A Rising Tide Lifts All Boats: The Effects of Common Ownership on Corporate Social Responsibility

“Diversification is meant to be one of our risk-management tools. But if you’re facing systemic risk, you can run, but you can’t hide. In other words, we can decide not to hold a company that’s producing emissions—that’s the divestment case. However, if the emissions continue, we’re still exposed to the risk of climate change.”

–Anne Simpson, Managing Investment Director of CalPERS (New York Times, June 23, 2021)¹

Introduction

The concentration of investor ownership has shifted immensely over the last century. Following insights from Berle and Means (1932), most U.S. publicly traded firms historically had widely dispersed ownership structures, comprised of many investors owning small stakes in a few firms, which became the basis for most managerialist theories of corporate governance. However, more recently, as trillions of dollars have been directed into various institutional vehicles, asset managers have consolidated, and fewer companies have assumed public status, the ownership of publicly traded firms has become concentrated among fewer investors, mostly institutional investors that invest on behalf of others (Useem 1993, Davis 2009a, 2010). Indeed, with only so many firms in which to put their money, institutional investors are now more likely to hold a substantial stake in multiple firms out of sheer necessity, including firms in the same industry, creating what Davis (2008) calls a “latent network” of firms connected by single investors. We refer to investors that simultaneously own shares in two or more firms in the same industry as common institutional owners (or “common owners” for short).

Although common ownership—and the power it bestows—may seem advantageous for investors, it has created a new investment challenge: When a single investor owns multiple firms in an industry, it becomes even more necessary to manage systematic risk²—the risk inherent in an industry—than it is for an investor that owns a single firm in the industry (Coffee 2021). Such risk is critical to manage because one firm’s errant behavior can inflict harm on an entire industry of bystander firms (Paruchuri and Misangyi 2015), amplifying losses for the industry and its common owners. At the same time, systematic risk cannot be reduced by diversification (owning more firms in the industry would only increase it) or divestment (divesting entire industries could create other portfolio losses). As the usual strategies to mitigate risk become ineffective, common owners must find ways to manage spillovers between industry peers or else risk large losses to their portfolios. Therefore, as common owners contend with an elevated exposure to systematic risk, the question arises: How can common owners use their unique boundary-spanning ownership positions in multiple firms to enact strategies that maximize their portfolio returns?

Most research underscores the devious influence common owners have on corporate strategy and overlooks their systematic risk dilemma. A number of studies conclude that common owners engage firms to foster cooperation and anticompetitive behaviors (Gutiérrez and Philippon 2017, Azar et al. 2018), restricting firms from executing intra-industry, winner-take-most strategies. Inkpen and Sundaram (2022, p. 557) summarize the state of this literature: “Why is common ownership a concern? The argument goes as follows: Shareholders with concurrent investments in competing firms will maximize portfolio returns rather than individual firm returns leading to owners wanting firms to cooperate more and compete less; cooperating firms will increase prices for consumers and their own profitability, hurting consumer surplus and favoring producer surplus.” We propose an alternative but overlooked strategy: Common owners impel firms in their portfolios to adopt corporate strategies that minimize negative externalities and create spillover benefits for other same-industry firms in their portfolios, amplifying their own investment returns in the process.

We test our “rising tide lifts all boats” theory in the context of corporate social responsibility (CSR). Studies have documented the spillover effects that one firm’s social responsibility—or lack thereof—can have on industry peers (Shi et al. 2022). Corporate irresponsibility can undermine peer firms’ reputations and attract stricter industry-wide regulation, depressing returns for all industry players (Jonsson et al. 2009). Conversely, pro-CSR activities can improve the industry’s overall image, which helps peers build legitimacy and attracts better talent to the space (Burbano 2016). Given that common owners are exposed to these spillovers, which increase systematic risk, we predict that they will look to manage that risk and generate additional returns by encouraging firms to improve their CSR.

Using data from 1995–2018 and multiple estimation techniques, we document a positive association between common ownership and firm CSR. We account for three contingencies to better understand the mechanisms and motivations behind our theory of how common owners use CSR spillovers to manage systematic risk. We find the relation between common ownership and CSR is concentrated in long-term common ownership. We also find common owners’ effect on CSR is stronger in stakeholder sensitive industries, in which CSR spillovers are most economically impactful. Last, supporting the idea that common owners are financially motivated, we find that they increase financially material CSR rather than financially immaterial CSR. We further corroborate our results and address endogeneity concerns by using a difference-in-differences methodology that exploits financial institution mergers to create variation in common ownership that is exogenous to firms’ CSR.

Our study offers a new and opposing view of the common owner. Alongside the growth in common ownership, research on this topic has ballooned in recent years but thus far has focused almost exclusively on competitive moves and outcomes (Lewellen and Lowry 2021). From this angle, some have concluded that by fostering cooperation and anticompetitive practices among firms, common owners harm business and society (Azar et al. 2018, Cheng et al. 2021). Although some legal scholars have seen a need for common owners to account for negative externalities (Condon 2020b) and Coffee (2021) considers the potential for an “optimistic upside” of common ownership, the channels through which common owners could strengthen business and society remain undertheorized and research on them has been limited to the competitive dynamics lens that dominates the common ownership literature (Dai and Qiu 2021). By developing new theory about common owners’ orientation to systematic risk and their propensity to look for spillovers, and by providing evidence that common ownership improves CSR across portfolio firms, we offer a new and contrasting view of common owners.

We further contribute to literature on the antecedents of CSR by showing how common owners can spur new investment in CSR. Researchers have shown how a firm’s CSR practices are shaped by different types of investors, but these findings remain mixed, where some investors seem to impel firms to pursue socially responsible strategies (Flammer et al. 2021) and others do not (DesJardine and Durand 2020). To explain these differences, scholars have dissected investors’ investment horizons and prosocial values to understand their attitudes toward CSR (DesJardine et al. 2021). We offer a different view by developing a “portfolio perspective” of CSR, which reasons that investors with boundary-spanning ownership structures are motivated to manage CSR because they stand to gain (and lose) more from each firm’s social responsibility (and irresponsibility) than other, more isolated investors. By exploring how common owners affect different types of CSR activity, we also contribute to recent management research on the distinction between financially material and immaterial CSR (Christophe and Lee 2020).

The practical implications of this study are significant. Based on evidence that common ownership induces firm cooperation and suppresses industry competition, there have been numerous proposals to more heavily regulate common ownership, including antitrust laws to scrutinize common owners’ transactions and requirements that institutional investors either restrict common ownership positions or commit to not actively influencing firm practices (Posner et al. 2016). Both the Organisation for Economic Co-operation and Development (2017) and the U.S. Federal Trade Commission (2018) have warned about the anticompetitive effects of common owners. However, before enforcing more restrictive policies, our study underscores the need to consider outcomes beyond firm competitive actions to form a fuller picture of common owners and their influence on business and society.

