technical CSR and Loan Spread is also negative and significant, although the coefficient was smaller at −0.06. All remaining correlations are within the acceptable range indicating no serious problems of multicollinearity.

Regression Results

Table 2, Model 1, presents the regression with Loan Spread as the dependent variable, and Institutional CSR as the independent variable. We present our estimates using clustered standard errors at the firm level. Institutional CSR has a negative and significant impact (p < .05), indicating support for Hypothesis 1. The coefficient of Institutional CSR indicates that a 1-point increase decreased loan spread by 3.65 basis points. In our sample of firms, with a range of Institutional CSR from −6 to 10, the impact is a lower spread by 58 basis points or 0.58% between the most socially responsible and least socially responsible firm.

Table 2.

Loan Spread Regression.

Independent variables	Model 1^a	Model 2^a
Institutional CSR	−3.656* (1.65)
Technical CSR	−3.205* (1.44)	−3.151* (1.46)
Institutional Strength		−1.019 (1.85)
Institutional Weakness		8.490** (2.98)
Log Loan Maturity	11.01*** (2.69)	11.24*** (2.67)
Log Loan Size	−5.935* (2.67)	−6.058* (2.66)
Number of Lenders	0.0399 (0.07)	0.0378 (0.07)
Leverage	129.0*** (20.22)	129.8*** (20.21)
Log Sales	−13.52*** (3.18)	−16.16*** (3.27)
Market to Book	−0.698 (3.18)	−0.965 (3.15)
Tangibility	−28.84^† (15.67)	−32.26* (15.76)
Profitability	−284.0*** (48.72)	−279.2*** (48.61)
Current Ratio	3.578 (3.69)	3.516 (3.63)
Previous Relationship	−1.387 (3.69)	−1.089 (3.67)
Litigation	12.13 (7.92)	13.01^† (7.85)
Investment Grade	−33.64*** (6.07)	−32.82*** (6.08)
Constant	345.6*** (53.08)	367.2*** (53.23)
Industry Controls?	Yes	Yes
Year Controls?	Yes	Yes
R ²	.39	.39
Adjusted R²	.38	.38
F	22.41	21.81

Note. Robust standard errors are in parentheses. CSR = corporate social responsibility.

N = 2,480.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Table 2, Model 1, also shows that Technical CSR is significant (p < .05) and has a negative impact on Loan Spread. The noteworthy result of Model 1 is that institutional CSR directed at secondary stakeholders has a significant effect independent of the effects of CSR directed at primary stakeholders. This result is important as it establishes the benefits of CSR investments aimed at secondary stakeholders. A test of equality of coefficients revealed that there is no significant difference between the coefficients of Institutional CSR and Technical CSR. As a robustness check of these results, we also conducted analyses where we constructed our Institutional CSR and Technical CSR variables using 3-year prior data from KLD. Our results (not reported) remained the same. Both CSR variables continued to have a negative impact on Loan Spread in these regressions.

The control variables in the models in Table 2, Model 1, behave by and large as expected. Loans with longer maturity on average are associated with higher spreads, likely because longer maturity exposes a bank to uncertainty and risk of default over a longer time period. Leverage is positively associated with loan spread, as debt makes a firm a more risky investment (Fisher, 1959).

As a next step, we ran the regression equation in Model 1 after disaggregating the Institutional CSR variable into Institutional Strengths and Institutional Weakness, respectively. Table 2, Model 2, presents the results. The results of this model are less sanguine than those presented in Model 1. Institutional Strengths has an insignificant impact on Loan Spread, whereas Institutional Weakness is positive and significant (p < .01). This suggests that the effects of institutional CSR observed in Model 2 are mainly due to the avoidance of concerns and negative events, rather than due to proactive investments in these activities, a result that is consistent with recent studies (e.g., Jayachandran et al., 2013). In other words, while engaging in institutional CSR does not create value for a firm and its stakeholders, having concerns in this area significantly raises the riskiness of the firm as well as the costs of transacting for the parties concerned.⁴ These results at first glance appear to call into question the benefits of secondary stakeholder activities. Jayachandran et al. (2013) similarly find that while strengths along the product social performance (i.e., the degree to which a product focused actions such as development and marketing are consistent with societal obligations) and environmental social performance dimensions do not improve firm performance (as measured by Tobin’s Q), weaknesses along these dimension hurt performance.

