The Corporate Opportunity Structure for Shareholder Activism: How Activist Hedge Funds Exploit Board Demographic Diversity

Published Online:https://doi.org/10.1287/orsc.2023.1679

Abstract

Inspired by research on social movements, we extend the idea that activists look for opportunities to target firms to the realm of financially motivated shareholder activists. Focusing on activist hedge funds, we argue that hedge fund campaigns are more likely to succeed when boards are slow and less united and that, compared with more homogeneous boards, demographically diverse boards tend to act more slowly and with less unity. Although these attributes make demographically diverse boards more effective under “normal” circumstances, they become a liability in confrontations with activist hedge funds. We, therefore, hypothesize that when subject to governance and performance problems, firms become more likely targets of activist hedge funds when they also have demographically diverse boards. To further probe our theory, we explore the opportunity recognition of activist hedge funds in two ways. First, we posit that this opportunity will be recognized and exploited primarily by experienced activist hedge funds. Second, we argue that activist hedge funds’ opportunity recognition is correct in so far that demographically diverse boards respond to activism campaigns in ways that are likely to benefit activist hedge funds. Using data on United States-based activism campaigns, we find support for our theory. By simultaneously studying problems and opportunities, this study establishes a foundation for examining when the disciplinary effect of shareholder activism may go awry and reveals why a strict business case for demographic diversity may be insufficient to align all shareholders behind board diversity.

Supplemental Material: The online appendix is available at https://doi.org/10.1287/orsc.2023.1679.

Financially-motivated shareholder activists (hereafter “shareholder activists”) often target firms that experience problems, such as deficient governance or underperformance (Brav et al. 2008, DesJardine and Shi 2022). By targeting such firms, shareholder activists seek to resolve these problems and to benefit from subsequent stock price improvements caused by their resolutions. Yet, research on corporate opportunity structures (Briscoe et al. 2014) reveals that activists target based not only on problems but also, on opportunities that increase the odds they can successfully intervene in corporate affairs to address those problems (Wang et al. 2021). Opportunities arise when activists assume that certain attributes of a firm—such as its size or board composition—increase the likelihood that they can successfully influence a firm (King 2008, Bartley and Child 2014, Gupta et al. 2018). Following the idea that social activists look for corporate opportunities, we expect that shareholder activists will also account for opportunities when deciding which firms to target to increase their likelihood of success. In this paper, we illuminate the corporate opportunity structure for shareholder activism by exploring the possibility that shareholder activists target based on both problems and opportunities.

Illuminating the corporate opportunity structure for shareholder activism is consequential not only because accounting for opportunities can help reconcile inconsistent findings on why shareholders target specific firms, which we unpack, but because targeting on opportunities raises fundamental concerns about the disciplinary effects of shareholder activism (Fama and Jensen 1983, Fos 2017). Supported by findings on the spillover effects of shareholder intervention (DesJardine et al. 2022a, Shi et al. 2022), a key justification for shareholder activism is that the mere threat of activism incentivizes managers to preemptively resolve governance and performance problems to avoid becoming subject to activist targeting. Gantchev et al. (2019, p. 1034), for instance, argue that shareholder activism has become “a primary external disciplining force” for managers, to which many practitioners agree (e.g., Freedman et al. 2019). However, if shareholder activists were to target firms based on opportunities, the threat of being targeted by activists may not lead managers to resolve problems in their firms but instead, to organize their companies in ways that make them more effective in confrontations with shareholder activists—even if this makes their companies less effective under “normal” circumstances. In such cases, opportunity-based targeting would incentivize managers to adopt ineffective organizational forms and thereby, create problems, contorting the disciplining effect of shareholder activism.

We explore the possibility that shareholder activists target based on opportunities by focusing on board demographic diversity. Compared with more homogenous boards, demographically diverse boards tend to make decisions (1) more slowly and (2) with less unity (Goodstein et al. 1994, Triana et al. 2014, Jung et al. 2023). Although these two aspects will make diverse boards more effective under “normal” circumstances—for example, because carefully contemplating divergent perspectives contributes to more innovative strategies (Miller and Triana 2009)—they become a liability in high-strain circumstances, when fast decisions made with unity are crucial (Jung et al. 2023). When firms underperform, for example, Triana et al. (2014, p. 614) warn that a diverse “board may become inert because it must now consider diverse views, even if it means becoming overloaded with information at a time when the board needs to act quickly.” We argue that confrontations with shareholder activists constitute a high-strain circumstance in which slower decision making and lower unity become a liability. Following shareholder activist interventions, slower boards will find it hard to catch up with activists in the race to mobilize support from other shareholders, which may force these boards to succumb to activists. Similarly, divided boards increase the likelihood that shareholder activists can find support—or at least sympathy—among some board members, which increases the odds those activists can obtain backing for their demands.

We analyze whether board demographic diversity is part of the opportunity structure for shareholder activism by focusing on activist hedge funds (DesJardine et al. 2021b). Because demographically diverse boards tend to act less swiftly and unitedly than more homogenous boards, we theorize that activist hedge funds are more likely to target firms with governance and performance problems when those firms’ boards are more demographically diverse. We thereby posit that problems and opportunities interact. Whereas problems are the underlying reason why activist hedge funds consider targeting a firm in the first place, opportunities influence whether activist hedge funds will be able to resolve those problems and realize the potential for subsequent share price returns. Our reasoning is supported by interviews with senior managers at activist hedge funds in which they acknowledged that they or their peers see board demographic diversity as an opportunity to influence firms. A portfolio manager at an activist hedge fund, for example, noted that “there will be more room for conflict on diverse boards, which can create opportunities for activists to have influence, by sowing further division and finding support from some directors on those boards.”

To further support our theory, we explore the opportunity recognition (Baron 2006) of activist hedge funds in two ways. First, based on the insight that experience increases the likelihood that actors recognize opportunities (e.g., Grégoire et al. 2010), we posit that dedicated activist hedge funds, which engage solely in activism, are most likely to recognize board demographic diversity as an opportunity, whereas this is unlikely for nondedicated activist hedge funds, which engage in activism only occasionally and are thus less experienced. We, therefore, hypothesize and find that dedicated activist hedge funds, instead of nondedicated activists, target firms with problems based on their boards’ demographic diversity. Second, based on the insight that the identification of an opportunity is a “conjecture” that can be either “correct” or “incorrect” (Shane and Venkataraman 2000, p. 220), we argue that activist hedge funds correctly identify this opportunity, which means that demographically diverse boards respond to activism campaigns in ways that are likely to benefit activist hedge funds. Specifically, we hypothesize and find that, relative to firms with less demographically diverse boards, firms with demographically diverse boards tend to (1) react less swiftly to hedge fund campaigns and (2) be more acquiescent in the aftermath of hedge fund campaigns.

Our study contributes to two streams of research. First, we contribute to research on shareholder governance by showing that shareholder activists target firms not only based on the problems they exhibit but also, on the opportunities activists can exploit. Specifically, we show that the problems of firms and the influence opportunities of activists have a multiplicative effect on one another. Future research can explore other elements of the corporate opportunity structure for shareholder activism and thereby, not only reconcile inconsistent findings on the antecedents of shareholder activism but examine when the disciplinary effect of shareholder activism goes awry, such as when firms avoid building demographically diverse boards because some activists exploit diversity as a targeting opportunity. Second, we contribute to research on board governance by problematizing the widespread assumption that a business case for diversity is sufficient to align shareholders behind board demographic diversity. We show that some activist shareholders evaluate board demographic diversity not in terms of its social or economic value—whether it resolves gender imbalance or creates returns for shareholders—but in terms of the opportunities it creates to influence firms.

Theory and Hypotheses

Why Opportunities Matter for Shareholder Activism

Existing research shows that shareholders activists target firms beset with governance and performance problems with the view of increasing the firms’ share price by “fixing” these problems. First, governance problems raise concerns among shareholder activists about mismanagement and the potential for managers to undertake self-serving actions at the expense of shareholders (Boyson and Mooradian 2011). By targeting firms with governance problems, shareholder activists can increase pressure and oversight on management to pay undivided attention to shareholders. Second, performance problems can attract the attention of shareholder activists (Brav et al. 2008). Shareholder activists tend to target firms with lower financial performance in hopes that they can initiate strategic and operational changes, such as spinning off divisions or cutting costs, that improve financial performance and create shareholder value. Bolstered by the idea that shareholder activists target firms based on problems, prior research has argued that shareholders activism has a disciplinary effect on managers, whereby managers are motivated to resolve problems in their companies to avoid becoming subject to activist intervention (Fama and Jensen 1983, Gantchev et al. 2019).

Research on corporate opportunity structures (Briscoe et al. 2014) suggests that shareholder activists target firms not only based on problems but also, based on opportunities. The corporate opportunity structure consists of organizational features that make a firm a more attractive target for activism because activists anticipate that these features increase the likelihood they can successfully influence the firm (King 2008). The corporate opportunity structure for shareholder activism is likely to differ from that of social activism because interactions between firms and social activists unfold differently than interactions between firms and shareholder activists. For example, an opportunity exists for social activists to target large firms because campaigns against large firms are more likely to attract media attention that benefits social activists (King and Soule 2007, Lenox and Eesley 2009), whereas an opportunity exists for shareholder activists to target small firms because acquiring sufficient shares to yield influence is less costly with smaller firms (Boyson and Mooradian 2011, Klein and Zur 2011). Examining whether and how opportunities influence the targeting decisions of shareholder activists cannot only clarify the antecedents of shareholder activism but also, illuminate the merits of the disciplinary effect of shareholder activism because targeting based on opportunities will not motivate managers to resolve problems but incentivize them to prepare for confrontations with activists in ways that distract them from resolving those problems.

We explore the corporate opportunity structure of shareholder activism by focusing on activist hedge funds. Activist hedge funds, which are shareholders who buy shares in firms whose management and operations they want to influence, are becoming increasingly influential (Ahn and Wiersema 2021, DesJardine et al. 2022b). In 2022, 967 firms were targeted by at least one activist hedge fund, 32% of which had market capitalizations exceeding U.S. $10 billion (Insightia 2023). Highlighting their important role in corporate governance, prior studies reveal that activist hedge funds can intervene dramatically in corporate affairs: for example, by influencing firm competitive actions (DesJardine et al. 2021b), asset divestitures (Chen and Feldman 2018), growth and downsizing activities (Shi et al. 2020), and firm performance (DesJardine and Durand 2020).

Background interviews we conducted suggest that activist hedge funds clearly think about opportunities.1 A portfolio manager at an activist hedge fund, for instance, noted that “as an activist, you’re looking for every opportunity you can to tip the scales in your favor.” Similarly, a founder and managing partner at an activist hedge pointed to a framework with which “pretty much any activist would agree,” the “value, plan, path” framework pioneered by activist hedge fund Starboard Value (2018), which highlights the importance of both problems (as activist hedge funds must identify firms at which they can unlock “value” through a clear “plan”) and opportunities (as a “path” must exist to influence those firms).

Why Hedge Fund Campaigns are More Likely to Succeed When Boards are Slow and Divided

As the overseer of corporate affairs, boards play a crucial role in deciding how firms respond to activists (Benton 2017, EY 2021, Canals 2022). When activist hedge funds launch a campaign, boards often enter direct negotiations with activists, which are influenced by the possibility that activist hedge funds can at any point bring their demands to vote before shareholders in a formal proxy fight. In negotiations between activist hedge funds and boards, the bargaining power of each side will be determined by the voting support it believes it has from other shareholders. If the targeted firm believes that its other shareholders will support the activist, the two sides may reach a settlement in which the firm gives in to the demands of the activist. If the targeted firm believes that it has the support of its other shareholders, the negotiations will likely fail, and the activist hedge fund may then initiate a proxy fight. Directors are often directly involved throughout this process by defining a firm’s defense strategy and meeting with various shareholders to solicit their support against the activist (Castañón Moats et al. 2021). Given these dynamics, we expect that hedge fund campaigns are more likely to succeed when target firms have boards that are (1) slow and (2) divided. We define a hedge fund campaign as successful when an activist hedge fund elicits changes in companies that benefit the activist hedge fund financially.

Hedge funds campaigns are more likely to succeed when the firms they target have boards that react slowly to their demands. Activist hedge funds “often begin planning their campaigns as much as a year in advance of the annual meeting,” which can include “coordinating with other hedge funds” (Latham & Watkins 2015, p. 1). To catch up with activist hedge funds, boards must “mount a defense quickly” (Lipton 2019) and create a “well-developed response plan” to “control its own narrative” (EY 2021, p. 5). If boards react slowly, the activist may get an upper hand in convincing other shareholders of its narrative, making it harder for the board to secure sufficient shareholder support and forcing a settlement with the activist. Our interviews show that activist hedge funds are aware that slow decision making by boards increases the likelihood that their campaigns will succeed. A portfolio manager at an activist hedge fund acknowledged that “we try to catch them off guard, because that allows for a faster, easier campaign to get what we want.” Similarly, the head of research at another activist hedge fund noted that “lightning strike activists … are very tactical” about timing and noted that “I’ve heard stories from companies we own where the CEO [chief executive officer] got a call on Sunday night saying, ‘You’ll have a letter Monday morning, we’re going public.’”