Theoretical Background

Common Owner Dilemma and Strategies to Manage Systematic Risk

We consider common owners as investors that simultaneously own shares in two or more firms in the same industry. Three clarifications of this definition are integral to our theory. First, we follow the tradition in the literature of defining common ownership within the same industry (Connelly et al. 2019) rather than between industries. Because our theory focuses on systematic risk within industries, studying common ownership within industries is most appropriate. Second, we follow studies that define common ownership as investing in two or more firms rather than only two firms (Park et al. 2019). We do this because CSR is likely to spill over to many of the focal firm’s peer firms, not just one, and common owners’ incentives to encourage CSR will be higher the more firms they own in the industry. Third, we underscore that common ownership is a structural property of an investor’s holdings, not a particular type of investor. In this sense, common ownership is akin to block ownership, whereby an investor of any type (e.g., pension fund) is considered a blockholder when holding a nontrivial percentage of a firm’s outstanding shares. Not all investors are common owners, and an investor may be a common owner in one industry and not another. Consistent with the literature (Connelly et al. 2019), we bound our theory to institutional investors because these are likely the most prevalent and influential common owners.

With the rising concentration of assets in recent decades (Davis 2009b), common ownership is steadily becoming commonplace. As shown in Figure 1, the fraction of U.S. public firms held by common owners increased from 38% in 1995 to 90% in 2018. Of course, by nature, the largest asset managers will almost always be common owners. However, given that there were 4,162 common owners in the United States at the end of 2018 managing at least $100 million in assets (the threshold to report holdings to securities regulators in the United States), common owners are a highly diverse group of investors. Dividing common owners into terciles based on the total market value of their equity holdings in 2018, the smallest common owners (e.g., Optimum Investment Advisors, Sowell Management Services, Vantage Financial Partners) had total holdings of $100,000 to $100 million; common owners in the middle group (e.g., Boston Financial Management, Gagnon Advisors, Valiant Capital Management) had $100 million to $1 billion; and the largest common owners (e.g., Bank of America, Berkshire Hathaway, BlackRock) had more than $1 billion to several trillion dollars’ worth of equity holdings.

As common ownership has become the new norm for many institutional investors, they now face a critical investment challenge: managing systematic risk. Systematic risk, also known as “market risk,” is the risk inherent to an entire market segment or industry. When an investor owns significant stakes in many firms across an industry, it becomes essential to manage systematic risk, though it cannot be done through further diversification (as it can for firm-specific risk). For common owners, firm-specific risk still matters, but not to the extent that systematic risk does. This is because one firm’s actions that negatively affect an entire industry are felt far more severely by common owners in that industry than by non-common owners holding shares in only one firm in that industry. For example, as the CalPERS executive in our opening quote explains, one firm’s carbon emissions can especially harm common owners by increasing the climate risk among multiple other firms in their portfolios. Since further diversification is infeasible and ineffective for managing systematic risk, common owners need more novel strategies to enhance their risk-adjusted returns.

Empirical studies have largely overlooked common owners’ systematic risk dilemma. By far, the dominant—and most-debated—perspective in the literature is that common owners coordinate inter-firm actions that reduce industry competition. The objective for common owners is to have firms in their portfolio compete less aggressively with each other (Azar et al. 2018) or less directly with each other (Connelly et al. 2019), thereby increasing the returns of the firms in aggregate. However, coordinating competition does not explain how common owners resolve their exposure to systematic risk; coordinating actions can be costly and, when deemed collusive or anticompetitive, pose severe regulatory penalties on investors, creating even bigger risks for these investors and leading scholars to question the idea that common owners orchestrate firm coordination (Gilje et al. 2020, Lewellen and Lewellen 2022). Offering a different view, we conjecture that common owners use more direct strategies to mitigate their exposure to systematic risk.

To manage that exposure, common owners could leverage one firm’s resources to invest in practices that benefit other firms in the industry. Doing so, any individual firm may be worse off expending its resources on a given strategy, but if that investment nets gains across other firms that the common owner can capture, the strategy becomes sound from the common owner’s portfolio perspective. For such a strategy to work, the common owner’s portfolio value “need not be highly or even positively correlated with the value of any given focal firm” (Goranova and Ryan 2022, p. 531) because, although the common owner may lose value by encouraging certain actions at a focal firm, the benefits returned to the common owner through value enhancements at other portfolio firms could outweigh those costs (Condon 2020b, Coffee 2021). In the same way that spillovers allow a single firm’s errant behavior to doubly hurt a common owner, a single firm’s positive behaviors, by benefiting industry peers, can multiply returns to the common owner and reduce their systematic risk. If so, we expect common owners will use CSR as a means to reduce negative spillovers and amplify positive ones.

Shareholder Returns from CSR

CSR refers to the voluntary integration of social and environmental practices in a firm’s operations (Bansal and Song 2017). Firms that embrace CSR commit to internalizing social and environmental concerns, for instance, by developing practices and policies to curb climate change or enhance social equality.³

Some studies have shown that a firm’s CSR can benefit shareholders by reducing firm-specific risks. Grounded in stakeholder theory, the insurance-based view of CSR suggests that socially responsible actions help firms earn the goodwill and support of stakeholders when facing negative events (Godfrey et al. 2009). Indeed, studies find that CSR mitigates firm risk (Sharfman and Fernando 2008) and that social irresponsibility increases it (Kölbel et al. 2017). Shareholders may therefore use CSR information to understand firm-specific risk (Amel-Zadeh and Serafeim 2018) and embrace CSR itself to reduce firm-specific risk (Krueger et al. 2020).

However, given the costs, not all investors will reap equivalent rewards from increasing firms’ CSR investments. Some studies find that the financial returns to CSR, on a firm-specific basis, can be negative (Manchiraju and Rajgopal 2017, Chen et al. 2018). When financial returns do exist, they tend to arise in the long term (Bansal and DesJardine 2014), creating opportunities for investors to profit in the short term by reducing spending on socially responsible activities (DesJardine et al. 2021). Therefore, although various types of investors, such as foreign investors (Mun and Jung 2018) and institutional investors (Cobb 2015) may seek to propel firms’ CSR for their own benefit, the returns remain subject to debate.

However, the returns to CSR—and the investment rationale for promoting CSR—become far more compelling when the spillover benefits from one firm’s CSR practices are accounted for. Spillover refers to the “impact of a focal organization’s event on the perceptions and decisions of peer organizations and their stakeholders” (Shi et al. 2022, p. 185). As we explain next, whereas one firm’s social irresponsibility can create costly negative spillovers on industry peers, pro-CSR activities can produce positive spillovers for those peers. However, to capture the spillover benefits from CSR, an investor needs to hold shares in peer firms. Fortunately for common owners, the same concentration of ownership within industries that amplifies their exposure to systematic risk also uniquely positions them to reap returns from CSR spillovers that are missed by non-common owners.