Given these results, we decided to undertake further investigation before moving on to testing our remaining hypotheses. Earlier, we argued that from a conceptual standpoint it is necessary to distinguish between four groups of firms based on the level and type of CSR engagement (Carroll, 1979): Superior Performers, Conformers, Mixed Performers, and Poor Performers. This distinction is important because firms that are sustained and coherent performers in terms of Institutional CSR are likely to be viewed favorably by stakeholders, whereas firms that show mixed and sporadic performance are likely to cause stakeholders to incur significant costs when transacting. In our sample, 29.8% of the loans were taken out by firms that were Superior Performers, 40% were Conformers, 15.8% were Mixed Performers, and the remaining 14.4% were Poor Performers.

Table 3 provides the results of the loans spread regressions with these four groups. Table 3, Models 3 to 6 include a dummy variable for each group added one at a time. In these columns, the coefficient of the dummy variable and its significance highlight whether the group experienced a higher or lower spread compared with the remaining three groups. Model 7 adds the three dummy variables Superior Performers, Conformers, and Poor Performers to the model simultaneously, with Mixed Performers serving as the comparison group.

Table 3.

Loan Spread Regression for Different Categories of Institutional CSR.

Independent variables	Model 3	Model 4	Model 5	Model 6	Model 7
Superior Performers	−13.37** (5.10)				−25.49*** (7.14)
Conformists		5.20 (5.16)			−20.47** (7.55)
Mixed Performers			16.53* (7.04)
Poor Performers				15.45^† (9.20)	0.31 (10.70)
Technical CSR	−3.280* (1.45)	−3.678* (1.43)	−3.457* (1.45)	−3.426* (1.42)	−2.925* (1.44)
Log Loan Maturity	10.90*** (2.69)	11.22*** (2.72)	10.90*** (2.70)	11.27*** (2.69)	10.88*** (2.66)
Log Loan Size	−6.058* (2.71)	−6.254* (2.74)	−6.303* (2.74)	−6.099* (2.70)	−6.140* (2.67)
Number of Lenders	0.03 (0.07)	0.04 (0.07)	0.03 (0.06)	0.04 (0.07)	0.03 (0.06)
Leverage	131.0*** (20.33)	129.6*** (20.63)	132.2*** (20.56)	128.5*** (20.09)	131.6*** (19.99)
Log Sales	−13.52*** (3.16)	−14.54*** (3.21)	−15.28*** (3.31)	−14.28*** (3.08)	−16.14*** (3.17)
Market to Book	−1.00 (3.16)	−2.31 (3.15)	−2.16 (3.09)	−1.47 (3.11)	−1.01 (3.15)
Tangibility	−27.04^† (15.46)	(24.48) (15.54)	−26.47^† (15.48)	−27.08^† (15.55)	−32.63* (15.40)
Profitability	−286.7*** (48.58)	−287.0*** (48.65)	−285.4*** (48.81)	−285.6*** (48.27)	−280.9*** (48.16)
Current Ratio	3.52 (3.72)	4.34 (3.73)	3.98 (3.70)	4.11 (3.69)	3.64 (3.68)
Previous Relationship	−1.77 (3.66)	−1.13 (3.72)	−1.24 (3.70)	−1.31 (3.68)	−1.33 (3.63)
Litigation	10.17 (7.78)	9.39 (7.81)	8.68 (7.88)	11.38 (7.69)	11.76 (7.74)
Investment Grade	−34.67*** (6.14)	−33.71*** (6.18)	−34.78*** (6.14)	−33.34*** (6.03)	−34.49*** (5.95)
Constant	353.9*** (52.12)	362.3*** (55.66)	366.8*** (54.47)	352.9*** (54.71)	390.3*** (53.09)
Industry Controls?	Yes	Yes	Yes	Yes	Yes
Year Controls?	Yes	Yes	Yes	Yes	Yes
R ²	.38	.38	.38	.38	.39
Adjusted R²	.377	.373	.376	.376	.38
F	23.44	22.71	23.88	22.58	22.2

Note. Robust standard errors are in parentheses. CSR = corporate social responsibility.