Hedge fund campaigns are also more likely to succeed when boards are divided. The lack of unity in divided boards increases the likelihood that activist hedge funds can find support—or at least sympathy—from some board members, which increases the odds an activist can reach advantageous settlement terms with the board (Bebchuk et al. 2020). A divided board is likely to be less effective in mobilizing support from shareholders because board members may communicate different views on the merits of the activist’s demands. The idea that divided boards create opportunities for activist hedge funds resonates with the insight from social movement research that divided elites create influence opportunities for social movements (McAdam 1996, Tarrow 2022). Divisions between hard-liners and soft-liners, for example, have created influence opportunities for opposition movements in autocratic regimes (Bermeo 1997, Tarrow 2022). Our interviews document that activist hedge funds know that their campaigns will be more successful when boards are divided. The head of research we interviewed reflected that “on a board that’s been together for a while, they’re used to closing ranks and they’re probably good at being responsive and speaking up for each other.” By contrast, as noted by a portfolio manager at an activist hedge fund, “when boards aren’t cohesive, when everybody doesn’t look the same way and think the same way, then you’re going to be better positioned as an activist to find some support for your ideas.”

Although activist hedge funds are aware that their odds of success increase when boards are slow and divided, it is hard for them to anticipate which boards will act slowly and lack unity. Most directors will not be willing to share information about internal dynamics with activist hedge funds out of fear of being sued, jeopardizing social capital, and being ousted from the corporate elite (Wiesenfeld et al. 2008). Given this information asymmetry, we argue that activist hedge funds will draw on observable indicators to assess whether boards are likely to respond swiftly and unitedly. In what follows, we argue that board demographic diversity could provide a valuable indicator for activist shareholders because it is highly visible and thereby, “provides investors with information” on organizational features that are “not readily observable in circumstances of information asymmetry” (Perrault 2015, p. 158). In line with this argument, a vice president at a firm that provides data to investors explained that “I have hedge fund activists call me and ask for all sorts of indicators about the board of directors: who they are, what are their backgrounds, how old they are, their gender and citizenship” and that activists are “using this data to inform their investment decisions and campaigns.”

Board Demographic Diversity as an Opportunity for Activist Hedge Funds

Research shows that demographic diversity slows down decision making and reduces group unity. Decision making tends to be slower because individuals in diverse groups spend more time anticipating and considering their own and others’ perspectives (Watson et al. 1993, Loyd et al. 2013) and because “diversity could reduce trust and increase the time for reaching consensus” (Andrevski et al. 2014, p. 824). Similarly, diversity can undermine unity because diverse individuals are sometimes less inclined to like, to show commitment to, and to be willing to coordinate with demographically different others (Tsui et al. 1992). Consequently, studies find that diverse individuals find it harder to act in concert (e.g., Acharya and Pollock 2021) because they have different communication styles (Gibson 1997), different norms of information sharing (Goodman et al. 1987), and different approaches to group work (Gibson and Zellmer-Bruhn 2001). Prior research suggests that these dynamics also affect boards (Triana et al. 2014).

That demographic diversity slows down boards and reduces their unity is advantageous in many circumstances but not all. Under “normal” circumstances, slow, critical, and investigative decision-making processes (e.g., Amason 1996) allow diverse boards to develop more innovative strategies (Miller and Triana 2009) and spot more promising opportunities (Estélyi and Nisar 2016). Compared with more homogenous boards, demographically diverse boards are also more effective in monitoring management (e.g., Bear et al. 2010), spend more time discussing strategic issues (Harrison et al. 2002), and are less influenced by each other’s perspective (Horwitz and Horwitz 2007). Under high-strain circumstances, however, slow decision making and a lack of unity can become a liability (Triana et al. 2014). Jung et al. (2023, p. 657) find that when firms are underperforming, “the urgent need to build consensus” among board members around a new strategy motivates companies to avoid appointing demographically diverse board members. We argue that confrontations with activist hedge funds constitute one of the high-strain circumstances in which board demographic diversity can become a liability for firms.

We posit that activist hedge funds see board demographic diversity as an opportunity to increase the likelihood that their campaigns will succeed because these activists anticipate that demographically diverse boards will (1) react more slowly and (2) act with less unity than more homogenous boards. Diverse boards will, for example, often include diverging views on what risks are acceptable because men are more likely to be overconfident than women (Croson and Gneezy 2009), members of racial minorities tend to be more accepting of risk than members of racial majorities (Glass and Cook 2020), and people from different countries have different preferences for uncertainty avoidance (Hofstede 1984). In confrontations with activist hedge funds, these diverging views can slow down decision making and undermine unity within demographically diverse boards. Our interviews document that activist hedge funds see this opportunity. Concerning response speed, a portfolio manager at an activist hedge fund noted, “if a board is made up of a bunch of diverse individuals that we think are going to find it harder to work together and will make decisions more slowly, then that’s going to tip things in our favor.” Similarly, a corporate defense lawyer reflected that members of demographically diverse boards are “going to see things from different perspectives and it’s going to take longer for them to have a discussion on [activist demands].” Concerning unity, a managing director at an activist hedge fund reflected, “you might assume this board you can penetrate because it has people from all over the map on it.” Similarly, a cofounder and chief investment officer at an activist hedge fund explained, “when you operate in a good old boys’ system, it’s easier if you’re the CEO to corral your guys around—when you have a more diverse board, it can be a lot more difficult to do that.”

We argue that problems and opportunities interact to shape the attractiveness of firms for hedge fund targeting. Problems are the reason why activist hedge funds consider targeting a firm in the first place (Denes et al. 2017). Without problems to resolve, activist hedge funds could not profit from subsequent increases in share prices; so, targeting a firm would be financially fruitless. Opportunities, in turn, influence whether activist hedge funds will be able to successfully intervene in a target firm to resolve those problems and thereby, realize the potential for subsequent increases in share prices. Stated differently, problems determine the size of the potential payout for activist hedge funds, whereas opportunities determine the likelihood of the potential payout. A portfolio manager at an activist hedge fund expressed this idea succinctly: “If two companies are performing equally badly, and you see one has a diverse set of directors that might make decision making harder and alliances between directors weaker, that can make activism easier—easier to convince some directors to come to your side and part against the majority.” We, therefore, hypothesize that board demographic diversity, by slowing down board decision making and hampering unity, amplifies the likelihood that activist hedge funds target firms with governance and performance problems.

Hypothesis 1(a).

Board demographic diversity strengthens the positive relationship between governance problems and the firms likelihood of being targeted by an activist hedge fund.

Hypothesis 1(b).

Board demographic diversity strengthens the positive relationship between performance problems and the firms likelihood of being targeted by an activist hedge fund.

Exploring the Opportunity Recognition of Activist Hedge Funds

We extend our main theory of why board demographic diversity presents an opportunity for activist hedge funds (H1a/1b) by exploring the opportunity recognition of activist hedge funds. Opportunity recognition (Baron 2006, Grégoire et al. 2010) is necessary because board demographic diversity will only affect targeting decisions if activist hedge funds recognize that their likelihood of success increases when targeting demographically diverse boards. Specifically, we explore whether experienced activist hedge funds are likely to recognize board demographic diversity as an opportunity (Hypotheses 2(a) and 2(b)) and whether this opportunity is a “correct” conjecture in the first place, in the sense that demographically diverse boards respond to hedge fund campaigns in ways that are likely to benefit activist hedge funds (Hypotheses 3(a) and 3(b)).

The first way in which we explore the opportunity recognition of activist hedge funds stems from the insight that experience influences whether actors recognize opportunities (e.g., Grégoire et al. 2010, Vandor and Franke 2016). In the context of entrepreneurial opportunities, Baron and Ensley (2006) show that experienced entrepreneurs are more likely to recognize promising new venture opportunities than less experienced entrepreneurs. Applied to our context, we distinguish between dedicated activist hedge funds, which derive their investment returns solely from activism, and nondedicated activist hedge funds, which engage in activism only occasionally and make most of their profits from nonactivist investment strategies, such as shorting stocks or trading currencies. We posit that dedicated activist hedge funds have gained more activism experience than nondedicated activist hedge funds because the former have launched more activism campaigns from which they could gain experience and/or have recruited partners or employees with more substantial activism experience that they can leverage to analyze a “company’s defenses and vulnerabilities” (Latham & Watkins 2015).

Extending our prior reasoning, dedicated activist hedge funds will have experienced numerous situations in which demographically diverse boards have responded slowly or lacked unity and are, therefore, well positioned to recognize that board demographic diversity could create an opportunity to successfully influence firms. Accordingly, when they identify firms with governance or performance problems, dedicated activist hedge funds will be likely to target those firms when they also have more demographically diverse boards. In contrast, because of their limited experience, nondedicated activist hedge funds are unlikely to recognize that demographically diverse boards could make it easier for activists to succeed in their campaigns and thus, may not target firms with governance and performance problems based on their boards’ diversity. We, therefore, hypothesize the following.

Hypothesis 2(a).

For dedicated activist hedge funds, board demographic diversity strengthens the positive relationship between governance problems and the firms likelihood of being targeted.

Hypothesis 2(b).

For dedicated activist hedge funds, board demographic diversity strengthens the positive relationship between performance problems and the firms likelihood of being targeted.

The second way in which we explore the opportunity recognition of activist hedge funds builds on the insights from entrepreneurship research that an opportunity is a “conjecture [that] is acted upon” and that this conjecture can be either “correct” or “incorrect” (Shane and Venkataraman 2000, p. 220). In our context, activist hedge funds would act based on a “correct” conjecture if targeting firms with demographically diverse board increases the likelihood that their campaigns succeed and act based on an “incorrect” conjecture if campaign success does not increase. We expect that, on average, activist hedge funds act based on correct conjectures. Compared with other types of investors, activist hedge funds are especially driven to increase their targeting success because they operate in a competitive environment, charging wealthy clients high performance-based fees in return for the possibility of outsized investment returns (Brav et al. 2015). To increase their odds of generating those returns, activist hedge funds will try to improve their targeting by learning from their own campaigns and the campaigns of other activists. Consequently, if board demographic diversity had little influence on the success of hedge fund campaigns, activist hedge funds would likely learn this and not target firms based on their boards’ demographic diversity.

Based on our theorizing about how hedge fund campaigns are more likely to succeed when boards are slow and lack unity and supported by our interview evidence, we expect that activist hedge funds will form two conjectures; compared with firms with less demographically diverse boards, firms with demographically diverse boards will (1) react less swiftly and (2) be more acquiescent in the aftermath of hedge fund campaigns (for how activism can change firm behavior in subsequent phases, see McDonnell et al. 2015). By acquiescence, we refer to firms’ tendencies to undertake activities that are likely to benefit activist hedge funds by increasing the share price in the short term. We expect that both conjectures are correct because response speed and firm behavior in the aftermath of hedge fund campaigns are relatively easy to observe and systematically tracked by data providers (by contrast, how unitedly boards respond to activist campaigns is much harder to observe, especially in the campaigns of other hedge funds). If these conjectures are correct, we should observe that firms with demographically diverse boards (1) react more slowly to hedge fund campaigns and (2) are more acquiescent to hedge funds campaigns relative to firms with less demographically diverse boards.

Hypothesis 3(a).

Board demographic diversity is positively associated with firmsresponse time to hedge fund campaigns.

Hypothesis 3(b).

Board demographic diversity is positively associated with firmsacquiescence to hedge fund campaigns.

Methods

To test our hypotheses, we use archival data spanning the years 2009–2018. We selected the firms that our sample comprises from ExecuComp, whose database contains all firms that have been listed on Standard & Poor’s Composite 1500 Index. To identify whether a firm in our sample was targeted by an activist hedge fund and if so, when, we relied on Insightia (formerly Activist Insight), which uses regulatory filings, activist websites, media accounts, and professional relationships to provide comprehensive information on activist hedge funds and their campaigns, including campaign demands and outcomes (Chen and Feldman 2018, de Figueiredo et al. 2019, Wiersema et al. 2020). Information on target firms’ response strategy was collected from FactSet. We collected financial and accounting data from S&P’s Compustat database and data on directors from BoardEx and Institutional Shareholder Services (ISS) Directors. Additionally, we collected data on institutional investor ownership from Thomson Reuters 13(F), data on executive compensation from ExecuComp, and data on foreign sales from Thomson Reuters Worldscope. From this sample, we excluded firms for which we did not have complete data for all the variables in our study.

Dependent Variables

Hypotheses 1(a) and 1(b): Activist Hedge Fund Targeting.

Our focal dependent variable for Hypotheses 1(a) and 1(b) identifies instances where firms are targets of hedge fund activism. Our final sample comprises 8,744 firm-year observations on 1,434 United States-based firms. Of the firm-years in our sample, 354 correspond to instances of hedge fund activism involving 253 firms targeted by 113 activist hedge funds. Of these, 190 firm-years correspond to targeting by a single activist campaign in a year. The remaining 164 firm-years correspond to targeting by multiple activist campaign in a year. Hedge fund activism equals one for firm-years associated with hedge fund activism and zero otherwise.