Theoretical Development

Portfolio Perspective of CSR: Why Common Owners Realize Greater Returns from CSR

Compared with non-common owners, common owners stand to benefit more from improvements in firms’ CSR. To understand why, it is necessary to consider the “portfolio perspective” taken by common owners and how CSR creates spillover benefits that reduce systematic risk and enhance value in common owners’ portfolios. As noted, because common owners hold shares in multiple firms in an industry, they are concerned with systematic risk, in particular how a focal firm’s actions can spill over onto an entire industry of bystander firms and affect the aggregate value of those peer firms. Therefore, compared with non-common owners, common owners more highly value—and are more likely to push for—investments in CSR that produce limited benefits for a single firm (e.g., a small reduction in firm-specific risk) as long as they enhance the aggregate value of other industry peers in the investor’s portfolio (e.g., a large reduction in systematic risk). In fact, as Coffee (2021, p. 604) explains, common owners “may knowingly accept, and even cause, losses at some firms in their portfolio if they expect that those losses will be outweighed by correlative gains at other portfolio firms.” Therefore, even if CSR is loss-inducing for a focal firm, relative to non-common owners, common owners will be better off pushing for more CSR as long as the total value portfolio firms realize from a single firm’s CSR exceeds the costs of that firm’s investment. We unpack two risk-related reasons why the value of CSR is far greater when accounting for spillovers between peer firms.

First, a single firm’s CSR can shape how stakeholders engage with other firms in an industry by altering their perceptions of that industry. External stakeholders, situated outside firm boundaries, rely on other firms’ actions to make attributions about similar firms whose internal dynamics they cannot observe (Jonsson et al. 2009). Attributions occur between firms in the same industry because audiences tend to categorize firms in an industry together (Porac et al. 1999). Stakeholders—the media, customers, investors, and others—therefore tend to generalize a single firm’s (ir)responsibility to other firms in the industry (Barnett and Hoffman 2008, Diestre and Rajagopalan 2014). As stakeholders “act based on perceptions, not objective reality” (Wry 2009, p. 156) and make attributions within industries, their engagements with firms will be shaped by other peer firms’ approach to CSR, both in terms of positive responses to prosocial behaviors (which reduce systematic risk for the common owner) and negative responses for irresponsible actions (which exacerbate it). For example, when a bank adopts a profit-sharing initiative, it may benefit other banks by shaping prospective employees’ beliefs that the banking industry is employee friendly. Therefore, when a single firm improves its CSR, other industry peers can realize improved stakeholder perceptions, which reduces systematic risk for the common owner.

Second, because responsibility tends to be generalized across firms in an industry, one firm’s CSR can also create spillovers for peer firms and mitigate systematic risk by reducing the likelihood of industry-wide regulation. Firms with socially responsible track records can help deter regulatory restrictions by improving an industry’s image and legitimacy, whereas those with less responsibility will increase the likelihood that regulators and policymakers impose new regulations or sanctions on the industry, raising the costs and reducing the economic prospects for all industry players. For example, in their study of the 1984 chemical leak in Bhopal, India for which the Union Carbide Corporation was responsible, Blacconiere and Patten (1994) found evidence of a significant negative intra-industry reaction by investors as regulators imposed tighter regulations on all chemical firms. Negative intra-industry market reactions have been documented in other incidents for which a single firm was culpable, such as oil spills (Hsu et al. 2013) and product recalls (Borah and Tellis 2016), consistent with investors anticipating regulatory costs for the entire industry.

Given the potential for value created (or destroyed) by one firm’s CSR action (or inaction) to spill over onto peer firms in an industry, the common owner’s exposure to systematic risk stands to be impacted—either positively or negatively—from each firm’s level of social responsibility. The amplified impacts of CSR on systematic risk arise because two conditions are met: (1) the common owner holds stock in multiple firms in an industry, and (2) the effects of CSR spill over between peer firms. Taking the illustrative example in Figure 2, when TD Bank adopts more generous paternal leave policies, job seekers will more likely be attracted to the industry as TD Bank has made the industry appear friendly toward parents; likewise, if TD Bank was caught in a harassment scandal, then stakeholders may see not only the company but also its industry peers in a more negative light and regulators may enforce more stringent human resources regulations throughout the industry. By influencing firms’ CSR, common owners can shape the nature of spillovers that occur within industries and thus manage their own systematic risk.

Common owners will therefore have a stronger preference than non-common owners for their portfolio firms to invest in CSR. In some cases, common owners will directly engage portfolio firms to improve their CSR, for instance, through private discussions with management (McCahery et al. 2016) and filing CSR-related resolutions (Flammer et al. 2021). In line with the opening quote by the CalPERS manager, several senior investment executives we interviewed explained that common owners develop teams tasked to “think about these [CSR] issues and intersect and deal with and engage with management” on CSR. As one interviewee explained, where non-common owners typically “don’t bother with the [hard work] of engagement and inquiry,” many common owners do because they have an incentive to use CSR as a spillover mechanism to manage systematic risk. Therefore, “when it comes to issues of corporate social responsibility,” anecdotal evidence suggests that common owners “are much more willing to trumpet their expectations [of] and influence” on firms (Condon 2020a). In other cases, common owners may not directly engage firms on CSR issues but rather “vote with their feet” (i.e., sell off shares or threaten to do so), which can align management with investor preferences for CSR (Dikolli et al. 2009).

Non-common owners may also push firms to improve their CSR but will have a less compelling financial rationale to do so. Able to manage their portfolio risk through divestment and unable to capture the spillover benefits from one firm’s CSR, non-common owners stand to gain or lose less from intervening in each individual firm’s CSR. In sum, realizing that common owners are highly motivated to reduce their systematic risk and that CSR is a primary channel through which they can both lower this risk and capture additional spillover benefits, we expect common owners to improve portfolio firms’ CSR activities. We therefore hypothesize the following:

Our central hypothesis hinges on the assumption that common owners encourage responsibility and discourage irresponsibility because of the impact of CSR spillovers on the common owners’ systematic risk. If this is the case, we would expect the positive relationship between common ownership and CSR to be concentrated in (1) common owners with long investment horizons; (2) stakeholder sensitive industries; and (3) CSR investments with more certain financial returns.

What Types of Common Owners Care Most About CSR?