N = 2,480.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Table 3, Model 3, shows that that the dummy variable Superior Performers had a significant negative coefficient and is associated with lower interest rates (p < .01). The coefficient indicates that loans taken out by firms in this group, which comprises of firms with Institutional Strengths > 0 and Institutional Weakness = 0, on average enjoyed a lower spread of about 13 basis points compared with the rest of the sample. The noteworthy point here is that proactively engaging in institutional CSR did contribute to lower spread, and the negative effects observed in Table 2, Model 1, are not being driven merely by higher spreads experienced by firms with weaknesses and concerns. Model 4 presents the results with the dummy variable Conformists added. The coefficient of this variable is insignificant, indicating that these firms did not enjoy a significantly lower spread compared with the remaining groups. Model 5 presents the results for Mixed Performers. The coefficient of Mixed Performers is positive and significant (p < .05). Loans by this group of firms on average experienced a higher spread of 16.5 basis points when compared with the rest of the sample. Combined with the results in Model 3, these results provide preliminary support for Hypothesis 4 and our argument that to derive the benefits of institutional CSR firms need to exhibit sustained superior performance, rather than mixed performance in terms of their secondary stakeholder activities. Model 6 presents the results for the subsample of Poor Performers. The coefficient for this group is also positive and significant (p < .10) with a value of 15.4. Finally Model 7 presents the results after adding the three dummies with Mixed Performers as the comparison group. As shown Superior Performers enjoyed significantly lower spreads by nearly 25 basis points (p < .001). Interestingly, even Conformists experienced lower spreads of about 20 basis points (p < .01). The Poor Performers, however, did not experience significantly lower spreads. It should be noted that the difference in coefficients between Superior Performers and Conformists is insignificant in this regression. These results, however, do not take into consideration heterogeneous levels of institutional CSR among Superior Performers, and are essentially group mean effects. We return to this point in the following section. As a robustness check, we also conducted analyses where we constructed the four dummy variables using there year prior KLD data as before. In these analyses, the Superior Performers dummy continued to have a negative impact, paralleling the results in Table 3, Model 3. These results further point to the importance of strengths in reducing interest rates and loan spreads.

Table 4 presents the results of our interaction hypotheses. In these models, we add the interaction terms to the basic model relating Institutional CSR to Loan Spread presented in Model 1 of Table 2. Model 8 in Table 4 presents the results after adding the interaction terms Institutional CSR × Log Loan Maturity used to test Hypothesis 2, and Institutional CSR × Leverage used to test Hypothesis 3. The interaction terms were formed after centering the interacting variables at the mean prior to multiplying them to reduce effects of multicollinearity. In this regression, the main term of Log Loan Maturity is positive and significant, whereas the interaction term of Institutional CSR × Log Loan Maturity is negative and significant (p < .05). This supports Hypothesis 2, and our argument that institutional CSR will mitigate the effect of loan maturity in enhancing interest rates and spread. Column 1 also shows that while the main term of Leverage is positive and significant, the interaction term Institutional CSR × Leverage is negative and significant (p < .001). This supports Hypothesis 3 and our proposition that institutional CSR will reduce the positive effects of leverage on interest rates and spread. More broadly, these findings support the idea that holding loan and firm characteristics constant, firms engaging in secondary stakeholder activities and institutional CSR will enjoy lower spreads not only because of lower default risk but also because banks are likely to incur lower costs in assessing creditworthiness, and in monitoring these firms ex-post.

Table 4.

Loan Spread Regression With Interaction Effects.

Independent variables	Model 8	Model 9	Model 10	Model 11
Institutional CSR	−3.126* (1.46)	−4.675* (2.14)	−3.172^† (1.68)	−3.182* (1.45)
Technical CSR	−3.380* (1.38)	−4.331* (1.78)	−2.910^† (1.65)	−3.433* (1.36)
Log Loan Maturity	10.61*** (2.64)	13.41** (4.89)	9.539** (3.40)	9.970*** (2.69)
Log Loan Size	−5.731* (2.61)	−1.65 (4.12)	−5.825^† (3.03)	−5.928* (2.60)
Number of Lenders	0.04 (0.06)	0.05 (0.09)	−0.02 (0.07)	0.04 (0.06)
Leverage	131.8*** (19.50)	128.0*** (36.89)	119.6*** (20.30)	130.9*** (19.10)
Log Sales	−14.83*** (3.02)	−19.24*** (3.50)	−19.32*** (3.08)	−14.80*** (2.98)
Market to Book	−1.44 (3.06)	−1.46 (3.90)	−2.43 (3.16)	−1.22 (3.05)
Tangibility	−30.43* (15.41)	−22.15 (20.87)	−33.93* (16.36)	−32.10* (15.46)
Profitability	−272.2*** (47.14)	−278.0*** (65.75)	−257.1*** (47.68)	−271.7*** (46.78)
Current Ratio	2.262 (3.64)	8.254 (6.42)	0.331 (3.86)	1.816 (3.62)
Previous Relationship	−2.449 (3.62)	−6.723 (4.23)	−4.57 (4.15)	−2.649 (3.60)
Litigation	9.944 (7.67)	7.612 (11.89)	16.33^† (8.65)	9.738 (7.71)
Investment Grade	−32.46*** (5.78)	−18.99* (8.09)	−24.07*** (6.38)	−32.37*** (5.74)
Institutional CSR × Log Maturity	−2.572* (1.25)	−3.372* (1.62)	−2.352^† (1.42)	−2.452* (1.24)
Institutional CSR × Log Leverage	−40.58*** (10.14)	−27.40* (12.63)	−27.10** (8.51)	−39.46*** (10.07)
Technical CSR × Log Maturity				−2.397^† (1.34)
Technical CSR × Log Leverage				−12.23 (10.43)
Constant	348.9*** (51.21)	259.4*** (75.76)	382.4*** (57.66)	355.3*** (51.81)
Industry Controls?	Yes	Yes	Yes	Yes
Year Controls?	Yes	Yes	Yes	Yes
n	2,480	740	1,732	2,480
R ²	.399	.495	.433	.401
Adjusted R²	.391	.472	.422	.393
F	24		21.18	23.71