Hypotheses 2(a) and 2(b): Dedicated Hedge Fund Targeting.

To test Hypotheses 2(a) and 2(b), we follow Insightia to identify dedicated activist hedge funds. According to Insightia, dedicated activist hedge funds are hedge funds “who proactively and systematically identify” promising companies to target and for “whom activist investments typically form a significant majority of their investment portfolios” (Activist Insight 2018, p. 3). Nondedicated activist hedge funds, by contrast, rely more significantly on nonactivist investment strategies (e.g., long-short strategies) to derive their portfolio returns and thus, target firms more irregularly. Following Insightia’s classification, we construct a nominal variable in which a firm-year receives a value of zero if it is associated with no hedge fund activism, a value of one if it is associated with targeting by a nondedicated activist hedge fund (224 firm-years), and a value of two if it is associated with targeting by a dedicated activist hedge fund (130 firm-years).

Hypothesis 3(a): Firm Response Time.

The dependent variable for Hypothesis 3(a) is the firms’ response time, which is measured as the number of days that have lapsed between the announcement date of a hedge fund campaign and the date the firm responds to the activist with a formal decision of how to proceed (e.g., settle with the activist or proceed to a proxy vote), as reported in Insightia. The longer a board takes to deliberate and respond to an activist, the longer the firms’ response time and the higher the value.

Hypothesis 3(b): Target Acquiescence.

Firm responses to hedge fund activism vary based on their acquiescence to activists’ underlying concern, namely in trying to bolster shareholder returns (Boyson and Pichler 2019). Acquiescent firms will undertake actions that aim to improve shareholder value in the aftermath of hedge fund activism. Based on prior research (Denes et al. 2017, Chen and Feldman 2018) and information from FactSet, we code target acquiescence as one when a target firm undertakes clear actions that aim to enhance shareholder value (e.g., repurchasing shares, issuing dividends, or divesting business units) in response to hedge fund activism and zero otherwise.

Governance and Performance Problems

Based on prior research, our study considers two types of problems that can attract the attention of activist hedge funds: governance and performance problems.

Starting with governance problems, we draw on existing governance research to identify four of the most prominent indicators of governance problems. We use these indicators to create a standardized index. The first indicator is CEO duality, which characterizes CEOs who also serve as the board’s chairperson. CEOs with a dual role are less concerned with controlling the board and have greater discretion to pursue their own agenda, which may weaken the board’s monitoring effectiveness (for a review, see Krause et al. 2014). CEO duality equals one if a CEO is also the chair of the board and zero otherwise. The second indicator pertains to directors appointed during CEO tenure, which measures the ratio of directors appointed during a CEO’s tenure to the total number of directors on the board (Westphal 1999). Directors appointed during a CEO’s tenure tend to be loyal and committed to the CEO and are, therefore, likely to support the CEO’s decisions with less concern about how those decisions may impact shareholder value (Coles et al. 2014). The third indicator is the average director tenure, measured as the average duration of tenure on a board of directors. Long director tenure can be detrimental to board monitoring (Huang and Hilary 2018). The fourth indicator is board busyness (Ferris et al. 2003, Field et al. 2013), measured as the ratio of the number of directors who hold three or more directorships to the total number of directors on a board (Fich and Shivdasani 2006). Directors who serve on many boards are often less effective monitors (Fich and Shivdasani 2006). We standardize each of these four indicators and add them together to obtain an overall index of governance problems.

We measure performance problems as a firm’s stock-market performance relative to that of its industry peers. We do not focus on accounting underperformance (e.g., return on assets) because stock performance can more directly influence activist hedge funds’ financial gains than accounting performance. We measure performance problems as the average shareholder returns of a firm’s industry peers (based on the Fama–French 48 industry classification) minus the firm’s shareholder returns, where shareholder returns equal the ratio of the share price at the end of a given year plus the value of dividends issued during that year to the share price at the beginning of the year minus one. Higher values indicate more severe performance problems.

Board Demographic Diversity

Social psychology research finds that people categorize one another based on their demographic characteristics (Messick and Mackie 1989, Fiske and Neuberg 1990, Stangor et al. 1992). Such social categorization emphasizes differences between groups and can, therefore, adversely affect group unity (Hogg and Terry 2000, Harrison and Klein 2007). Specifically, research has found consistent evidence that gender, race, and nationality promote social categorization (Messick and Mackie 1989, Fiske and Neuberg 1990, Stangor et al. 1992, Zellmer-Bruhn et al. 2008) and thus, undermine group unity (Harrison et al. 1998, Chatman and Flynn 2001, Dahlin et al. 2005). Following this research, we assess board demographic diversity using three characteristics of directors: (1) gender diversity, (2) racial diversity, and (3) nationality diversity.

Gender diversity and racial diversity are especially fitting as gender and race are “known as two of, if not the most, powerful attributes that people use to categorize others because their boundaries are visible and, more importantly, perceived to be fixed” (Jung et al. 2023, p. 661). In support, researchers have found that social categorizations on gender and race tend to resist manipulations designed to decrease such categorization (Hewstone et al. 1991, Stangor et al. 1992). Nationality is often aligned with people’s countries of origin and serves as a superordinate determinate of identity that can reflect significant differences in working styles and values and thus, contribute to social categorizations that challenge group unity (Zellmer-Bruhn et al. 2008). Dahlin et al. (2005, p. 1111), for instance, explain that “diversity in nationality might segment a team and interfere with team members’ ability to work together effectively.” Nationality has the additional advantage of proxying directors’ ethnic diversity, which can also affect social categorization (Dinesen et al. 2020), but is very hard to measure given data on directors’ ethnicities are largely unavailable.2

Gender, race, and nationality are categorical variables and thus, require that we operationalize diversity in terms of attribute variety (Harrison and Klein 2007). Following the suggestion of Harrison and Klein (2007), we measure gender, racial, and nationality diversity using the Blau index (1pi2), where pi stands for the proportion of directors with specific attributes (Blau 1977). A value of zero represents complete homogeneity, whereas higher values indicate greater diversity (Harrison and Sin 2006). To measure gender diversity, we used data derived from BoardEx to calculate the percentage of female directors and the percentage of male directors, and then, we applied the Blau index. To measure racial diversity, we collected information on the directors’ race from the ISS Directors database. For directors with missing race information in the database, we follow existing research to classify their race based on their names. Following the ISS Directors, we classified the directors in our sample into the categories White, African American, Asian, Hispanic, and “other.” We then applied Blau’s index to all five categories to calculate racial diversity within a board. To measure nationality diversity, we collected directors’ citizenship information from BoardEx. BoardEx provides around 44% of directors’ country of citizenship information. If such information is not available, we follow Barrios et al. (2022) to use their country of first board appointment reported in BoardEx to proxy their country of citizenship. Based on such information, we calculated nationality diversity using the Blau index. Having calculated the three components of demographic diversity, we standardized the respective indexes (with a mean of zero and a standard deviation (SD) of one) and added them together to measure board demographic diversity.

Control Variables

Hypotheses 1(a), 1(b), 2(a), and 2(b) concern the relationship between a firm’s board demographic diversity and hedge fund activism, and we use a firm-year–level data set to test these hypotheses. We include the following variables that can influence the likelihood of a firm being chosen as a hedge fund activism target. We control for firm size (measured as the natural logarithm of a firm’s total assets) because activist hedge funds tend to target smaller firms (Greenwood and Schor 2009). Firms with poor accounting performance are also more likely to become targets of hedge fund activism, so we also include firm accounting performance. Here, we control for return on equity (ROE), measured as the ratio of operating income after depreciation to shareholder equity. We also control for a firm’s value using market-to-book ratio, which is measured as the ratio of market value to book value. In addition, firms with slower revenue growth may be faced with a higher likelihood of hedge fund activism; thus, we control for revenue growth rate, measured as the ratio of the current year’s revenues to the prior year’s revenues minus one. Because activist hedge funds are more likely to target firms with unconstrained resources (Clifford 2008), we control for a firm’s slack resources. We follow existing research (Bromiley 1991, Chen 2008, Tyler and Caner 2016) to create a composite slack index that comprises three types of slack: available slack (measured as the ratio of current assets to current liabilities), absorbed slack (measured as the ratio of working capital to sales), and potential slack (measured as the ratio of total equity to total debt).

Activist hedge funds may target firms with higher levels of corporate investment to prevent managers from engaging in excessive spending or “empire building.” We, therefore, control for corporate-level investment using the sum of expenditures on research and development (R&D), new capital, and acquisitions. We also include business diversification as a control because activist hedge funds may target highly diversified firms (Chen and Feldman 2018). We used two components of an entropy measure to calculate business diversification: related and unrelated diversification (Hoskisson et al. 1993). Unrelated diversification arises from business units that operate in industry groups with different two-digit Standard Industrial Classification (SIC) codes, with the firm’s total assets as the reference. Related diversification arises from business units that operate in the same two-digit SIC industry group. We also control for payout ratio, measured as the ratio of total dividends to net income, because firms with lower payout ratios are more likely to be targeted by an activist hedge fund (Brav et al. 2008). Research also suggests that a firm’s corporate social responsibility (CSR) can influence a firm’s likelihood of being targeted by activist hedge funds (DesJardine et al. 2021a). Therefore, we control for a firm’s CSR using data from KLD. Following DesJardine et al. (2021a), we focus on the environment, community, diversity, employee relations, and human rights dimensions. We exclude categories related to board demographic diversity in measuring CSR. Because the number of categories covered for each dimension varies over time and comparing counts of these scores across years is not feasible (Deng et al. 2013), we divide the total number of strengths by the respective number of strength indicators for each dimension to measure CSR for each firm-year. We also control for foreign sales (measured as the ratio of revenues from foreign sales to total revenues) because the degree of a firm’s international diversification may increase the likelihood of hedge fund activism.

Institutional ownership can also affect a firm’s probability of being targeted because activist hedge funds often seek the support of other institutional investors (Appel et al. 2019). For that reason, we control for the percentage of shares owned by two types of institutional investors (Bushee 1998, 2001) denoted by dedicated institutional ownership and transient institutional ownership (Shi et al. 2017, Connelly et al. 2019). We also control for board size (measured as the number of board members that a firm has) because larger boards may slow down decision making and harm board effectiveness (Judge and Miller 1991). Research also shows that CEOs who own more equity have greater ownership power (Finkelstein 1992). As this could reduce the likelihood of hedge fund activism, we control for CEO equity ownership, which equals the percentage of firms’ outstanding shares that are owned by the CEO. The next control, CEO option pay, is measured as the ratio of a CEO’s option pay to that CEO’s total compensation. We included this control because activist hedge funds tend to tie a CEO’s compensation to the firm’s performance, so fewer options mean that a firm is less likely to perform well in the future (Ertimur et al. 2011). We also control for CEO tenure because long-tenured CEOs tend to be more entrenched, making their firms more attractive to activist hedge funds. We measure CEO tenure as the number of years since an executive was appointed CEO. Lastly, we control for industry fixed effects based on the Fama–French 48 industry classification and year fixed effects in all regressions.

Hypothesis 3(a) (Hypothesis 3(b)) tests the relationship between board demographic diversity and response time (target acquiescence), and we use a campaign-level data set to test these hypotheses. In addition to all preceding control variables, we add the following controls. We include dedicated activist hedge funds because dedicated activist hedge funds may pose a greater threat to target firms and affect response time and response strategy. We control for governance problems and performance problems because firms with such problems are anticipated to face stronger pressures to respond to activists. We also control for campaign demand fixed effects. Insightia classifies campaign demands into board activism, business strategy, balance sheet activism, acquisition activism, renumeration, and other.

Analysis

We used probit regressions to test Hypotheses 1(a) and 1(b) because these models are appropriate for modeling the probability of two alternatives (Hoetker 2007). We did not include firm fixed effects in these models because doing so would exclude any firms associated with a time-invariant dependent variable, namely any firms that had not been targeted by an activist hedge fund, which would create selection bias. The analyses for Hypotheses 1(a) and 1(b) are conducted at the firm-year level. Because the dependent variable for Hypotheses 2(a) and 2(b) is nominal, we tested these hypotheses using multinomial logit models that allow us to estimate targeting decisions by both dedicated and nondedicated activist hedge funds using the firm-years in which no hedge fund activism occurred as the baseline. We retain firm-years in which no activism occurred because our goal is to estimate what factors affect dedicated activist hedge funds’ targeting decisions, and firm-years with no activism targeting serve as counterfactuals.

For Hypotheses 1(a), 1(b), 2(a), and 2(b), although we control for a host of variables, omitted variable bias could exist if there is unobservable heterogeneity that simultaneously influences board demographic diversity and our set of hedge fund activism dependent variables. We address this concern by implementing the control function approach (Heckman and Robb 1985), which uses a first-stage regression to model an endogenous explanatory variable as a function of exogenous instruments and then, controls for the residual created from the first stage in subsequent second-stage regressions. Compared with two-stage least squares, the control function approach is more generalized because it can be used in nonlinear models, including the probit regressions employed in our analyses (Wooldridge 2010). Given that board demographic diversity is our key variable of interest, we treated it as endogenous, which required identifying relevant and exogenous instruments. Relevance suggests that instrumental variables affect the predictor variable, whereas exogeneity requires that the instrumental variables do not directly influence the dependent variables. We identified two instruments.