As noted, a single firm’s CSR activities can have considerable spillover effects on its peers, making CSR a useful strategic tool for managing systematic risk. However, many investments in CSR are likely to produce benefits predominantly in the long run, making a long investment horizon helpful to see the returns to CSR (Bansal and DesJardine 2014). Continuing with our example, the benefits of investing in parental-friendly workplace policies are likely to take time to manifest and even more time before spillover benefits materialize for industry peers, which is the driving force for common owners looking to reduce their systematic risk. Thus, CSR will primarily benefit common owners that hold stocks long enough for CSR spillovers to manifest. It is in this spirit that we theorize about long-term common owners.

We define long-term common owners as those with stable portfolios that hold positions in companies for long periods of time. Using the classification of Bushee (1998), prior studies (DesJardine and Bansal 2019, Flammer et al. 2021) classify long-term investors as dedicated and quasi-index institutional investors, both of which have long holding periods. They differ, however, in their investment strategies: dedicated institutional investors typically hold concentrated positions in relatively few companies (Connelly et al. 2010), whereas quasi-index institutional investors typically hold diversified positions in many companies because their investment strategy requires them to track general indexes (Appel et al. 2016).⁴ Still, when they have common ownership, their long horizons will bestow both types with the patience needed to benefit from CSR spillovers and will therefore be highly likely to use CSR as a tool to mitigate systematic risk in their portfolios. Thus, we account for both dedicated and quasi-index common owners in the following hypothesis:

Where Do Common Owners Matter Most?

CSR spillovers occur because stakeholders categorize firms together and attribute (or generalize) one firm’s behavior to other firms in the same industry (Diestre and Rajagopalan 2014). However, the financial impact of stakeholders’ attributions on firms and their common owners is likely to be particularly strong in industries where stakeholders will be more inclined to act on their attributions. We refer to these as stakeholder sensitive industries and theorize that they manifest in two distinct ways.

First, CSR spillovers will matter more to common owners in industries where stakeholder trust and corporate reputation are most valuable for business. Prior research shows that one firm’s reputation and trustworthiness impacts that of other firms in the same industry (Barnett and Hoffman 2008). Yet, such generalizations will more strongly impact common owners’ systematic risk when the reputations of firms in an industry have a meaningful impact on their bottom lines. For this to happen, reputational spillovers caused by one firm’s socially responsible or errant behavior needs to influence how stakeholders engage with the firm and its peers. Research finds that reputation matters mostly in industries that interact with, deal with, or sell directly to endconsumers, because end consumers are sensitive to firm reputation and can reward socially responsible firms by buying their products or services, and penalize irresponsible firms by not doing so (Dyreng et al. 2016). Reputation is also highly important for industries that depend on support from local communities because the risks inherent to certain industrial activities requires firms to build trust to obtain a “social license to operate” (Brammer and Millington 2004).

Second, CSR spillovers are more likely to matter to common owners in industries where regulators respond strongly to errant behaviors. In industries where firm actions do not carry a high risk of serious damage to the public, regulators will be less likely to penalize the entire industry with costly regulation when one firm behaves poorly. In such industries, lawmakers may instead deal with the irresponsible firm in isolation, allowing peer firms to escape penalties spilling over onto their business (Huang et al. 2017). However, in industries where irresponsibility can cause significant harm to the public, regulators tend to be less forgiving; consistent with this, certain industries, such as those involved in resource extraction, attract higher public attention and scrutiny and are therefore easier targets of political action (Liang and Renneboog 2017). Because accidents in these industries are especially costly to the environment and/or society, one firm’s errant behavior is more likely to draw stricter regulations intended to deter other peers from engaging in similar behavior. As an Amoco executive lamented, “We still have to live with the sins of our brothers. We were doing fine until Exxon spilled all that oil. Then we were painted with the same brush as them” (Hoffman 2001, p. 189). On the other hand, by behaving well (e.g., instituting compliance programs designed to reduce environmental accidents), firms may benefit from more relaxed regulatory monitoring and oversight. Thus, in industries that garner strong regulatory responses, one firm’s errant behavior can reduce the economic prospects of all industry firms, while good behavior can reduce the costs associated with regulation.

In their effort to manage systematic risk, common owners will be especially attuned to firm behaviors in stakeholder sensitive industries where these types of reputational and regulatory spillovers are most likely to manifest. In such industries, common owners stand to lose or benefit most from one firm’s CSR affecting the actions of key stakeholders. Common owners will therefore more intensely influence firms to improve CSR in industries that are particularly sensitive to stakeholder influence. We therefore hypothesize the following:

What Types of CSR Do Common Owners Impact Most?

Financially material CSR (“material CSR”) refers to CSR issues that are most value-relevant for firm financial performance, whereas financially immaterial CSR (“immaterial CSR”) refers to CSR issues that are unlikely to produce financial value. The financial materiality of a CSR issue depends on how central it is to the industry (Grewal et al. 2020). For example, greenhouse gas emissions can be a material issue for energy firms, but immaterial for firms in the financial services industry, which are relatively low emitters. The most widely used framework for distinguishing material and immaterial CSR issues from an investor viewpoint is that of the Sustainability Accounting Standards Board (SASB). Using SASB’s standards, research documents that material CSR investments have meaningful predictive power over firms’ future financial performance (Khan et al. 2016). However, CSR being immaterial does not mean it is irrelevant or unimportant; SASB’s focus is on investors, but a CSR issue immaterial to investors may still be important or valuable to other stakeholders. For instance, a large bank’s reduction of its emissions could help a country achieve its climate change objectives (e.g., carbon neutrality of the financial sector) and therefore be important to politicians, regulators, and the government while having a limited effect on the bank’s bottom line.

If common owners encourage CSR for financial reasons, then increases in CSR should be driven mostly by improvements in material CSR. Material CSR has the highest potential for financial impacts between firms—and therefore potential benefit for the common owner—because such investments touch on financially relevant CSR issues in an industry. Common owners, with sizable stakes in multiple same-industry firms, are in a unique position to learn which CSR issues are most material to an industry. They thus have greater incentive and opportunity than non-common owners do to evaluate each CSR issue “on its merits in the context of materiality to the company’s long-term financial performance” (Federal Trade Commission 2018, p. 74). In turn, they can understand which forms of CSR are most likely to have spillover effects within the industry and can focus their efforts on altering these material CSR activities to manage systematic risk. Although many non-common owners will also be financially motivated to support material CSR over immaterial CSR, common owners will be even more compelled to do so given their sensitivity to systematic risk and the substantial spillover effects produced by material CSR. By comparison, due to firms’ resource limitations and the limited potential for financially relevant spillovers, common owners are unlikely to encourage increases in immaterial forms of CSR and may even discourage such. In closing, we hypothesize the following:

Methods

Sample

Our primary data source is the Thomson Reuters Institutional Holdings (13F) database, which consolidates 13F filings (containing details of equity ownership required by the Securities and Exchange Commission) for investment managers with more than $100 million in holdings in publicly traded U.S. firms. We collected institutional ownership data for the 22,665 firms with a non-missing Committee on Uniform Securities Identification Procedures (CUSIP) identifier and four-digit Standard Industrial Classification (SIC) code from 1995 onwards. We then merged in CSR data from Morgan Stanley Capital International’s (MSCI) Kinder, Lyndenberg, & Domini (KLD) data set, which provides data until 2018 and thus defines the end of our sample period. This left us with 5,887 firms with non-missing CSR data. Last, we collected financial data from Compustat, corporate misconduct data from Violation Tracker, and the Reputation Risk (RepRisk) Index from Wharton Research Data Services (WRDS). Our final sample includes 4,769 firms, held by 2,396 distinct common owners and encompassing 31,024 firm-year observations between 1995 and 2018.