Note. Robust standard errors are in parentheses. CSR = corporate social responsibility.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Hypothesis 4 posits that the risk-reducing effects of institutional CSR will be strongest among firms that show superior performance along this dimension rather than among firms with mixed performance. In part, this argument was supported by the positive coefficient of Mixed Performers in Table 3. As a further test of Hypothesis 4, we also divided our full sample into subsamples of Superior Performers and Mixed Performers and re estimated our regression with the interaction terms separately in the two subsamples. The expectation was that the interaction hypotheses will be strongly supported in the sample of Superior Performers, but will be weaker in the case of Mixed Performers. Model 9 presents the results for the subsample of Superior Performers. This regression essentially comprises of those firms where institutional CSR is greater than 0. The main term Institutional CSR is negative and significant (p < .05) and both interaction terms Institutional CSR × Log Loan Maturity and Institutional CSR × Leverage are also negative and significant (p < .05). In contrast, in the regression for Mixed Performers (results not reported), neither the main term nor the interaction terms were significant indicating that effects of institutional CSR were strongest among firms that showed sustained and coherent performance.

As a further test of the argument that developing strengths has an effect on reducing risks and spread, we also conducted our regression in a subsample of firms comprising both Superior Performers and Conformists. This analysis is important as it gives an idea of the effects of institutional CSR as it increases from 0 (rather than just being positive and greater than 0 as in model 9). Model 10 presents the results. Institutional CSR has a slightly weaker negative impact in this subsample (p = .059). In addition, the interaction between Institutional CSR × Log Loan Maturity is also weaker (p = .09) while staying negative. The interaction between Institutional CSR × Leverage is strongly negative and significant.

Next, we also added the interaction terms Technical CSR × Log Loan Maturity and Technical CSR × Leverage to our full sample to see whether our hypothesized interaction effects with Institutional CSR hold even after adding these interaction terms. As shown in Model 11 of Table 4, all interaction terms along with the main term of institutional CSR remain negative and significant In contrast, while Technical CSR × Log Loan Maturity was marginally significant and negative, Technical CSR × Leverage insignificant. These results once again suggest that the effects of institutional CSR are strong enough and persist even after for controlling for the effects of technical CSR.

Discussion

This article examines the effect of institutional CSR, or CSR directed at secondary stakeholders, on the bank loan market. The loan market is an interesting context because it allows us to examine not only to what extent CSR directed at secondary stakeholders reduces risks inherent in the firm but also transactions costs (because banks incur substantial costs in terms of assessing creditworthiness and monitoring the utilization of their loans). Theoretically our argument for the effects of institutional CSR in reducing interest rates and loan spread draws on two principles. First, institutional CSR helps build legitimacy with various stakeholders. In this regard, it is particularly effective compared with technical CSR as it ensures conformity with societal norms, and institutes “societal governance” (Brammer et al., 2012). Second, sustained institutional CSR activities also reassure stakeholders that the firm is committed to norms of fairness and reciprocity in its transactions. From a stakeholder standpoint, the effect it has is that it reduces search costs and ex-ante costs of writing contracts, as well as ex-post monitoring costs. These two effects together result in institutional CSR having a reducing impact on loan spread, as value is created simultaneously for the bank as well as the firm.

Countering these effects, the authors argued that the benefits of institutional CSR are derived only as long as the firm does not experience any negative events along this dimension. While prior institutional CSR and secondary stakeholder activities can buffer the firm from such events in the short run and limit its downside risk (Godfrey et al., 2009), these events will nevertheless cause stakeholders to reevaluate the firm in subsequent exchanges leading to enhanced search costs and monitoring costs. The cross-sectional prediction these arguments suggest is that only firms with consistent institutional CSR activities and superior CSP along this dimension will enjoy lower spreads in their loan transactions.