The first instrument is percentage of foreign-born residents, which equals the number of residents in a U.S. county born outside the United States divided by the total number of county residents. Firms located in counties with a higher proportion of foreign-born residents are more likely to promote demographic diversity on their board of directors resulting from isomorphic community pressures (Pugh et al. 2008, Marquis and Battilana 2009). However, the percentage of foreign-born residents at the county level is unlikely to directly influence a firm’s likelihood of being targeted by an activist hedge fund. We collected data on county demographics from the U.S. Census Bureau.

The second instrument is social capital at the county level, which refers to the norms and networks facilitating collective action (Woolcock 2001). Firms located in counties with a higher social capital index are more likely to have an inclusive culture and therefore, a higher level of board demographic diversity. However, social capital at the county level should not exert a direct influence on a firm’s likelihood of being targeted by an activist hedge fund. We used the county-level social capital index created by the Northeast Regional Center for Rural Development at the Pennsylvania State University. Rupasingha et al. (2006) provided a detailed description of this index, which is derived from principal component analysis and includes four factors: (1) response rates in U.S. Census surveys, (2) voter turnout in presidential elections, (3) the total numbers of 10 types of social organizations, and (4) the total numbers of nonprofits.

The dependent variable for Hypothesis 3(a), response time, is the number of days that pass until a company responds to a hedge fund campaign. Although Tobit models and linear regressions have been used to model time use variables, such models assume normality. Because of the challenges with the normality assumption, Brown and Dunn (2011) suggest modeling time use data using generalized linear models (GLMs) with a Poisson-gamma distribution, which are sounder theoretically and empirically than either Tobit or linear models for this type of analysis. Theoretically, the Poisson-gamma model is built on “a flexible three-parameter distribution that can model the episodic nature” of time use data (Brown and Dunn 2011, p. 531). Empirically, the Poisson-gamma model is superior to Tobit and linear models in terms of residual analyses, model interpretation, and model performance (Brown and Dunn 2011).3 Because the variance and mean of response time differ (i.e., overdispersed), violating the assumption of the Poisson distribution, we also test this hypothesis using GLMs with a negative binomial distribution. We use probit regressions to test Hypothesis 3(b) because the dependent variable, target acquiescence, is a dummy. The analyses for Hypotheses 3(a) and 3(b) are conducted at the activist-campaign level to allow us to model the nature of responses to different campaigns.

We used Heckman selection models to address the possibility for sample selection bias associated with testing Hypotheses 3(a) and 3(b) (Heckman 1979). Sample selection may occur if unobservable heterogeneity drives both whether a firm-year is associated with hedge fund activism and the firms’ response. In a first-stage probit regression, we estimated the likelihood of a firm being a target of hedge fund activism in a year and then, calculated the inverse Mills ratio based on the probit regression, which we control for in all second-stage models. The dependent variable in the first stage receives a value of one if a firm is a target of hedge fund activism and zero otherwise. We included all control variables in the model as well as industry and year fixed effects. We followed Shi et al. (2020) to also include the percentage of firms that were targets of hedge fund activism in the same metropolitan statistical area (MSA), MSA activism target percent, as an exclusion restriction. Hedge fund activism is costly, and activist hedge funds may target firms located in the same MSA to reduce campaign costs. Meanwhile, the percentage of firms targeted for activism in the same MSA should not influence a board’s decision making and firms’ response time and response strategy. In untabulated results, we found that the coefficient for MSA activism target percent is 6.78 (p < 0.01); however, when we used MSA activism target percent to predict response time, its coefficient is positive but statistically not significant. Based on the first-stage probit regression, we calculated the inverse Mills ratio and controlled for it in the second-stage regression.4

We cluster standard errors by firm to handle potential correlations among residuals within firms (Petersen 2009). All independent, moderating, and control variables are measured in year t − 1, and the dependent variable is measured in year t.

Results

Tables 1 and 2 display the descriptive statistics and correlations for all variables used in testing Hypotheses 1 and 2 and in testing Hypothesis 3, respectively.

Table

Table 1. Descriptive Statistics and Correlations for Variables Used in Hypotheses 1 and 2

Table 1. Descriptive Statistics and Correlations for Variables Used in Hypotheses 1 and 2

Variable1234567891011121314151617181920
1Hedge fund activism1.00
2Board demographic diversity0.041.00
3Governance problems0.000.161.00
4Performance problems0.050.020.051.00
5Firm size0.010.400.310.041.00
6Return on equity−0.020.070.04−0.010.101.00
7Market-to-book ratio−0.05−0.02−0.03−0.17−0.250.101.00
8Revenue growth rate−0.04−0.08−0.09−0.11−0.110.020.191.00
9Slack resources−0.02−0.16−0.120.03−0.44−0.070.310.121.00
10Corporate investment intensity0.02−0.05−0.040.02−0.01−0.100.080.180.171.00
11Business diversification0.030.160.160.030.260.04−0.13−0.10−0.17−0.111.00
12Payout ratio−0.020.060.060.000.120.030.00−0.06−0.11−0.060.031.00
13CSR0.040.320.150.000.420.080.05−0.09−0.110.030.160.071.00
14Foreign sales (%)0.020.180.010.030.040.010.10−0.020.230.080.16−0.070.111.00
15Dedicated institutional ownership0.040.080.01−0.010.09−0.010.10−0.08−0.07−0.010.000.060.13−0.021.00
16Transient institutional ownership0.01−0.05−0.09−0.08−0.18−0.010.010.040.08−0.03−0.05−0.10−0.190.02−0.071.00
17Board size0.010.350.180.030.630.09−0.16−0.08−0.34−0.090.200.110.31−0.010.02−0.161.00
18CEO equity ownership−0.02−0.14−0.050.00−0.25−0.030.070.020.11−0.05−0.10−0.02−0.13−0.11−0.07−0.07−0.231.00
19CEO option pay−0.010.060.050.030.02−0.010.180.050.110.050.05−0.070.020.13−0.090.040.01−0.081.00
20CEO tenure0.01−0.140.060.02−0.14−0.040.060.020.09−0.05−0.07−0.01−0.10−0.060.00−0.03−0.140.42−0.051.00
Mean0.040.310.860.018.030.221.650.09−0.400.170.610.300.3325.020.070.169.490.020.147.77
Standard deviation0.201.841.650.381.690.491.320.251.050.250.570.710.8027.800.070.092.290.040.197.34


Notes. N = 8,744. Correlations greater than 0.02 are associated with a p-value lower than 0.05.

Table

Table 2. Descriptive Statistics and Correlations for Variables Used in Hypotheses 3(a) and 3(b)

Table 2. Descriptive Statistics and Correlations for Variables Used in Hypotheses 3(a) and 3(b)

Variable12345678910111213141516171819202122
1Response time1.00
2Target acquiescence0.051.00
3Board demographic diversity0.050.061.00
4Firm size0.050.080.421.00
5Return on equity0.020.050.110.301.00
6Market-to-book ratio−0.040.000.12−0.110.021.00
7Revenue growth rate0.01−0.03−0.020.03−0.030.171.00
8Slack resources−0.040.04−0.16−0.33−0.120.24−0.021.00
9Corporate investment intensity−0.03−0.02−0.04−0.07−0.120.160.300.281.00
10Business diversification0.01−0.020.220.380.15−0.08−0.07−0.13−0.141.00
11Payout ratio0.020.010.100.100.110.06−0.05−0.05−0.040.071.00
12CSR0.020.010.330.550.230.070.04−0.140.060.260.111.00
13Foreign sales (%)−0.050.050.170.080.060.06−0.110.080.000.180.020.141.00
14Dedicated institutional ownership−0.020.020.050.180.040.140.05−0.06−0.030.030.050.140.061.00
15Transient institutional ownership−0.040.15−0.030.080.070.12−0.03−0.030.020.03−0.03−0.130.070.251.00
16Board size0.040.050.360.670.18−0.020.02−0.25−0.060.280.080.380.080.170.051.00
17CEO equity ownership0.120.01−0.09−0.22−0.05−0.03−0.030.06−0.04−0.08−0.07−0.10−0.13−0.040.01−0.191.00
18CEO option pay−0.030.130.130.070.020.130.010.020.020.09−0.02−0.050.10−0.10−0.040.05−0.071.00
19CEO tenure0.060.06−0.080.050.040.030.03−0.010.010.00−0.01−0.08−0.080.020.070.080.36−0.021.00
20Dedicated activist hedge funds−0.02−0.060.030.070.050.04−0.06−0.04−0.04−0.050.040.040.060.07−0.060.09−0.01−0.010.011.00
21Governance problems−0.040.000.250.380.140.010.01−0.18−0.080.300.060.210.090.120.160.290.020.040.22−0.031.00
22Performance problems0.06−0.040.020.02−0.04−0.120.05−0.24−0.04−0.030.000.060.090.08−0.050.070.02−0.090.030.040.071.00
Mean149.080.090.846.970.061.330.02−0.280.190.470.190.3522.820.070.128.740.010.146.070.35−0.110.02
Standard deviation198.330.291.942.150.631.030.311.690.380.550.830.8328.420.080.112.320.030.206.770.481.760.59


Notes. N = 431. Correlations greater than 0.09 are associated with a p-value lower than 0.05.

Table 3 reports the results used to test Hypotheses 1(a) and 1(b). In the first-stage regression, Model 1 shows that the coefficient estimate for percentage of foreign-born residents is positive (β = 2.26, p < 0.01) and that the coefficient estimate for social capital is positive (β = 0.15, p < 0.01), in line with our theory. To examine whether our instrumental variables can be considered exogenous and relevant, we instrument board demographic diversity on the percentage of foreign-born residents and social capital. Using “ivreg2” in Stata, we find evidence that the two instruments appear relevant and exogenous; the Cragg–Donald F statistic is 79, far greater than the critical value of 19 (Stock and Yogo 2005), and the result of a Hansen test fails to reject the null hypothesis of instrument exogeneity (p = 0.66). Models 2 and 3 are used to test Hypotheses 1(a) and 1(b), in which we control for the first-stage residuals.

Table

Table 3. The Moderating Effects of Board Demographic Diversity on Hedge Fund Activism Targeting

Table 3. The Moderating Effects of Board Demographic Diversity on Hedge Fund Activism Targeting

VariablesModel 1Model 2Model 3Model 4Model 5Model 6
First stageHedge fund activismDedicatedNondedicated
Board demographic diversity × Governance problems0.02***0.02**0.09***0.02
(0.01)(0.01)(0.03)(0.02)
Board demographic diversity × Performance problems0.16***0.15***0.35**0.15
(0.06)(0.06)(0.16)(0.11)
Board demographic diversity0.080.120.08−0.610.52
(0.15)(0.15)(0.15)(0.51)(0.39)
Governance problems−0.01−0.00−0.01−0.100.02
(0.02)(0.02)(0.02)(0.07)(0.06)
Performance problems0.22*0.25**0.26**0.410.81***
(0.13)(0.10)(0.10)(0.37)(0.25)
Firm size0.24***−0.06−0.06−0.060.07−0.20
(0.03)(0.05)(0.05)(0.05)(0.20)(0.13)
Return on equity0.01−0.10**−0.11**−0.10**−0.14−0.21*
(0.05)(0.04)(0.04)(0.04)(0.09)(0.12)
Market-to-book ratio0.00−0.21***−0.22***−0.22***−0.28**−0.70***
(0.03)(0.05)(0.04)(0.04)(0.13)(0.14)
Revenue growth rate−0.20**−0.15−0.18−0.19−0.83−0.27
(0.08)(0.18)(0.18)(0.18)(0.80)(0.65)
Slack resources−0.01−0.06−0.06−0.05−0.20−0.06
(0.04)(0.04)(0.04)(0.04)(0.14)(0.11)
Corporate investment intensity−0.26*0.22*0.23*0.22*0.470.51
(0.14)(0.13)(0.13)(0.13)(0.48)(0.32)
Business diversification0.050.090.090.090.120.19
(0.08)(0.06)(0.06)(0.06)(0.21)(0.17)
Payout ratio0.02−0.05−0.05−0.05−0.00−0.12
(0.03)(0.04)(0.04)(0.04)(0.21)(0.10)
CSR0.27***0.080.080.080.45*0.00
(0.05)(0.06)(0.06)(0.06)(0.23)(0.16)
Foreign sales0.01***−0.00−0.00−0.000.01−0.01
(0.00)(0.00)(0.00)(0.00)(0.01)(0.01)
Dedicated institutional ownership−0.172.00***2.01***2.01***7.04***1.61
(0.64)(0.59)(0.59)(0.60)(1.64)(1.66)
Transient institutional ownership0.71*0.560.520.563.41**0.10
(0.41)(0.43)(0.43)(0.43)(1.45)(1.11)
Board size0.11***−0.01−0.01−0.010.09−0.07
(0.02)(0.02)(0.02)(0.02)(0.08)(0.07)
CEO equity ownership−0.54−1.69*−1.78*−1.79*−5.64−3.72
(0.94)(0.93)(0.92)(0.92)(4.32)(2.72)
CEO option pay0.29*−0.17−0.21−0.200.17−0.78
(0.17)(0.19)(0.18)(0.19)(0.73)(0.55)
CEO tenure−0.01***0.01**0.01**0.01***0.03*0.03**
(0.00)(0.00)(0.01)(0.00)(0.02)(0.01)
Percentage of foreign-born residents2.26***
(0.43)
Social capital0.15***
(0.06)
Residuals−0.05−0.07−0.060.67−0.46
(0.15)(0.15)(0.15)(0.52)(0.39)
Constant−3.04***−1.64***−1.59***−1.65***−5.61***−1.89
(0.29)(0.48)(0.48)(0.48)(1.64)(1.34)
Observations8,7448,7448,7448,7448,7448,744
Industry fixed effectsYesYesYesYesYesYes
Year fixed effectsYesYesYesYesYesYes
Adjusted R20.324
χ2261.1303.1300.420,77220,772
Log likelihood−1,319−1,315−1,312−1,488−1,488
Pseudo-R20.1060.1090.1110.1280.128


Notes. Standard errors clustered by firm are reported in parentheses. Two-tailed tests.