Dependent Variables

CSR Performance.

Following prior research (DesJardine et al. 2021), we use data from KLD to measure CSR. Among CSR data providers, KLD works well for our study because it covers U.S. firms over a long period of time (since 1995). KLD assesses firms on strength and concern topics using a binary system; for each topic, firms are assigned a value of one, indicating the presence of that criterion, or zero, indicating its absence. Although KLD strengths and concerns can capture positive CSR (e.g., a diverse workforce) and negative CSR (e.g., involvement in a child labor–related legal case), our theory relates to both dimensions—the existence of strengths and mitigation of concerns—leading us to follow prior studies by constructing a net measure of CSR (Barnett and Salomon 2012) that deducts KLD concerns from strengths in the environment, community, employee relations, diversity, product, and human rights categories. Because the maximum number of strengths and concerns varies over time, we follow prior research (Lins et al. 2017) and divide the number of strengths (or concerns) a firm has in a given category by the maximum possible number of strengths (or concerns) in that category-year. This results in indices with a range of zero to one for each category-year, across both strengths and concerns. Next, we compute a net CSR index for each category-year that subtracts the concern index from the strength index and thus ranges from −1 to +1. Finally, we sum the net CSR indices across the six categories to calculate our primary dependent variable CSR performance. We also separate this measure into strengths (CSR strengths) and concerns (CSR concerns) by taking the sum of the respective indices across the six categories.

Material and Immaterial CSR Performance.

We further separate our measure of CSR performance into CSR that is investor-relevant (financially material) and CSR that matters more to non-equity stakeholders (financially immaterial) (Khan et al. 2016). SASB classifies material and immaterial CSR topics by industry. We follow prior research and use SASB’s industry-specific standards to map the 53 KLD subcategories as material or immaterial for the 11 sectors (representing 77 industries) covered by SASB. For each firm-year, we measure material CSR performance by subtracting an index of material concerns from an index of material strengths (using the process described in the previous section) and immaterial CSR performance using the same method for the immaterial categories. We detail our SASB mapping process in the online appendix.

Independent Variables

Common Institutional Ownership.

We follow prior literature that defines common ownership as occurring when an institutional investor simultaneously owns at least two firms operating in the same primary four-digit SIC industry (Connelly et al. 2019). Although researchers have developed different ways to operationalize this definition, we adopt that of Connelly et al. (2019). Specifically, our independent variable of interest—common institutional ownership (CIO)—is a ratio of which the numerator is the sum of an institutional investor’s fractional holdings in two firms operating in the same industry (common fractional holdings) and the denominator is the sum of the fractional holdings of the pair’s non-common institutional investors plus the sum of the common fractional holdings (i.e., the numerator), aggregated across all of the pair’s common institutional investors.⁵ We aggregate this firm-pair measure across all of a firm’s same-industry peers to form a firm-level continuous measure capturing each firm’s common ownership (Lewellen and Lowry 2021). Importantly, following Connelly and colleagues, to ensure they have sufficient influence, we require that a common owner has a blockholding (>1% ownership) stake in each firm.⁶

Relative to other measures in the literature, the Connelly et al. (2019) measure (Connelly measure) is desirable for two reasons. First, “count” measures that count the unique number of investors that own a focal firm and any of its same-industry peers or that count the number of same-industry peers that share at least one investor with the focal firm (He and Huang 2017, Park et al. 2019) are intuitive but fail to reflect the decomposition of ownership between two firms (e.g., the value of such a measure would be the same whether an investor owned one share or all of the shares of a commonly owned firm). The Connelly measure, however, reflects the degree of ownership held by each common owner. Second, “average” measures, which take the arithmetic or product-based average of an investor’s ownership in commonly owned firms (Gilje et al. 2020, Lewellen and Lowry 2021), do not account for the proportion of common institutional ownership relative to total institutional ownership, which the Connelly measure does. Scaling common ownership in this way is important because it accounts for the overall influence of common owners relative to other (non-common) investors. Despite these advantages of the Connelly measure, we compute alternate measures of CIO, which we discuss later.

Long-Term and Short-Term Common Ownership.

To measure common ownership by long-term investors, we map the investor classification data of Bushee (1998) to our institutional investor holdings (13F) data. We consider long-term investors as those Bushee classifies as dedicated and quasi-index and short-term investors as those classified as transient (Brochet et al. 2015, Flammer et al. 2021). Therefore, we compute dedicated (quasi-index) CIO as the sum of a dedicated (quasi-index) institutional investor’s fractional holdings in two firms operating in the same industry (the numerator), divided by the numerator plus the sum of the fractional holdings of the pair’s non-common institutional investors (the denominator), aggregated across all of the pair’s dedicated common institutional investors.⁷ We again form a firm-level measure by aggregating the firm-pair measure across all of a firm’s same-industry peers. Our measure of transient CIO is measured in the same way but uses the holdings of transient institutional investors.

Stakeholder Sensitive Industries.

We examine whether the relation between common ownership and CSR is particularly strong in industries that are most sensitive to stakeholder attributions. We create an indicator variable, stakeholder sensitive industry, that equals one for firms belonging to industries that bear high financial impact from regulatory or reputational attributions and zero otherwise. To construct this measure, we follow Brammer and Millington (2004) (B&M), who identify 41 industries with a high “risk of regulatory interference” and 21 industries for which reputation is most valuable because firms are consumer-facing or require public trust to maintain their social license to operate. Under both classifications, the “potentially significant environmental or social consequences” (Brammer and Millington 2004, p. 1417) of certain industries intensifies stakeholders’ responses to firm behaviors.⁸

To assess the validity of the 41 industries that are sensitive to regulation, we download data on environmental and social penalties from Violation Tracker from 2000–2021. Consistent with our focus on regulatory costs stemming from regulatory actions, Violation Tracker collects penalty data from regulatory agencies and related sources.⁹ We calculate the average dollar amount of social penalties (from “employment-related offences” and “safety-related offenses”) and environmental penalties (from “environment-related offences”) for each industry. We find considerable overlap between the 41 industries in the B&M list and the industries with the highest (i.e., 95th percentile) environmental or social penalties according to Violation Tracker.¹⁰