Our empirical evidence supports these various arguments: We find institutional CSR has a negative impact on loan spread. This result was observed even after controlling for the effects of technical CSR directed at primary stakeholders. Since the inception of stakeholder theory, practitioners and scholars have called into question whether CSR confers any benefits upon a firm and its stakeholders. While the debate remains vibrant, arguably the benefits of technical CSR directed at primary stakeholders are less questionable, given that these stakeholders engage in direct exchanges with the firm and have immediate power. In contrast, there is far more ambiguity concerning whether institutional CSR directed specifically at secondary stakeholders confers benefits, because secondary stakeholders do not engage in direct exchanges and hence lack the immediacy that primary stakeholders exert. Against this backdrop, arguably it is investments related to secondary stakeholders that are a particularly critical test of whether CSR has value in capitalistic economies, and whether it should be undertaken for instrumental reasons rather than for purely normative purposes (Donaldson & Preston, 1995; Jones, 1995; McWilliams & Siegel, 2001). Our results support the view that institutional CSR does indeed create value for stakeholders, and concomitantly for the firm, independent of the effects of technical CSR.

In this regard, it is also worth noting that recently there has been considerable interest in terms of understanding how social responsibility impacts the cost of capital of firms (Dhaliwal et al., 2011; Sharfman & Fernando, 2008) and their financial constraints (Cheng, Ioannou, & Serafeim, 2014). This research differs from prior work which has looked at socially responsible investing as a way to understand whether investors value social responsibility (Barnett & Salomon, 2006). Our findings complement these studies. Thus, while Dhaliwal et al. (2011) show the effects on the cost of equity, Sharfman and Fernando (2008) demonstrate the impact in bond markets, and Cheng et al. (2014) show that CSR eases capital constraints, we provide another piece of the puzzle by looking at the loan market. Where we differ is that we specifically highlight the effects of institutional CSR and technical CSR separately on the cost of capital.

Our study also extends these studies in another important respect by highlighting the moderating effects of CSR with respect to leverage and loan maturity, two central factors studied in the loan literature that have been consistently found to increase interest rates and loan spread. We view these moderating effects as supportive of our arguments that institutional CSR lowers transaction costs and creates value. These moderating effects have also not been examined in other studies looking at bank loans and CSR such as Goss and Roberts (2011). In our context, the argument was that for the same level of perceived riskiness in terms of leverage and maturity, banks will incur lower costs in terms of assessing creditworthiness and in terms of ex-post monitoring in the case of CSR firms. Supporting these arguments, we found that while loan maturity had a positive impact on spread, institutional CSR tempered this positive impact to some extent and reduced the effects of maturity in enhancing spread. Similarly, we found that while leverage increases loan spread, institutional CSR mitigates the positive impact of leverage on spread. The latter result is particularly noteworthy because since Fisher (1959), the literature has consistently found leverage to be positively associated with loan spread and interest rates. Our study, however, indicates that the impact of leverage varies depending upon the institutional CSR of the firm.

Apart from the direct and moderating effects, we also contribute by highlighting the contingent effects of institutional CSR. As noted earlier, we found that only firms that were superior and consistent performers (i.e., firms with strong CSP) captured the associated risk-reducing benefits. In contrast, firms that displayed mixed performance (or inconsistent CSP) and had both strength and concerns experienced higher spreads, and did not seem to be associated with lower risks (cf. Brammer & Millington, 2008). These results attest to the growing recognition in the literature that institutional CSR is more than a matter of simply investing in activities related to secondary stakeholders (Margolis & Walsh, 2003). Rather, it is an activity that needs to be managed judiciously with careful attention to various aspects, including choosing issues carefully, and inculcating a culture within the firm where such activities are ingrained. Interestingly, our analysis also reveals that simply conforming to expectations of socially responsible behavior (i.e., doing neither good nor bad) is less detrimental than demonstrating mixed performance, as evidenced by the negative significant coefficient of Conformists in the regression in Table 3, Model 7, with Mixed Performers as the comparison group.

Future research could build on this study in various ways. First, the authors relied on secondary data from the KLD database. Although the measures we used are well established, future studies may strengthen the analysis and results by supplementing it for example with specific CSR financial investment data culled from social responsibility reports. Second, future studies could examine in what manner a firm attained a particular CSR status, such as Mixed Performers, instead of examining cross-sectional effects as we did. As noted earlier, prior literature points both to the buffering effects of reputation assets such as CSR when there are negative events (Ahluwalia, Burnkrant, & Unnava, 2000), as well as to the amplification of negative events due to expectancy violation. Given these opposing predictions, we think there is potential to disentangle these effects in greater depth by looking at the spreads of firms that had preexisting strengths and experienced negative events subsequently, as opposed to firms that had preexisting concerns and then added strengths subsequently. If the impact of concerns is weaker in the first group, it would provide strong support for the buffering effects of CSR when there are negative events.