 *p < 0.1; **p < 0.05; ***p < 0.01.

Results from Model 2 support Hypothesis 1(a). The coefficient for governance problems × board demographic diversity (β = 0.02, p < 0.01) is positively associated with the likelihood of hedge fund activism. To interpret this result, we follow Mize (2019) to calculate average marginal effect (AME) of governance problems when board demographic diversity takes different values. Presented in Figure 1, the AME of governance problems increases as board demographic diversity increases, and the AME is statistically significant when board demographic diversity exceeds three. The probability that a firm with high governance problems is targeted by an activist hedge fund more than doubles, increasing from 1.7% to 5.0%, when board demographic diversity shifts from a low value (mean − 1 SD) to a high value (mean + 1 SD). For comparative purposes, DesJardine et al. (2021a) find that the likelihood of hedge fund activism increases from 3.04% to 5.11% when CSR increases by two standard deviations. In this sense, the moderating effect of board demographic diversity is larger than the main effect of CSR on hedge fund activism and stronger than most other predictors of hedge fund activism reviewed by Denes et al. (2017), including takeover defenses, liquidity, institutional ownership, and sales growth.

Figure 1. (Color online) AME of Governance Problems Across Board Demographic Diversity
Note. The two dashed lines represent 95% confidence intervals.

Results from Model 3 support Hypotheses 1(b). The coefficient for performance problems × board demographic diversity is 0.16 (p < 0.01), suggesting that board demographic diversity amplifies the likelihood that performance problems lead to hedge fund activism. Shown in Figure 2, the AME of performance problems increases as board demographic diversity increases, and the AME is statistically significant when board demographic diversity is greater than two. In terms of magnitude, we find that the likelihood of targeting more than triples, from 1.3% to 5.1%, when board demographic diversity increases from a low to high value.

Figure 2. (Color online) AME of Performance Problems Across Board Demographic Diversity
Note. The two dashed lines represent 95% confidence intervals.

Model 4 includes both interaction terms. The coefficients for governance problems × board demographic diversity (β = 0.02, p < 0.05) and performance problems × board demographic diversity (β = 0.15, p < 0.01) are both economically and statistically similar to the prior results, suggesting that each type of problem has a distinct effect with board demographic diversity in driving hedge fund activism. To ensure the robustness of our findings, we also test Hypotheses 1(a) and 1(b) using ordinary least squares (OLS) regressions. Reported in Table A1 in the online appendix, the coefficients for both interaction terms remain positive and statistically significant.

Hypotheses 2(a) and 2(b) posit that dedicated activist hedge funds, instead of nondedicated activists, target firms with governance and performance problems based on their boards’ demographic diversity. Using multinomial logistic regression, Model 5 compares firm-years with no hedge fund activism (as the baseline) with those with activism undertaken by dedicated activist hedge funds. As reported, the coefficients for governance problems × board demographic diversity (β = 0.09, p < 0.01) and performance problems × board demographic diversity (β = 0.35, p < 0.05) are both statistically significant in the hypothesized directions. In Model 6, which compares firm-years with no hedge fund activism with those with activism undertaken by nondedicated activist hedge funds, neither of the interaction terms are statistically significant. Hypotheses 2(a) and 2(b) are supported.

Table 4 presents the results used to test Hypotheses 3(a) and 3(b), which examine firms’ responses to hedge fund activism. In support of Hypothesis 3(a) and derived from a GLM with a Poisson-gamma distribution, the coefficient in Model 1 for board demographic diversity is 0.10 (p < 0.05), suggesting that firms’ response time increases by 50 days when board demographic diversity increases from one standard deviation below to one standard deviation above its mean. To account for possible overdispersion in the variable response time, we also test Hypothesis 3(a) using a GLM with a negative binomial distribution. Reported in Model 2, the coefficient for board demographic diversity is 0.11 (p < 0.05), providing further support for Hypothesis 3(a).

Table

Table 4. Board Demographic Diversity and Target Response Time and Acquiescence

Table 4. Board Demographic Diversity and Target Response Time and Acquiescence

VariablesModel 1Model 2Model 3
Response timeTarget acquiescence
Board demographic diversity0.10**0.11**0.14**
(0.04)(0.05)(0.07)
Firm size−0.05−0.17**0.14
(0.05)(0.08)(0.09)
Return on equity−0.14−0.33*0.12
(0.14)(0.19)(0.20)
Market-to-book ratio0.030.07−0.06
(0.07)(0.09)(0.17)
Revenue growth rate0.040.18−0.11
(0.16)(0.41)(0.76)
Slack resources−0.01−0.08−0.11
(0.11)(0.21)(0.22)
Corporate investment intensity−0.57***−0.62**−1.13*
(0.20)(0.25)(0.61)
Business diversification−0.18−0.18−0.62***
(0.14)(0.17)(0.22)
Payout ratio0.070.20−0.04
(0.08)(0.15)(0.10)
CSR0.080.14−0.05
(0.08)(0.11)(0.17)
Foreign sales−0.00−0.000.00
(0.00)(0.00)(0.00)
Dedicated institutional ownership−0.99−2.06*−0.32
(1.02)(1.23)(2.23)
Transient institutional ownership0.030.482.06
(0.73)(0.96)(1.40)
Board size0.050.09*−0.05
(0.03)(0.05)(0.07)
CEO equity ownership3.72**−0.930.45
(1.48)(2.97)(4.10)
CEO option pay−0.260.121.83***
(0.27)(0.36)(0.58)
CEO tenure0.010.02*0.01
(0.01)(0.01)(0.02)
Dedicated activist hedge funds0.170.23*−0.08
(0.11)(0.13)(0.22)
Governance problems−0.050.03−0.10
(0.04)(0.06)(0.09)
Performance problems0.30*0.55***−0.07
(0.18)(0.21)(0.22)
Constant4.24***3.99***−2.75**
(0.74)(0.88)(1.19)
Observations431431357
Campaign demand fixed effectsYesYesYes
Industry fixed effectsYesYesYes
Year fixed effectsYesYesYes
Log likelihood−24,648−2,333−123.5
χ275.07
Pseudo-R20.186


Notes. Standard errors clustered by firm are reported in parentheses. The number of observations is smaller in Model 3 than in Models 1 and 2 because of the drop of singletons in Model 3. Two-tailed tests.

 *p < 0.1; **p < 0.05; ***p < 0.01.

In support of Hypothesis 3(b), results from a probit regression in Model 3 show that the coefficient for board demographic diversity is 0.14 (p < 0.05). When board demographic diversity increases from one standard deviation below to one standard deviation above its mean, the likelihood that a target firm will respond to hedge fund activism by taking proshareholder actions more than doubles, increasing from 9% to 19%.

Robustness Checks and Supplementary Analyses

Matched Sample Analyses.

To further assess the robustness of our findings, we replicate our main analyses using a matched sample, with the aim of reducing possible imbalances between firms that were targeted by activist hedge funds and those that were not targeted. Following related research (Cheng et al. 2015), we take two steps to identify the matched sample. First, we estimate a probit regression using the firm-year associated with hedge fund activism as the dependent variable, which equals one for firms targeted in the respective year and zero otherwise. For completeness, we include all control variables used to test Hypotheses 1(a) and 1(b) (listed in Table 1) as well as governance problems and performance problems as predictors of hedge fund activism. Using this model, we estimate the likelihood that each firm-year is associated with hedge fund activism. Second, we use one-to-one nearest-neighbor propensity score matching without replacement to identify the closest “look-alike” control firms based on the set of matching variables. Here, we match firms exactly by Fama–French 48 industry code and calendar year and find the most suitable pairs of targeted and nontargeted firms based on the matching variables. Using this approach, the matched sample comprises 337 firm-years associated with hedge fund activism and 337 firm-years without activism. Reported in Table A2 in the online appendix, postmatching t tests suggest that none of the matching variables statistically differ between the two groups at the level of p < 0.05.

Table 5 reports results from conditional logit regressions where each matched pair is a separate group. In Model 1, the coefficient for board demographic diversity × governance problems is 0.06 (p < 0.05), supporting Hypothesis 1(a). The results indicate that the marginal effect of governance problems on a firm’s probability of being targeted is −0.05 (p < 0.10) when board demographic diversity takes a low value and 0.06 (p < 0.01) when governance problems takes a high value. In Model 2, the coefficient for board demographic diversity × performance problems is 0.27 (p < 0.05), supporting Hypothesis 1(b). Here, the marginal effect of performance problems on a firm’s probability of being targeted is 0.24 (p < 0.05) when board demographic diversity takes a low value and 0.68 (p < 0.01) when board demographic diversity takes a high value. Model 3 reports similar results with both interaction terms included.

Table

Table 5. Conditional Logit Regressions Using a Matched Sample

Table 5. Conditional Logit Regressions Using a Matched Sample

VariablesModel 1Model 2Model 3
Board demographic diversity × Governance problems0.06**0.06**
(0.03)(0.03)
Board demographic diversity × Performance problems0.27**0.27**
(0.13)(0.13)
Board demographic diversity−0.04−0.03−0.07
(0.06)(0.06)(0.06)
Governance problems−0.010.03−0.02
(0.06)(0.06)(0.06)
Performance problems0.640.92***0.88**
(0.48)(0.35)(0.36)
Firm size−0.08−0.11−0.12
(0.11)(0.10)(0.10)
Return on equity−0.11−0.17−0.18
(0.18)(0.18)(0.18)
Market-to-book ratio−0.36*−0.43**−0.47***
(0.20)(0.17)(0.18)
Revenue growth rate−1.11**−1.15**−1.22**
(0.49)(0.50)(0.52)
Slack resources−0.27−0.24−0.25
(0.17)(0.17)(0.17)
Corporate investment intensity0.450.480.48
(0.41)(0.42)(0.44)
Business diversification0.140.190.19
(0.20)(0.19)(0.20)
Payout ratio0.060.050.03
(0.13)(0.12)(0.12)
CSR0.110.190.21
(0.18)(0.16)(0.17)
Foreign sales−0.01−0.01−0.01
(0.00)(0.00)(0.00)
Dedicated institutional ownership2.142.782.87
(1.88)(1.85)(1.87)
Transient institutional ownership3.22**3.06**3.28**
(1.37)(1.38)(1.39)
Board size0.020.030.02
(0.05)(0.05)(0.05)
CEO equity ownership−1.65−2.93−3.39
(2.58)(2.55)(2.63)
CEO option pay0.390.310.26
(0.48)(0.48)(0.48)
CEO tenure0.020.020.02*
(0.01)(0.01)(0.01)
Observations674674674
χ225.2124.1727.16
Log likelihood−220.2−220.3−218
Pseudo-R20.05740.05670.0669


Note. Standard errors clustered by firm are reported in parentheses.

 *p < 0.1; **p < 0.05; ***p < 0.01.

Accounting for CSR as an Alternative Explanation.

Because activist hedge funds target more socially responsible firms (DesJardine et al. 2021a), an alternative explanation for our findings is that targeting is driven by a firm’s CSR, which positively correlates with board demographic diversity, rather than diversity itself. In our main analyses, we addressed this concern by demonstrating that, relative to more homogenous boards, diverse boards respond more slowly to activism campaigns (Hypothesis 3(a)) and are more likely to acquiesce to activists (Hypothesis 3(b)), which support our theory that it is the board’s demographic diversity—not the firm’s social responsibility—that drives targeting. We further addressed this concern by conducting interviews to test our theory, empirically controlling for CSR, and using a control function approach. Further in line with our theory, DesJardine et al. (2021a) find that workplace diversity is the only dimension of CSR that does not directly increase a firm’s odds of being targeted, which suggests that diversity does not signal the waste of resources that is central to their theory.