To assess the list of 21 industries where reputation and trust are most valuable, we download RepRisk data from WRDS. Consistent with our focus on industry vulnerability to consumer and public perceptions, RepRisk uses media and news sources to calculate an index that measures a firm’s reputational risk exposure to CSR issues.¹¹ We average the RepRisk Index for each industry from 2000 to 2016 (the last year for which RepRisk data are available on WRDS). Again, we find considerable overlap between the 21 industries in the B&M list and the industries with the highest (i.e., 95th percentile) CSR-related reputational risks per the RepRisk Index.¹²

Given these results, we code firms in SIC codes 1000–1499 (Mining and Oil & Gas), 2600–2699 (Pulp & Paper), 2800–2899 (Chemicals), 4400–4499 and 1781 (Water Sourcing and Distribution), 2900–2999 (Energy Production), 4600–4699 and 4900–4939 (Utilities), 2100–2199 (Tobacco), and 5181–5182, 5813, 2084, and 2085 (Alcohol) as operating in a stakeholder sensitive industry.

Control Variables

We include control variables to mitigate the possibility that our results are driven by confounding factors. Foremost, because blockholder ownership has been shown to impact CSR performance (Neubaum and Zahra 2006), we control for this variable while taking care to exclude common institutional ownership to ensure that we do not conflate this measure with our common ownership predictor of interest. Specifically, non-common blockholder ownership is the percentage of a firm’s outstanding shares held by a firm’s non-common blockholders (that own at least 1% of the firm).¹³

To control for the influence of firm size on CSR, we include firm size as one plus the natural logarithm of the book value of total assets. Because firms that perform better may also have greater capacity to invest in CSR, we include three performance measures: return on assets (ROA), measured as income before extraordinary items over total assets; market-to-book ratio, measured as the ratio of the market value to the book value of total assets; and earnings volatility, measured as the standard deviation of the return on assets over the prior five years. Firms with a greater capacity to take on new debt may also invest more in new CSR initiatives. We thus consider differences in financing by including leverage as the ratio of total debt to total assets. Finally, we include sales growth, measured as the annual percentage change in sales, to control for the possibility that growing firms invest more in CSR and attract common owners.

Analyses

Our main analyses use ordinary least squares (OLS) regressions with firm fixed effects that control for observed and unobserved time-invariant firm characteristics and year fixed effects that control for common macroeconomic shocks that affect all firms. The results, reported in Table A1 (columns 4-6) of the online appendix, are similar if we lag the independent variables by one year. We winsorize all continuous variables at the 1% level to control for extreme outliers. Robust standard errors are clustered by firm to account for autocorrelation in the data within firms across time.

Results

Table 1 reports descriptive statistics and correlations, and Tables 2 and 3 report our main results. Hypothesis 1 predicts that common ownership is positively related to CSR. In Model 1 of Table 2, the coefficient estimate of CIO is 0.001 (p < 0.001), consistent with Hypothesis 1. In terms of economic significance, increasing CIO by one standard deviation increases CSR performance by 0.06 units (practically speaking, firms engage in 0.06 more strength than concern activities).¹⁴ Given that the average CSR performance in our sample is −0.04 (negative because the number of concerns slightly exceeds the number of strengths), a 0.06-unit increase shifts the average firm’s net negative CSR score to a net positive score of 0.02; this increase is 9% of the standard deviation of CSR performance.¹⁵ Models 2 and 3 show that the increase in CSR performance is driven by an increase in CSR strengths (β = 0.001, p < 0.001) rather than by a decrease in CSR concerns (the coefficient in Model 3 is close to zero and insignificant).

Table 1. Descriptive Statistics and Correlation Matrix

Table 2. OLS Regressions for Common Institutional Ownership and CSR Performance

Table 3. Stakeholder Sensitive Industry and Material vs. Immaterial CSR Performance

To make our main results more concrete, we use a recently developed methodology from the finance literature (Di Giuli and Kostovetsky 2014) to translate the increase in CSR performance into dollar terms. We find that the estimated CSR expenditures associated with a one standard deviation increase in common ownership amount, on average, to 2.7% of firms’ net income.¹⁶

Hypothesis 2 predicts that long-term common ownership is positively associated with CSR. In Models 4–6, we replace CIO with dedicated CIO, quasi-index CIO, and transient CIO. In line with Hypothesis 2, the coefficient estimates on dedicated CIO and quasi-index CIO, both of which capture long-term common ownership, are positive and significant in Models 4 and 5, which use CSR performance and CSR strengths, respectively, as the dependent variable. In Model 6, where CSR concerns is the dependent variable, the coefficient estimate on dedicated CIO is close to zero and insignificant, whereas it is negative and significant for quasi-index CIO. The estimates in Model 4 suggest that increasing dedicated CIO (β = 0.001, p < 0.001) and quasi-index CIO (β = 0.002, p < 0.001) each by one standard deviation increases CSR performance by 0.1 units—nearly 16% of its standard deviation.¹⁷ Moreover, as its coefficient estimates in Models 4–6 are close to zero and insignificant, transient CIO has a negligible relation with CSR performance. F tests find that the coefficient estimates on dedicated CIO and quasi-index CIO in Models 4 and 5 are significantly larger than the coefficient estimates on transient CIO (p < 0.001).

Hypothesis 3 predicts that the positive relationship between common ownership and CSR is stronger in stakeholder sensitive industries than in other industries. In line with this prediction, as reported in Model 7 of Table 3, the coefficient estimate of the interaction of CIO × stakeholder sensitive industry is 0.02 (p < 0.001). Given the coefficient on the interaction term, the relationship between CIO and CSR performance strengthens so that increasing common ownership by one standard deviation increases CSR performance by 1.30 units or shifts the average firm’s CSR performance from −0.04 net CSR concerns to 1.26 net CSR strengths.¹⁸ Again, the coefficients on the interaction terms in Models 8 and 9 are consistent with a statistically significant increase in CSR strengths (β = 0.014, p < 0.001) and decrease in CSR concerns (β = −0.006, p < 0.05).

Hypothesis 4 posits that common owners increase firms’ material CSR more than their immaterial CSR. Model 10 of Table 3 uses the same specification as Model 1 of Table 2 but uses material CSR performance as the dependent variable. The coefficient estimate of 0.002 (p < 0.05) indicates that CIO has a positive and significant relation with financially material CSR. In Model 11, however, where immaterial CSR performance is the dependent variable, the coefficient on CIO is −0.002 (p < 0.01), indicating that CIO is negatively and significantly related to financially immaterial CSR. Based on a system of seemingly unrelated regressions, which uses a common sample for the two regressions to test whether the coefficients differ, the coefficients on CIO across the two models are statistically significantly different (p < 0.001).