Third, in the context of bank loans, future studies could also broaden the analysis by examining the impact of social responsibility on nonprice loan features. In our study, we concentrated on loan maturity and spread. Future studies could examine whether other covenants in loan contracts, such as maintaining minimum liquidity and financial ratios, are as rigorous for CSR firms as they are for non-CSR firms. This multi dimensional approach would help further establish the role of CSR in terms of reducing transaction costs. Fourth, future studies could examine in depth whether the firm at the other end of the transaction, in our case the bank, is influenced by transacting with CSR firms. In our study, we posited that in return for various benefits, banks will provide loans at lower spreads to firms engaging in institutional CSR. Hence, it would be instructive to examine whether banks’ performance is enhanced by lending to CSR firms, for example, through a lower percentage of defaults and bad loans. Furthermore, it would also be useful to study whether banks’ response to CSR varies across countries, given that institutions play a key role in shaping the value of socially responsible behavior (Aguilera & Jackson, 2003; Gardberg & Fombrun, 2006; Matten & Moon, 2008; Scherer & Palazzo, 2007). It is possible that effect sizes and the relationships we obtained in our analyses are weaker in the U.S. lending market compared with other countries which have a more socially proactive orientation. These issues notwithstanding, this study takes a further step toward demonstrating the effects of secondary stakeholder activities and institutional CSR in creating value, and considerable research potential remains in this regard.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Bill Francis (PhD, University of Toronto) is the Warren H. and Pauline U. Bruggeman distinguished professor of finance at Rensselaer Polytechnic Institute, Lally School of Management. He conducts his research on issues in initial public offerings, corporate restructurings, and international asset-pricing. He has published in journals such as Journal of Financial & Quantitative Analysis, Journal of Financial Economics, Journal of International Money and Finance, Journal of Macroeconomics, and Review of Financial Studies.

Pamela Harper (PhD, Rensselaer Polytechnic Institute) is an assistant professor of management at Marist College. She focuses her research on social issues in management and firm governance. Her articles have appeared in such journals as The American Journal of Management and R&D Management. Prior to academia, she spent 11 years with General Electric in various management assignments in the United States, Europe, and Asia.

Shyam Kumar (PhD, University of Illinois) is an associate professor of management and organization, Rensselaer Polytechnic Institute, Lally School of Management. His research interests are in the areas of firm boundaries, and alliances and acquisitions. His research has appeared in such journals as Advances in Strategic Management, Journal of Business Research, Managerial and Decision Economics, Organization Science, and Strategic Management Journal.

References

Aguilera

R. V.

Jackson

(2003). The cross-national diversity of corporate governance: Dimensions and determinants. Academy of Management Review, 28, 447-465.

Aguilera

R. V.

Rupp

D. E.

Williams

C. A.

Ganapathi

(2007). Putting the S back in corporate social responsibility: A multilevel theory of social change in organizations. Academy of Management Review, 32, 836-863.

Aguinis

Glavas

(2012). What we know and don’t know about corporate social responsibility: A review and research agenda. Journal of Management, 38, 932-968.

Ahluwalia

Burnkrant

R. E.

Unnava

H. R.

(2000). Consumer response to negative publicity: The moderating role of commitment. Journal of Marketing Research, 37, 203-214.

Bansal

Roth

(2000). Why companies go green: A model of ecological responsiveness. Academy of Management Journal, 43, 717-736.

Barclay

M. J.

Smith

C. W.

(1995). The maturity structure of corporate debt. Journal of Finance, 50, 609-631.

Barnett

M. L.

(2007). Stakeholder influence capacity and the variability of financial returns to corporate social responsibility. Academy of Management Review, 32, 794-816.

Barnett

M. L.

Salomon

R. M.

(2006). Beyond dichotomy: The curvilinear relationship between social responsibility and financial performance. Strategic Management Journal, 27, 1101-1122.

Basu

Palazzo

(2008). Corporate social responsibility: A process model of sensemaking. Academy of Management Review, 1, 122-136.

10.

Baum

J. A. C.

Oliver

(1991). Institutional linkages and organizational mortality. Administrative Science Quarterly, 36, 187-218.

11.