Nevertheless, to further reconcile this concern, we retested our models after interchanging board demographic diversity with CSR as the moderator. If CSR creates an opportunity to influence firms, we should find that our main relationships hold using CSR as the moderator. Results for such analyses are reported in Model 1 of Table A3 in the online appendix and use the same measure of CSR as defined earlier. We find that the coefficient estimates for governance problems × CSR (β = 0.01, p = 0.491) and performance problems × CSR (β = −0.10, p = 0.275) are statistically not significant, suggesting that CSR does not fulfill the same role as board demographic diversity in shareholder activists’ opportunity structure.

Isolated Effects of Gender Diversity, Racial Diversity, and Nationality Diversity.

In our main analyses, we created an index of board demographic diversity based on diversity in gender, race, and nationality. To investigate the extent to which each type of diversity drives our findings, we interact governance problems and performance problems with gender diversity, racial diversity, and nationality diversity separately. Reported in Model 1 of Table A4 in the online appendix, we find that the coefficients for nationality diversity × governance problems (β = 0.34, p < 0.05), gender diversity × performance problems (β = 1.58, p < 0.05), and nationality diversity × performance problems (β = 2.14, p < 0.01) are positive and statistically significant. Conversely, we do not find statistically significant results when gender diversity and racial diversity are interacted with governance problems or when racial diversity is interacted with performance problems. From these results, we discern that racial diversity seems to have little effect on its own in driving hedge fund activism, which could arise if activist hedge funds perceive that directors from different racial groups will not be impeded when they are of similar gender and from similar countries, such as a Black Canadian female director working alongside a White Canadian female director.

Board Demographic Diversity and Director Turnover.

To provide further evidence that board demographic diversity can influence board unity, we investigate the relationship between board demographic diversity and director turnover. If board demographic diversity decreases board unity, we should anticipate a positive relationship between board demographic diversity and director turnover, meaning that a highly diverse board will experience a larger number of director departures. To test this idea, while controlling for board size, we measure director turnover as the number of directors departing a firm’s board in a year. Given that director turnover is a count, we conduct Poisson and negative binomial regressions. We include all control variables used in predicting hedge fund activism in addition to governance problems and performance problems. Model 1 of Table 6 uses Poisson regression and shows that the coefficient for board demographic diversity is 0.04 (p < 0.01). Model 2 uses negative binomial regression and shows that the coefficient for board demographic diversity is 0.04 (p < 0.01). Both findings imply that board diversity may impair board unity, as so far as it is reflected in director turnover.

Table

Table 6. Board Demographic Diversity and Director Turnover

Table 6. Board Demographic Diversity and Director Turnover

VariablesModel 1Model 2
Director turnover
Board demographic diversity0.04***0.04***
(0.01)(0.01)
Governance problems−0.03***−0.03***
(0.01)(0.01)
Performance problems0.070.06
(0.05)(0.05)
Firm size0.13***0.12***
(0.02)(0.02)
Return on equity−0.03−0.04
(0.03)(0.03)
Market-to-book ratio−0.04**−0.04**
(0.02)(0.01)
Revenue growth rate−0.23**−0.19**
(0.10)(0.09)
Slack resources0.030.02
(0.03)(0.03)
Corporate investment intensity−0.08−0.08
(0.07)(0.07)
Business diversification0.11***0.12***
(0.03)(0.03)
Payout ratio−0.04**−0.05***
(0.02)(0.02)
CSR0.05***0.07***
(0.02)(0.02)
Foreign sales−0.00−0.00
(0.00)(0.00)
Dedicated institutional ownership0.140.09
(0.35)(0.33)
Transient institutional ownership−0.28−0.21
(0.21)(0.21)
Board size0.04***0.05***
(0.01)(0.01)
CEO equity ownership−1.08**−1.03**
(0.51)(0.49)
CEO option pay0.060.09
(0.09)(0.09)
CEO tenure−0.02***−0.02***
(0.00)(0.00)
Constant−1.00***−0.97***
(0.16)(0.17)
Observations8,8188,818
Industry fixed effectsYesYes
Year fixed effectsYesYes
Log likelihood−13,533−12,992


Notes. Standard errors clustered by firm are reported in parentheses. Model 1 uses Poisson regression, and Model 2 uses negative binomial regression.

 **p < 0.05; ***p < 0.01.

Discussion

Our paper begins to reveal the corporate opportunity structure for shareholder activism by testing the idea that shareholder activists see board demographic diversity as an opportunity to influence firms. Situating our study in the context of hedge fund activism, we hypothesized and found that activist hedge funds target firms that exhibit governance and performance problems at higher rates when they also have more demographically diverse boards. We further explored the opportunity recognition of activist hedge funds by showing that dedicated activist hedge funds, rather than nondedicated activists, recognize and exploit this opportunity and that activist hedge funds correctly recognize this opportunity given that demographically diverse boards respond to hedge fund campaigns in ways that are likely to benefit intervening hedge funds. Our theory and findings contribute to research on shareholder governance and board governance, and they open up novel directions for future research.

Theoretical Contributions

Our first set of contributions is to research on shareholder governance (e.g., Goranova and Ryan 2014, Aguilera et al. 2015, Shi and Hoskisson 2021). Inspired by research on how social activists use corporate opportunity structures (Briscoe et al. 2014), we show that shareholder activists not only consider the problems firms exhibit in their targeting decisions but also, look to take advantage of opportunities. Specifically, we find that the size of the problems that a firm exhibits and the opportunities activists have to influence the firm have a multiplicative effect, in which problems determine the size of the potential payout for activist hedge funds, whereas opportunities determine the likelihood of the potential payout. Although we explored the opportunity that arises from board demographic diversity, we expect that many other influence opportunities for shareholders exist. Opportunities may arise, for instance, if CEOs have a deliberative approach to decision making (Frisch 2008) or if firms employ decentralized management (Bernstein et al. 2016); although such new governance models may work under “normal” circumstances, they could become a liability in high-strain confrontations with shareholder activists. The composition of a firm’s investor base may also create opportunities given that some investor bases may make it easier for shareholder activists to form alliances with other shareholders, and alliances are central considerations for activists (Wang et al. 2021). Organizational features such as these could influence the likelihood that a firm is targeted without constituting a “problem” that shareholder activists seek to address.

Uncovering the corporate opportunity structure for shareholder activism can advance research on shareholder governance by clarifying the antecedents of activism. Research on these antecedents highlights important inconsistencies in what drives shareholder activists to target firms (for a review, see Denes et al. 2017). Studies show, for example, that low dividend yields can increase (Brav et al. 2008) or have no effect (Boyson et al. 2017) and that high R&D expenditures can increase (Boyson and Mooradian 2011) or have no effect (Klein and Zur 2009) on the likelihood a firm is targeted by activist hedge funds. Viewed through our theoretical framework, whether activists see small dividends or high R&D expenditures as a problem worth targeting firms for could depend on the opportunities those activists see to successfully influence those firms. In short, examining problems alone, while ignoring opportunities, can proliferate mixed results on the antecedents of shareholder activism.

Analyzing the corporate opportunity structure for shareholder activism is also important because of the consequences that arise if shareholders target based on opportunities. Leading governance scholars and practitioners argue that shareholder activism is beneficial because corporate managers anticipate being targeted by activists and preemptively alter their behaviors to avoid being targeted (Fama and Jensen 1983). Yet, if shareholder activists target based on opportunities, the threat of targeting they brandish will not incentivize managers to preemptively resolve problems but to prepare for confrontations with activists in ways that may distract them from resolving these problems. More dangerous, this threat could cause managers to make value-destroying decisions that seek to remove the opportunities activists could exploit. Considering our findings, for instance, firms may avoid building demographically diverse boards because some activists see these as a targeting opportunity. Based on this insight, future research could analyze when the disciplinary effect of shareholder activism goes awry, with the aim to inform governance and policy outcomes that will safeguard against the potentially perverse outcomes of opportunity-based shareholder activism.

Our second set of contributions is to research on board governance (e.g., Post and Byron 2015, Creary et al. 2019, Jung et al. 2023). In recent years, shareholders have been rallying to increase board demographic diversity at unprecedented levels (Katz and McIntosh 2020). BlackRock, the world’s largest asset manager, has informed firms in its portfolio that it expects at least two female directors on every board (Fuhrmans 2018). Alongside the highly visible activities of such shareholders, our study suggests that some shareholder activists may be “welcoming” board demographic diversity for other reasons. In particular, our paper finds that activist hedge funds see board demographic diversity as an opportunity for shareholder influence, which is consequential because by targeting firms with demographically diverse boards, activist hedge funds could slow down or even derail the mainstreaming of board demographic diversity. A principal–principal conflict (Young et al. 2008) may thus exist between shareholders who want to foster board demographic diversity because it enhances the long-term performance of firms and shareholders who want to exploit demographically diverse boards.

Our paper has implications for the increasing number of researchers (e.g., Miller and Triana 2009) and practitioners (see Hirsch 2021) who try to prove a business case for board demographic diversity. Our findings suggest that conclusive evidence for a business case—even if it were to be found—would not be enough to align all shareholders behind board demographic diversity because some shareholder activists assess board demographic diversity not based on whether it creates long-term value or resolves focal organizational problems (e.g., gender imbalance) but based on whether it creates opportunities to intervene in corporate affairs.

Limitations and Future Research Directions

We highlight two limitations of our analysis that future research could address. First, although we were able to examine how board demographic diversity directly affects the response time of boards to activist campaigns, we could not do the same for board unity. Although our interviews suggest that activist hedge funds believe that demographically diverse boards will act less unitedly than more homogenous boards in confrontations with activists, further work is needed to examine this idea, which could be done by surveying boards to understand whether shareholder activist interventions are more likely to elicit conflicts between directors on more demographically diverse boards. Second, our analysis assumes that activist hedge funds are exclusively financially motivated. Although there is evidence that this is valid in most cases (Brav et al. 2008, Denes et al. 2017), cases exist in which these activists raise nonfinancial issues in their campaigns (DesJardine and Bansal 2021). Future research could measure the values of activist hedge funds by looking at how they vote on environmental and social shareholder proposals and explore whether these values influence the likelihood that they target firms with demographically diverse boards.

Practical Implications

Our research has important implications for nonactivist shareholders and policy makers. In almost all campaigns, shareholder activists depend on the support of other shareholders to influence firms. Although shareholder activists are unlikely to publicly acknowledge that they target firms based on their boards’ demographic diversity—and instead, as a portfolio manager we interviewed noted, “say something like ‘This board isn’t functioning as it should’”—our statistical analysis suggests that targeting based on diversity happens. Nonactivist shareholders who see the benefits of board demographic diversity may, therefore, want to think carefully about what is driving an activist’s interventions and which activist campaigns they choose to support. For policy makers, our paper highlights that the very features that make demographically diverse boards more effective under “normal” circumstances—that they make decisions more slowly and critically than homogenous boards—can become a liability in confrontations with shareholder activists. If policy makers wish to encourage greater board demographic diversity and help society benefit from the advantages that arise from such boards, they should consider prohibiting activist “lightning strikes” (Emmerich et al. 2021), as the Netherlands did in 2021 (Loyens & Loeff 2021). In sum, by better understanding the opportunities that some shareholder activists exploit, we can strengthen corporate governance systems that facilitate corporate progress on organizational outcomes that are important to both business and society.

Acknowledgments

The authors thank Brayden King for his expert editorial guidance and three anonymous reviewers for their constructive feedback in the review process. All authors contributed equally.

Endnotes

1 We interviewed six senior managers at activist hedge funds to understand their targeting decisions: (1) a cofounder and chief investment officer, (2) a founder and managing director, (3) a founder and managing partner, (4) a head of research, (5) a managing director, and (6) a portfolio manager. We also interviewed (7) a vice president at a firm that provides data to investors to understand what data activist hedge funds seek on target firms and (8) a senior corporate defense lawyer to understand how targeting unfolds at different types of firms. We recorded six of the interviews and took detailed notes for the remaining two interviews that our interviewees subsequently signed off. The interviews had an average length of approximately 30 minutes.

2 For example, there is no available data source to identify the social groups with common national or cultural traditions (i.e., ethnic groups) to which directors belong, and it is hard to form accurate measures based on available biographic or visual data on directors (without speaking to each director individually).

3 Hypothesis 3(a) investigates whether board demographic diversity can predict cross-sectional heterogeneity in terms of response time across firms that have been targeted by an activist hedge fund. A common method to test time use variables is survival analysis. However, given that the focal advantage of survival analysis is to cope with censoring in time use variables, which does not occur in our sample because all targeted firms have responded to hedge fund activism, survival analysis is not necessary or ideal to test our theory.

4 We use a firm-year level data set to implement the first-stage regression and calculate the inverse Mills ratio, whereas the analyses for Hypotheses 3(a) and 3(b) are at the activist campaign level. To ensure that our findings are not biased by the inclusion of the inverse Mills ratio calculated at the firm-year level, we generated bias-corrected standard errors using a bootstrap technique with 5,000 replications. Using this approach, the statistical significance in our findings did not change.