Supplementary Analyses

Natural Experiment to Address Endogeneity.

A central concern of our main results is that an omitted variable simultaneously drives common ownership and CSR. For instance, board members can influence decisions that attract higher investment from common owners while also influencing investments in CSR. Another concern is the potential for stock selection, namely, common owners are attracted to firms that invest heavily in CSR. Still another possibility is that our results are attributable to firms imitating other firms’ CSR (Cao et al. 2019). We address these endogeneity concerns by using mergers between financial institutions to generate exogenous variation in common ownership that is independent of firms’ CSR. Such mergers, largely due to financial sector deregulation and exogenous to any portfolio firm attributes, increase common ownership for same-industry firms owned by either of the merging institutions.

Following He and Huang (2017), we identify 14 financial institution mergers during our sample period (listed and explained in more detail in the online appendix).¹⁹ A firm is included in the treatment group (i.e., experienced an increase in common ownership) if, in the quarter immediately preceding the merger’s announcement, the firm is blockheld by one of the merging institutions and the other merging institution does not blockhold that firm but does blockhold at least one other firm in the same industry. A firm is included in the control group (i.e., experienced no increase in common ownership) if, in the quarter before the merger’s announcement, the firm is blockheld by a financial institution that blockholds a treated firm and the other merging institution does not blockhold any of its same-industry peers (see Figure A1 in the online appendix). Because control firms experience no change in common ownership but are still owned by one of the merging institutions, there is little possibility that differences in the merging institutions’ stock selection and investment preferences confound our results.

We estimate a difference-in-differences (DID) model in the three years before and after each merger. The variable of interest is treat × post, which captures the effect of the exogenous change in common ownership on the CSR of treated firms compared with its effect on the control firms. We include all prior control variables and year fixed effects and firm-merger fixed effects. We cluster standard errors at the firm level. In untabulated analyses, we find validity for the parallel trends assumption.

In line with Hypothesis 1, results from Model 12 in Table 4 show a positive and significant relationship between CIO and CSR performance (β = 0.226, p < 0.10), attributable largely to increases in CSR strengths (Model 13) and not reductions in CSR concerns (Model 14). In support of Hypothesis 2, long-term common owners push for CSR (the coefficient of dedicated CIO is positive and significant in Model 15), driven by increases in CSR strengths (the coefficients of dedicated CIO and quasi-index CIO are both positive and significant at the 10% level in Model 16). Once again, we find no significant relation between transient CIO and CSR.

Table 4. Difference-in-Differences Analyses Using Financial Institution Mergers

We next find that the positive relation between CIO and CSR is stronger in stakeholder sensitive industries. The coefficient estimate of treat × post × stakeholder sensitive industry in Model 18 is positive and significant (β = 0.808, p < 0.05), again attributable to both increases in CSR strengths (Model 19: β = 0.562, p < 0.01) and decreases in CSR concerns (Model 20: β = −0.247, p < 0.10). Finally, results in Models 21 and 22 show that the main relation between CIO and CSR is positive for material CSR performance (β = 0.104, p < 0.10) and negative for immaterial CSR performance (β = −0.186, p < 0.10).

Robustness Checks

To ensure the robustness of our findings, we took the following steps, and tabulated the results in the online appendix.

Alternative Measures of CSR.

We computed three alternate measures of CSR performance. The first is similar to our main measure but includes the corporate governance category from KLD (CSR perf + GOV). The second is the sum of the environmental and social pillar scores from Thomson Reuters ASSET4 (ASSET4 E+S). The third uses the “Overall ESG Score” from MSCI’s Intangible Value Added database (MSCI ESG Score). As shown in columns 1–3 of Table A1 in the online appendix, the results from using these alternate measures of CSR performance align with our main findings.

Alternative Measures of Common Ownership.

In constructing our main measure, we excluded common owners with less than 1% ownership, consistent with Connelly et al. (2019) and others (Gilje et al. 2020), to ensure that common owners have significant incentives to engage and influence firms. However, we also tested our results using a more restrictive blockholding threshold that requires institutional investors to own at least five percent of a firm’s outstanding shares (He and Huang 2017). We also calculated a product-based Connelly measure that uses the product of the common fractional holdings (rather than the summation); this approach helps to account for symmetry in the extent to which an investor owns a similar stake in both firms (Lewellen and Lowry 2021).²⁰ Re-estimating our main models using these alternative measures, we found no material changes in the results, as shown in Table A2 in the online appendix.

Subdimensions of Stakeholder Sensitive Industry Moderator.

To ascertain whether our results for Hypothesis 3 hold for both subdimensions of stakeholder sensitive industry, we split the measure into its two components: reputation sensitive industries that are most impacted by reputational attributions and regulation sensitive industries that are most impacted by regulatory attributions.²¹ In Table A3 in the online appendix, we find that both components are important moderators (statistically and economically) of the relation between common ownership and CSR performance.

Discussion

Our study takes a new look at the investment priorities and corporate influence of common owners. The vast rise of common ownership, driven by fewer investors each expanding their stakes in more and more companies (Davis 2010), has created a new investment challenge for the common owner: how to manage systematic risk. By owning multiple firms across an industry, common owners bear unparalleled exposure to the risks that can result when one firm’s actions affect the value of industry peers, amplifying gains or losses to the common owner’s portfolio. We theorize that common owners seek to protect themselves from systematic risk and to enhance their returns by improving CSR among portfolio firms. Common owners stand to gain far more than non-common owners do from one firm’s CSR investment because it can mitigate negative externalities and create positive spillovers for other firms in their portfolios.

We find, in support of our theory, that common ownership has a positive association with CSR. Our main results show that a one standard deviation increase in common ownership is associated with a 0.06-unit increase in CSR, or 9% of the standard deviation of our focal KLD-based CSR performance measure. We find statistically and economically similar results using alternate CSR measures constructed from ASSET4 and MSCI ESG data. We also find that the main relation between common ownership and CSR is attributable to common owners with long investment horizons and becomes stronger in stakeholder sensitive industries, where CSR spillovers are most economically impactful. Moreover, we find that common ownership is most strongly associated with increases in financially material CSR rather than financially immaterial CSR. Acknowledging that the relationship between common ownership and CSR is potentially endogenous and that other explanations may exist, we apply a DID approach that exploits exogenous variation in common ownership and find further support for our main findings. We believe our findings contribute to several streams of research, which we now discuss.