Bharath

S. T.

Dahiya

Saunders

Srinivasan

(2011). Lending relationships and loan contract terms. Review of Financial Studies, 24, 1141-1203.

12.

Bhattacharya

C. B.

Sen

(2004). Doing better at doing good: When, why and how consumers respond to corporate social initiatives. California Management Review, 47(1), 9-24.

13.

Brammer

Jackson

Matten

(2012). Corporate social responsibility and institutional theory: New perspectives on private governance. Socio-Economic Review, 10, 3-28.

14.

Brammer

Millington

(2008). Does it pay to be different? An analysis of the relationship between corporate social and financial performance. Strategic Management Journal, 29, 1325-1343.

15.

Campbell

J. L.

(2007). Why would corporations behave in socially responsible ways? An institutional theory of corporate social responsibility. Academy of Management Review, 32, 946-967.

16.

Carroll

A. B.

(1979). A three-dimensional conceptual model of corporate social performance. Academy of Management Review, 4, 497-505.

17.

Chava

Roberts

(2008). How does financing impact investment? The role of debt covenants. Journal of Finance, 63, 2085-2121.

18.

Cheng

Ioannou

Serafeim

(2014). Corporate social responsibility and access to finance. Strategic Management Journal, 35, 1-23.

19.

Cochran

P. L.

Wood

R. A.

(1984). Corporate social responsibility and financial performance. Academy of Management Journal, 27, 42-56.

20.

Cornell

Shapiro

(1987). Corporate stakeholders and corporate finance. Financial Management, 16, 5-14.

21.

Dhaliwal

D. S.

Oliver Zhen

Tsang

Yong George

(2011). Voluntary nonfinancial disclosure and the cost of equity capital: The initiation of corporate social responsibility reporting. Accounting Review, 86, 59-100.

22.

DiMaggio

P. J.

Powell

W. W.

(1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48, 147-160.

23.

Donaldson

Preston

(1995). The stakeholder theory of the modern corporation: Concepts, evidence, and implications. Academy of Management Review, 20, 65-91.

24.

Drucker

Puri

(2005). On the benefits of concurrent lending and underwriting. Journal of Finance, 60, 2763-2799.

25.

Fisher

(1959). Determinants of risk premiums on corporate bonds. Journal of Political Economy, 67, 217-237.

26.

Freeman

R. E.

(1994). The politics of stakeholder theory. Business Ethics Quarterly, 4, 409-421.

27.

Gardberg

N. A.

Fombrun

C. J.

(2006). Corporate citizenship: Creating intangible assets across institutional environments. Academy of Management Review, 31, 329-346.

28.

Godfrey

P. C.

Merrill

C. A.

Hansen

J. M.

(2009). The relationship between corporate social responsibility and shareholder value: An empirical test of the risk management hypothesis. Strategic Management Journal, 30, 425-445.

29.

Goss

Roberts

G. S.

(2011). The impact of corporate social responsibility on the cost of bank loans. Journal of Banking and Finance, 35, 1794-1810.

30.

Graham

J. R.

Qiu

(2008). Corporate misreporting and bank loan contracting. Journal of Financial Economics, 88(3), 44-61.

31.

Hillman

A. J.

Keim

G. D.

(2001). Shareholder value, stakeholder management, and social issues: What’s the bottom line? Strategic Management Journal, 22, 125-139.

32.

Hull

C. E.

Rothenberg

(2008). Firm performance: The interactions of corporate social performance with innovation and industry differentiation. Strategic Management Journal, 29, 781-789.

33.

Jayachandran

Kalaignanam

Eilert

(2013). Product and environmental social performance: Varying effect on firm performance. Strategic Management Journal, 34, 1255-1264.

34.

Jensen

M. C.

Meckling

W. H.

(1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305-360.

35.

Jones

T. M.

(1995). Instrumental stakeholder theory: A synthesis of ethics and economics. Academy of Management Review, 20, 404-437.

36.

King

Lenox

(2002). Exploring the locus of profitable pollution reduction. Management Science, 48, 289-299.

37.

Kock

C. J.

Santaló

Diestre

(2012). Corporate governance and the environment: What type of governance creates greener companies? Journal of Management Studies, 49, 492-514.

38.

Lev

Petrovits

Radhakrishnan

(2010). Is doing good for you? How corporate charitable contributions enhance revenue growth. Strategic Management Journal, 31, 182-200.

39.

Margolis

J. D.

Walsh

J. P.

(2003). Misery loves companies: Rethinking social initiatives by business. Administrative Science Quarterly, 48, 268-305.

40.