References

  • Acharya AG, Pollock TG (2021) Too many peas in a pod? How overlaps in directors’ local and global status characteristics influence board turnover in newly public firms. Acad. Management J. 64(5):1472–1496.CrossrefGoogle Scholar
  • Activist Insight (2018) Frequently Asked Questions & Help Guide (Activist Insight, London).Google Scholar
  • Aguilera RV, Desender K, Bednar MK, Lee JH (2015) Connecting the dots: Bringing external corporate governance into the corporate governance puzzle. Acad. Management Ann. 9(1):483–573.CrossrefGoogle Scholar
  • Ahn AM, Wiersema MF (2021) Activist hedge funds: Beware the new titans. Acad. Management Perspect. 35(1):96–122.CrossrefGoogle Scholar
  • Amason AC (1996) Distinguishing the effects of functional and dysfunctional conflict on strategic decision making: Resolving a paradox for top management teams. Acad. Management J. 39(1):123–148.CrossrefGoogle Scholar
  • Andrevski G, Richard OC, Shaw JD, Ferrier WJ (2014) Racial diversity and firm performance: The mediating role of competitive intensity. J. Management 40(3):820–844.CrossrefGoogle Scholar
  • Appel IR, Gormley TA, Keim DB (2019) Standing on the shoulders of giants: The effect of passive investors on activism. Rev. Financial Stud. 32(7):2720–2774.CrossrefGoogle Scholar
  • Baron RA (2006) Opportunity recognition as pattern recognition: How entrepreneurs “Connect the dots” to identify new business opportunities. Acad. Management Perspect. 20(1):104–119.CrossrefGoogle Scholar
  • Baron RA, Ensley MD (2006) Opportunity recognition as the detection of meaningful patterns: Evidence from comparisons of novice and experienced entrepreneurs. Management Sci. 52(9):1331–1344.LinkGoogle Scholar
  • Barrios JM, Bianchi PA, Isidro H, Nanda D (2022) Boards of a feather: Homophily in foreign director appointments around the world. J. Accounting Res. 60(4):1293–1335.CrossrefGoogle Scholar
  • Bartley T, Child C (2014) Shaming the corporation: The social production of targets and the anti-sweatshop movement. Amer. Sociol. Rev. 79(4):653–679.CrossrefGoogle Scholar
  • Bear S, Rahman N, Post C (2010) The impact of board diversity and gender composition on corporate social responsibility and firm reputation. J. Bus. Ethics 97(2):207–221.CrossrefGoogle Scholar
  • Bebchuk LA, Brav A, Jiang W, Keusch T (2020) Dancing with activists. J. Financial Econom. 137(1):1–41.CrossrefGoogle Scholar
  • Benton RA (2017) The decline of social entrenchment: Social network cohesion and board responsiveness to shareholder activism. Organ. Sci. 28(2):262–282.LinkGoogle Scholar
  • Bermeo N (1997) Myths of moderation: Confrontation and conflict during democratic transitions. Comparative Politics 29(3):305–322.CrossrefGoogle Scholar
  • Bernstein E, Bunch J, Canner N, Lee M (2016) Beyond the holacracy hype. Harvard Bus. Rev. 94(7):38–49.Google Scholar
  • Blau PM (1977) Inequality and Heterogeneity: A Primitive Theory of Social Structure (Free Press, New York).Google Scholar
  • Boyson NM, Mooradian RM (2011) Corporate governance and hedge fund activism. Rev. Derivatives Res. 14(2):169–204.CrossrefGoogle Scholar
  • Boyson NM, Pichler P (2019) Hostile resistance to hedge fund activism. Rev. Financial Stud. 32(2):771–817.CrossrefGoogle Scholar
  • Boyson NM, Gantchev N, Shivdasani A (2017) Activism mergers. J. Financial Econom. 126(1):54–73.CrossrefGoogle Scholar
  • Brav A, Jiang W, Kim H (2015) Recent advances in research on hedge fund activism: Value creation and identification. Annual Rev. Financial Econom. 7(1):579–595.CrossrefGoogle Scholar
  • Brav A, Jiang W, Partnoy F, Thomas R (2008) Hedge fund activism, corporate governance, and firm performance. J. Finance 63(4):1729–1775.CrossrefGoogle Scholar
  • Briscoe F, Chin MK, Hambrick DC (2014) CEO ideology as an element of the corporate opportunity structure for social activists. Acad. Management J. 57(6):1786–1809.CrossrefGoogle Scholar
  • Bromiley P (1991) Testing a causal model of corporate risk taking and performance. Acad. Management J. 34(1):37–59.CrossrefGoogle Scholar
  • Brown JE, Dunn PK (2011) Comparisons of tobit, linear, and poisson-gamma regression models: An application of time use data. Sociol. Methods Res. 40(3):511–535.CrossrefGoogle Scholar
  • Bushee BJ (1998) The influence of institutional investors on myopic R&D investment behavior. Accounting Rev. 73(3):305–333.Google Scholar
  • Bushee BJ (2001) Do institutional investors prefer near-term earnings over long-run value? Contemporary Accounting Res. 18(2):207–246.CrossrefGoogle Scholar
  • Canals J (2022) Boards of directors and corporate strategy: Some reflections on Pankaj Ghemawat’s contributions. Strategy Sci. 7(2):143–147.LinkGoogle Scholar
  • Castañón Moats M, DeNicola P, Malone L (2021) The Director’s Guide to Shareholder Activism (Harvard Law School Forum on Corporate Governance, Cambridge, MA).Google Scholar
  • Chatman JA, Flynn FJ (2001) The influence of demographic heterogeneity on the emergence and consequences of cooperative norms in work teams. Acad. Management J. 44(5):956–974.CrossrefGoogle Scholar
  • Chen W-R (2008) Determinants of firms’ backward- and forward-looking R&D search behavior. Organ. Sci. 19(4):609–622.LinkGoogle Scholar
  • Chen S, Feldman ER (2018) Activist-impelled divestitures and shareholder value. Strategic Management J. 39(10):2726–2744.CrossrefGoogle Scholar
  • Cheng CSA, Huang HH, Li Y (2015) Hedge fund intervention and accounting conservatism. Contemporary Accounting Res. 32(1):392–421.CrossrefGoogle Scholar
  • Clifford CP (2008) Value creation or destruction? Hedge funds as shareholder activists. J. Corporate Finance 14(4):323–336.CrossrefGoogle Scholar
  • Coles JL, Daniel ND, Naveen L (2014) Co-opted boards. Rev. Financial Stud. 27(6):1751–1796.CrossrefGoogle Scholar
  • Connelly BL, Shi W, Hoskisson RE, Koka BR (2019) Shareholder influence on joint venture exploration. J. Management 45(8):3178–3203.CrossrefGoogle Scholar
  • Creary S, McDonnell M-H, Ghai S, Scruggs J (2019) When and why diversity improves your board’s performance. Harvard Bus. Rev. (March 27), https://hbr.org/2019/03/when-and-why-diversity-improves-your-boards-performance.Google Scholar
  • Croson R, Gneezy U (2009) Gender differences in preferences. J. Econom. Literature 47(2):448–474.CrossrefGoogle Scholar
  • Dahlin KB, Weingart LR, Hinds PJ (2005) Team diversity and information use. Acad. Management J. 48(6):1107–1123.CrossrefGoogle Scholar
  • de Figueiredo RJP Jr, Feldman ER, Rawley E (2019) The costs of refocusing: Evidence from hedge fund closures during the financial crisis. Strategic Management J. 40(8):1268–1290.CrossrefGoogle Scholar
  • Denes MR, Karpoff JM, McWilliams VB (2017) Thirty years of shareholder activism: A survey of empirical research. J. Corporate Finance 44:405–424.CrossrefGoogle Scholar
  • Deng X, Kang J-k, Low BS (2013) Corporate social responsibility and stakeholder value maximization: Evidence from mergers. J. Financial Econom. 110(1):87–109.CrossrefGoogle Scholar
  • DesJardine MR, Bansal P (2021) Engine No. 1’s big win over Exxon shows activist hedge funds joining fight against climate change. The Conversation (May 26), https://theconversation.com/engine-no-1s-big-win-over-exxon-shows-activist-hedge-funds-joining-fight-against-climate-change-159983.Google Scholar
  • DesJardine MR, Durand R (2020) Disentangling the effects of hedge fund activism on firm financial and social performance. Strategic Management J. 41(6):1054–1082.CrossrefGoogle Scholar
  • DesJardine MR, Shi W (2022) The downside of displaying agentic values: Evidence from shareholder activism. Organ. Sci., ePub ahead of print November 2, https://doi.org/10.1287/orsc.2022.1637.LinkGoogle Scholar
  • DesJardine MR, Grewal J, Viswanathan K (2022a) A rising tide lifts all boats: The effects of common ownership on corporate social responsibility. Organ. Sci., ePub ahead of print October 21, https://doi.org/10.1287/orsc.2022.1620.LinkGoogle Scholar
  • DesJardine MR, Marti E, Durand R (2021a) Why activist hedge funds target socially responsible firms: The reaction costs of signaling corporate social responsibility. Acad. Management J. 64(3):851–872.CrossrefGoogle Scholar
  • DesJardine MR, Shi W, Sun Z (2021b) Different horizons: The effects of hedge fund activism vs. corporate shareholder activism on strategic actions. J. Management 48(7):1858–1887.Google Scholar
  • DesJardine MR, Zhang M, Shi W (2022b) How shareholders impact stakeholder interests: A review and map for future research. J. Management 49(1):400–429.Google Scholar
  • Dinesen PT, Schaeffer M, Sønderskov KM (2020) Ethnic diversity and social trust: A narrative and meta-analytical review. Annual Rev. Polital Sci. 23(1):441–465.CrossrefGoogle Scholar
  • Emmerich AO, Norwitz TS, Savitt W (2021) The SEC Should Address the Risk of Activist “Lightning Strikes” (Harvard Law School Forum on Corporate Governance, Cambridge, MA).Google Scholar
  • Ertimur Y, Ferri F, Muslu V (2011) Shareholder activism and CEO pay. Rev. Financial Stud. 24(2):535–592.CrossrefGoogle Scholar
  • Estélyi KS, Nisar TM (2016) Diverse boards: Why do firms get foreign nationals on their boards? J. Corporate Finance 39:174–192.CrossrefGoogle Scholar
  • EY (2021) What Boards Need to Know About Shareholder Activism (EY, London).Google Scholar
  • Fama EF, Jensen MC (1983) Separation of ownership and control. J. Law Econom. 26(2):301–325.CrossrefGoogle Scholar
  • Ferris SP, Jagannathan M, Pritchard AC (2003) Too busy to mind the business? Monitoring by directors with multiple board appointments. J. Finance 58(3):1087–1111.CrossrefGoogle Scholar
  • Fich EM, Shivdasani A (2006) Are busy boards effective monitors? J. Finance 61(2):689–724.CrossrefGoogle Scholar
  • Field L, Lowry M, Mkrtchyan A (2013) Are busy boards detrimental? J. Financial Econom. 109(1):63–82.CrossrefGoogle Scholar
  • Finkelstein S (1992) Power in top management teams: Dimensions, measurement, and validation. Acad. Management J. 35(3):505–538.CrossrefGoogle Scholar
  • Fiske ST, Neuberg SL (1990) A continuum of impression formation, from category-based to individuating processes: Influences of information and motivation on attention and interpretation. Zanna MP, ed. Advances in Experimental Social Psychology (Academic Press, Cambridge, MA), 1–74.Google Scholar
  • Fos V (2017) The disciplinary effects of proxy contests. Management Sci. 63(3):655–671.LinkGoogle Scholar
  • Freedman A, Fein M, Robertson I (2019) Fall of the Ivory Tower: Controlled Companies and Shareholder Activism (Harvard Law School Forum on Corporate Governance, Cambridge, MA).Google Scholar
  • Frisch B (2008) When teams can’t decide. Harvard Bus. Rev. 86(11):121–126.Google Scholar
  • Fuhrmans V (2018) How to get more women in the boardroom? Some try blunt force. Wall Street Journal (April 25), https://www.wsj.com/articles/how-to-get-more-women-in-the-boardroom-some-try-blunt-force-1524648602.Google Scholar
  • Gantchev N, Gredil OR, Jotikasthira C (2019) Governance under the gun: Spillover effects of hedge fund activism. Rev. Finance 23(6):1031–1068.CrossrefGoogle Scholar
  • Gibson CB (1997) Do you hear what I hear? A framework for reconciling intercultural communication difficulties arising from cognitive styles and cultural values. Erez M, Earley PC, eds. New Perspectives on International Industrial/Organizational Psychology (Jossey-Bass, San Francisco), 335–362.Google Scholar
  • Gibson CB, Zellmer-Bruhn ME (2001) Metaphors and meaning: An intercultural analysis of the concept of teamwork. Admin. Sci. Quart. 46(2):274–303.CrossrefGoogle Scholar
  • Glass C, Cook A (2020) Pathways to the glass cliff: A risk tax for women and minority leaders? Soc. Problems 67(4):637–653.CrossrefGoogle Scholar