Theoretical Implications

Foremost, our study offers a new perspective on the motives common owners have to shape firms’ strategic actions. By far, the most popular focus in the literature has been to approach common ownership from a competitive angle. Debate exists among academics (Goranova and Ryan 2022, Lewellen and Lowry 2021) and practitioners (OECD 2017) as to whether common owners seek to improve their returns by coordinating actions among portfolio firms to soften industry competition. Yet, manipulating competitive dynamics does not explain how common owners address an arguably greater investment challenge: managing systematic risk. Motivated by this investment dilemma, our study reveals a new strategy that common owners use to manage systematic risk—leveraging CSR. By encouraging firms in their portfolio to be more socially responsible, and discouraging irresponsible or deviant behaviors, common owners can manage spillovers in ways that mitigate systematic risk and improve their overall portfolio value. Although some legal scholars have recently discussed the need for common owners to manage firm externalities (Condon 2020b) and a potential “upside” of common ownership (Coffee 2021), we are the first to venture beyond competitive dynamics (Dai and Qiu 2021) to provide empirical evidence on these matters.²² Our study shows that common owners are motivated to use CSR as a tool to manage systematic risk and maximize portfolio returns. By showing the positive effects common owners have on CSR, our study also contrasts sharply with the view that common owners harm society and the economy by encouraging corporate collusion (Federal Trade Commission 2018).

Second, our study contributes to the burgeoning literature on the role of investors in shaping firm CSR. Prior studies have linked CSR-related outcomes to ownership by various types of investors, including, among others, activist hedge funds (DesJardine and Durand 2020) and short- and long-term investors (Cobb 2015). Although prior studies have been useful in uncovering how different types of investors impact firms’ CSR, we thus far know little about the role of common owners. This is surprising not only because common owners arguably have the strongest incentives of all investors to control negative corporate externalities but also because research has already shown that common owners can profoundly impact a variety of corporate practices and strategies, such as competitive actions (Connelly et al. 2019) and mergers and acquisitions (Goranova et al. 2010). Integrating CSR research with studies on common ownership in management, law, accounting, and finance, we offer theory and supportive evidence on how common owners enhance the CSR activities of their portfolio firms.

Approaching CSR from a portfolio perspective, which accounts for spillover effects, provides new insight on the financial rationale some investors have to promote socially responsible corporate activities and to discourage irresponsible ones. Prior studies typically attribute investors’ promotion of CSR to one of two rationales: Either investors are socially motivated to impose their social preferences on firms (Guay et al. 2004) or they are financially motivated to use CSR as a strategy to improve firm-specific profits (DesJardine and Durand 2020). Our study presents evidence for a third and entirely overlooked rationale: investors promote CSR to reap the spillover benefits to their portfolios and maximize portfolio returns. Scholars have long debated investors’ financial motives for CSR, but this conversation has remained almost exclusively at the firm level, asking for each individual firm, “Does it pay to be green?” We inform this discussion by revealing how the financial rationale for CSR becomes far more persuasive when approached from a portfolio perspective rather than any individual firm’s perspective.

We also contribute to research on financially material and immaterial CSR. Management scholars have long warned that research should not treat CSR as a monolithic construct but rather should think about how CSR can be partitioned to capture its various dimensions. Whereas some studies have considered specific CSR actions, such as the adoption of environmental standards (Dowell et al. 2000), others have disaggregated CSR into internal and external CSR (Hawn and Ioannou 2016), primary and secondary CSR (Hillman and Keim 2001), tactical and strategic CSR (DesJardine et al. 2019), responsibility and irresponsibility (Kölbel et al. 2017), and various dimensions that capture separate CSR activities (Flammer 2015). Although recent accounting and finance research has considered CSR that is material and immaterial from an investor standpoint (Khan et al. 2016, Grewal et al. 2020), few management studies have made this distinction. We contribute by considering how common owners’ emphasis on their portfolio value leads to changes in CSR driven primarily by increases in material CSR and decreases in immaterial CSR. By making this distinction, we also provide further evidence that common owners concern themselves with CSR for financial reasons.

Practical Implications

Our results show that common owners encourage CSR to reduce their exposure to systematic risk and maximize their portfolio returns. Given the vast sums invested in ESG funds and other socially responsible investment vehicles, our study illuminates an important driver behind the mainstream investment community’s recent push for improved CSR among publicly traded companies. Although critics argue that ESG investing does not improve CSR outcomes but mainly serves to preempt and delay regulation, we show that common owners’ need to manage systematic risk creates a potent market mechanism that forces firms to internalize their negative externalities, thereby improving the CSR of firms held in common owners’ portfolios. Rather than replace regulation, common owners’ incentives to maximize positive CSR spillovers and minimize negative ones can help society better manage CSR issues, including climate change, particularly in jurisdictions in which such regulation has been lacking. Relatedly, given the benefits common ownership can have on the environment and society by improving firms’ CSR practices, our findings suggest that policymakers may not want to be too hasty in restricting common ownership amid suggestions that common owners incentivize anticompetitive behaviors (Azar et al. 2018).

For other stakeholders wanting to see improved CSR, including governments, consumers, and employees, our study suggests that faster progress can be made by supporting common owners. In particular, socially oriented investors without common ownership may advance their non-financial objectives by supporting the pro-CSR agendas of common owners, for instance, by voting in support of their pro-CSR shareholder resolutions. Finally, an executive who sees the value of CSR but faces investor pushback can use our study to appeal to the firm’s common owners and demonstrate how the returns to CSR for these investors are uniquely amplified when the spillover benefits from CSR are accounted for.

Limitations and Directions for Future Research

We acknowledge several limitations of our work, which we hope present opportunities for future research. First, a limitation of our empirical approach is that we did not observe the common owners’ actual engagements with managers. The closest we came is through our interviews, drawing on anecdotal evidence, and through prior studies showing that common owners do engage managers, and influence firms via the threat of exit. Future research could use a more observational approach and go “behind closed doors” to better understand how common owners’ influence over managers unfolds. Second, although we looked at common owners’ investment horizons, common owners may be further subdivided by, for example, investor type (such as pension funds and foundations) and geographic location. Beyond their investment horizons, do common owners in different places and with different investment styles differentially affect CSR? Third, we assume that common owners are uniquely positioned to capture the financial spillovers from one firm’s CSR onto industry peers. Future research could directly test this assumption by using portfolio-level return models to compare the financial returns of CSR to common owners versus non-common owners.

Conclusion

Common owners have received considerable criticism over concerns that they limit competition. However, looking more broadly, we see how they can benefit society by promoting corporate social activities. A unique feature of our theory regarding common owners’ exposure to systematic risk is that it is not bound to CSR but can guide inquiries into the other ways that common owners might benefit business and society by helping firms internalize their externalities when other governance mechanisms, such as independent directorships and regulatory penalties, fail to do so. Without incentives to see the bigger landscape, some firms will myopically advance their own agendas at the expense of all others in a “tragedy of the commons.” By having to account for the impact of each firm’s actions on the larger system, common owners can incentivize firms to internalize their externalities.