Matten

Moon

(2008). “Implicit” and “explicit” CSR: A conceptual framework for a comparative understanding of corporate social responsibility. Academy of Management Review, 33, 404-424.

41.

Mattingly

J. E.

Berman

S. L.

(2006). Measurement of corporate social action: Discovering taxonomy in the Kinder Lydenburg Domini ratings data. Business & Society, 45, 20-46.

42.

McGuire

Dow

Argheyd

(2003). CEO incentives and corporate social performance. Journal of Business Ethics, 45, 341-359.

43.

McGuire

Sundgren

Schneeweiss

(1988). Corporate social responsibility and firm financial performance. Academy of Management Journal, 31, 854-872.

44.

McWilliams

Siegel

(2001). Corporate social responsibility: A theory of the firm perspective. Academy of Management Review, 26, 117-127.

45.

McWilliams

Siegel

(2010). Creating and capturing value: Strategic corporate social responsibility, resource based theory and sustainable competitive advantage. Journal of Management, 26, 117-127.

46.

Meyer

J. A.

Rowan

(1977). Institutionalized organizations: Formal structure as myth and ceremony. American Journal of Sociology, 83, 340-363.

47.

Mishina

Block

E. S.

Mannor

M. J.

(2012). The path dependence of organizational reputation: How social judgment influences assessments of capability and character. Strategic Management Journal, 33, 459-477.

48.

Mitchell

R. K.

Agle

B. R.

Wood

D. J.

(1997). Toward a theory of stakeholder identification and salience: Defining the principle of who and what really counts. Academy of Management Review, 22, 853-886.

49.

Orlitzky

Schmidt

F. L.

Rynes

S. L.

(2003). Corporate social and financial performance: A meta-analysis. Organization Studies, 24, 403-441.

50.

Pfarrer

M. D.

Pollock

T. G.

Rindova

V. P.

(2010). A tale of two assets: The effects of firm reputation and celebrity on earnings surprises and investors’ reactions. Academy of Management Journal, 53, 1131-1152.

51.

Rhee

Haunschild

P. R.

(2006). The liability of good reputation: A study of product recalls in the U.S. automobile industry. Organization Science, 17, 101-119.

52.

Rindova

V. P.

Fombrun

C. J.

(1999). Constructing competitive advantage: The role of firm-constituent interactions. Strategic Management Journal, 20, 691-710.

53.

Roberts

P. W.

Dowling

G. R.

(2002). Corporate reputation and sustained superior financial performance. Strategic Management Journal, 23, 1077-1093.

54.

Scherer

A. G.

Palazzo

(2007). Toward a political conception of corporate responsibility: Business and society seen from a Habermasian perspective. Academy of Management Review, 32, 1096-1120.

55.

Scott

W. R.

Meyer

J. W.

(1983). The organization of societal sectors. Beverly Hills, CA: SAGE.

56.

Sen

Bhattacharya

C. B.

(2001). Does doing good always lead to doing better? Consumer reactions to corporate social responsibility. Journal of Marketing Research, 38, 225-244.

57.

Servaes

Tamayo

(2013). The impact of corporate social responsibility on firm value: The role of customer awareness. Management Science, 59, 1045-1061.

58.

Sharfman

M. P.

Fernando

C. S.

(2008). Environmental risk management and the cost of capital. Strategic Management Journal, 29, 569-592.

59.

Strahan

P. E.

(1999). Borrower risk and the price and non-price terms of banks loans (Staff Reports No. 90). Federal Reserve Bank of New York. Retrieved from http://www.newyorkfed.org/medialibrary/media/research/staff_reports/sr90.pdf

60.

Surroca

Tribò

A. J.

Waddock

(2010). Corporate responsibility and financial Performance: The role of intangible resources. Strategic Management Journal, 31, 463-490.

61.

Tang

Hull

C. E.

Rothenberg

(2012). How corporate social responsibility engagement strategy moderates the CSR–financial performance relationship. Journal of Management Studies, 49, 1274-1303.

62.

Ulmer

R. R.

(2001). Effective crisis management through established stakeholder relationships: Malden Mills as a case study. Management Communication Quarterly, 14, 590-615.

63.

Vitell

S. J.

Davis

D. L.

(2004). The relationship between ethics and job satisfaction: An empirical investigation. Journal of Business Ethics, 9, 489-494.

64.

Wang

Bansal

(2012). Social responsibility in new ventures: Profiting from a long-term orientation. Strategic Management Journal, 33, 1135-1153.

65.

Williamson

O. E.

(1985). The economic institutions of capitalism. New York, NY: Free Press.