  • Goodman PS, Ravlin E, Schminke M (1987) Understanding groups in organizations. Cummings LL, Staw BM, eds. Leadership, Participation, and Group Behavior (JAI Press, Greenwich, CT), 323–385.Google Scholar
  • Goodstein J, Gautam K, Boeker W (1994) The effects of board size and diversity on strategic change. Strategic Management J. 15(3):241–250.CrossrefGoogle Scholar
  • Goranova M, Ryan LV (2014) Shareholder activism: A multidisciplinary review. J. Management 40(5):1230–1268.CrossrefGoogle Scholar
  • Greenwood R, Schor M (2009) Investor activism and takeovers. J. Financial Econom. 92(3):362–375.CrossrefGoogle Scholar
  • Grégoire DA, Barr PS, Shepherd DA (2010) Cognitive processes of opportunity recognition: The role of structural alignment. Organ. Sci. 21(2):413–431.LinkGoogle Scholar
  • Gupta VK, Han S, Mortal SC, Silveri S, Turban DB (2018) Do women CEOs face greater threat of shareholder activism compared with male CEOs? A role congruity perspective. J. Appl. Psych. 103(2):228–236.CrossrefGoogle Scholar
  • Harrison DA, Klein KJ (2007) What’s the difference? Diversity constructs as separation, variety, or disparity in organizations. Acad. Management Rev. 32(4):1199–1228.CrossrefGoogle Scholar
  • Harrison DA, Sin H (2006) What is diversity and how should it be measured. Konrad AM, Prasad P, Pringle J, eds. Handbook of Workplace Diversity (Sage, Thousand Oaks, CA), 191–216.CrossrefGoogle Scholar
  • Harrison DA, Price KH, Bell MP (1998) Beyond relational demography: Time and the effects of surface- and deep-level diversity on work group cohesion. Acad. Management J. 41(1):96–107.CrossrefGoogle Scholar
  • Harrison DA, Price KH, Gavin JH, Florey AT (2002) Time, teams, and task performance: Changing effects of surface- and deep-level diversity on group functioning. Acad. Management J. 45(5):1029–1045.CrossrefGoogle Scholar
  • Heckman JJ (1979) Sample selection bias as a specification error. Econometrica 47(1):153–161.CrossrefGoogle Scholar
  • Heckman JJ, Robb R (1985) Alternative methods for evaluating the impact of interventions: An overview. J. Econometrics 30(1):239–267.CrossrefGoogle Scholar
  • Hewstone M, Hantzi A, Johnston L (1991) Social categorization and person memory: The pervasiveness of race as an organizing principle. Eur. J. Soc. Psych. 21(6):517–528.CrossrefGoogle Scholar
  • Hirsch L (2021) The business case for boardroom diversity. New York Times (January 23), https://www.nytimes.com/2021/01/23/business/dealbook/diversity-board-directors.html.Google Scholar
  • Hoetker G (2007) The use of logit and probit models in strategic management research: Critical issues. Strategic Management J. 28(4):331–343.CrossrefGoogle Scholar
  • Hofstede G (1984) Culture’s Consequences: International Differences in Work-Related Values (Sage, Thousand Oaks, CA).Google Scholar
  • Hogg MA, Terry DJ (2000) Social identity and self-categorization processes in organizational contexts. Acad. Management Rev. 25(1):121–140.CrossrefGoogle Scholar
  • Horwitz SK, Horwitz IB (2007) The effects of team diversity on team outcomes: A meta-analytic review of team demography. J. Management 33(6):987–1015.CrossrefGoogle Scholar
  • Hoskisson RE, Hitt MA, Johnson RA, Moesel DD (1993) Construct validity of an objective (entropy) categorical measure of diversification strategy. Strategic Management J. 14(3):215–235.CrossrefGoogle Scholar
  • Huang S, Hilary G (2018) Zombie board: Board tenure and firm performance. J. Accounting Res. 56(4):1285–1329.CrossrefGoogle Scholar
  • Insightia (2023) Shareholder Activism in 2022: The Definitive Statistical Analysis of Shareholder Activism with Data Compiled, Analyzed, and Published by Insightia (Insightia, New York).Google Scholar
  • Judge WQ, Miller A (1991) Antecedents and outcomes of decision speed in different environmental contexts. Acad. Management J. 34(2):449–463.CrossrefGoogle Scholar
  • Jung H, Lee YG, Park SH (2023) Just diverse among themselves: How does negative performance feedback affect boards’ expertise vs. ascriptive diversity? Organ. Sci. 34(2):657–679.LinkGoogle Scholar
  • Katz DA, McIntosh LA (2020) Corporate Governance Update: Raising the Stakes for Board Diversity (Harvard Law School Forum on Corporate Governance, Cambridge, MA).Google Scholar
  • King B (2008) A social movement perspective of stakeholder collective action and influence. Bus. Soc. 47(1):21–49.CrossrefGoogle Scholar
  • King BG, Soule SA (2007) Social movements as extra-institutional entrepreneurs: The effect of protests on stock price returns. Admin. Sci. Quart. 52(3):413–442.CrossrefGoogle Scholar
  • Klein A, Zur E (2009) Entrepreneurial shareholder activism: Hedge funds and other private investors. J. Finance 64(1):187–229.CrossrefGoogle Scholar
  • Klein A, Zur E (2011) The impact of hedge fund activism on the target firm’s existing bondholders. Rev. Financial Stud. 24(5):1735–1771.CrossrefGoogle Scholar
  • Krause R, Semadeni M, Cannella AA (2014) CEO duality: A review and research agenda. J. Management 40(1):256–286.CrossrefGoogle Scholar
  • Latham & Watkins (2015) Anticipating Activism: Implications for Your 2016 Annual Meeting of Stockholders (Latham & Watkins, Los Angeles).Google Scholar
  • Lenox MJ, Eesley CE (2009) Private environmental activism and the selection and response of firm targets. J. Econom. 18(1):45–73.Google Scholar
  • Lipton M (2019) Dealing with Activist Hedge Funds and Other Activist Investors (Harvard Law School Forum on Corporate Governance, Cambridge, MA).Google Scholar
  • Loyd DL, Wang CS, Phillips KW, Robert B, Lount J (2013) Social category diversity promotes premeeting elaboration: The role of relationship focus. Organ. Sci. 24(3):757–772.LinkGoogle Scholar
  • Loyens & Loeff (2021) Statutory Cooling-Off Period Will Enter into Force on May 1, 2021 (Loyens & Loeff, Amsterdam).Google Scholar
  • Marquis C, Battilana J (2009) Acting globally but thinking locally? The enduring influence of local communities on organizations. Res. Organ. Behav. 29:283–302.CrossrefGoogle Scholar
  • McAdam D (1996) Conceptual origins, current problems, future direction. McAdam D, McCarthy JD, Zald MN, eds. Comparative Perspectives on Social Movements: Political Opportunities, Mobilizing Structures, and Cultural Framings (Cambridge University Press, Cambridge, UK), 23–40.CrossrefGoogle Scholar
  • McDonnell M-H, King BG, Soule SA (2015) A dynamic process model of private politics: Activist targeting and corporate receptivity to social challenges. Amer. Sociol. Rev. 80(3):654–678.CrossrefGoogle Scholar
  • Messick DM, Mackie DM (1989) Intergroup relations. Annual Rev. Psych. 40(1):45–81.CrossrefGoogle Scholar
  • Miller T, Triana MDC (2009) Demographic diversity in the boardroom: Mediators of the board diversity–firm performance relationship. J. Management Stud. 46(5):755–786.CrossrefGoogle Scholar
  • Mize TD (2019) Best practices for estimating, interpreting, and presenting nonlinear interaction effects. Sociol. Sci. 6:81–117.CrossrefGoogle Scholar
  • Perrault E (2015) Why does board gender diversity matter and how do we get there? The role of shareholder activism in deinstitutionalizing old boys’ networks. J. Bus. Ethics 128(1):149–165.CrossrefGoogle Scholar
  • Petersen MA (2009) Estimating standard errors in finance panel data sets: Comparing approaches. Rev. Financial Stud. 22(1):435–480.CrossrefGoogle Scholar
  • Post C, Byron K (2015) Women on boards and firm financial performance: A meta-analysis. Acad. Management J. 58(5):1546–1571.CrossrefGoogle Scholar
  • Pugh SD, Dietz J, Brief AP, Wiley JW (2008) Looking inside and out: The impact of employee and community demographic composition on organizational diversity climate. J. Appl. Psych. 93(6):1422–1428.CrossrefGoogle Scholar
  • Rupasingha A, Goetz SJ, Freshwater D (2006) The production of social capital in US counties. J. Socio-Economics 35(1):83–101.CrossrefGoogle Scholar
  • Shane S, Venkataraman S (2000) The promise of entrepreneurship as a field of research. Acad. Management Rev. 25(1):217–226.CrossrefGoogle Scholar
  • Shi W, Hoskisson RE (2021) Understanding and Managing Strategic Governance (Wiley, Hoboken, NJ)Google Scholar
  • Shi W, Connelly BL, Hoskisson RE (2017) External corporate governance and financial fraud: Cognitive evaluation theory insights on agency theory prescriptions. Strategic Management J. 38(6):1268–1286.CrossrefGoogle Scholar
  • Shi W, Wajda D, Aguilera RV (2022) Interorganizational spillover: A review and a proposal for future research. J. Management 48(1):185–210.CrossrefGoogle Scholar
  • Shi W, Connelly BL, Hoskisson RE, David J, Ketchen J (2020) Portfolio spillover of institutional investor activism: An awareness–motivation–capability perspective. Acad. Management J. 63(6):1865–1892.CrossrefGoogle Scholar
  • Stangor C, Lynch L, Duan C, Glas B (1992) Categorization of individuals on the basis of multiple social features. J. Personality Soc. Psych. 62(2):207–218.CrossrefGoogle Scholar
  • Starboard Value (2018) Capitalize for Kids (Starboard Value, New York).Google Scholar
  • Stock JH, Yogo M (2005) Testing for weak instruments in linear IV regression. Andrews DWK, Stock JH, eds. Identification and Inference for Econometric Models: A Festschrift in Honor of Thomas Rothenberg (Cambridge University Press, Cambridge, UK), 80–108.CrossrefGoogle Scholar
  • Tarrow S (2022) Power in Movement: Social Movements and Contentious Politics, 4th ed. (Cambridge University Press, Cambridge, UK).CrossrefGoogle Scholar
  • Triana MDC, Miller TL, Trzebiatowski TM (2014) The double-edged nature of board gender diversity: Diversity, firm performance, and the power of women directors as predictors of strategic change. Organ. Sci. 25(2):609–632.LinkGoogle Scholar
  • Tsui AS, Egan TD, O’Reilly CA (1992) Being different: Relational demography and organizational attachment. Admin. Sci. Quart. 37(4):549–579.CrossrefGoogle Scholar
  • Tyler BB, Caner T (2016) New product introductions below aspirations, slack and R&D alliances: A behavioral perspective. Strategic Management J. 37(5):896–910.CrossrefGoogle Scholar
  • Vandor P, Franke N (2016) See Paris and … found a business? The impact of cross-cultural experience on opportunity recognition capabilities. J. Bus. Venturing 31(4):388–407.CrossrefGoogle Scholar
  • Wang CS, Whitson JA, King BG, Ramirez RL (2021) Social movements, collective identity, and workplace allies: The labeling of gender equity policy changes. Organ. Sci., ePub ahead of print October 21, https://doi.org/10.1287/orsc.2021.1492.LinkGoogle Scholar
  • Watson WE, Kumar K, Michaelsen LK (1993) Cultural diversity’s impact on interaction process and performance: Comparing homogeneous and diverse task groups. Acad. Management J. 36(3):590–602.CrossrefGoogle Scholar
  • Westphal JD (1999) Collaboration in the boardroom: Behavioral and performance consequences of CEO-board social ties. Acad. Management J. 42(1):7–24.CrossrefGoogle Scholar
  • Wiersema M, Ahn A, Zhang Y (2020) Activist hedge fund success: The role of reputation. Strategic Management J. 41(13):2493–2517.CrossrefGoogle Scholar
  • Wiesenfeld BM, Wurthmann KA, Hambrick DC (2008) The stigmatization and devaluation of elites associated with corporate failures: A process model. Acad. Management Rev. 33(1):231–251.CrossrefGoogle Scholar
  • Woolcock M (2001) The place of social capital in understanding social and economic outcomes. Canadian J. Public Policy 2:11–17.Google Scholar
  • Wooldridge JM (2010) Econometric Analysis of Cross Section and Panel Data (MIT Press, Cambridge, MA).Google Scholar
  • Young MN, Peng MW, Ahlstrom D, Bruton GD, Jiang Y (2008) Corporate governance in emerging economies: A review of the principal–principal perspective. J. Management Stud. 45(1):196–220.CrossrefGoogle Scholar
  • Zellmer-Bruhn ME, Maloney MM, Bhappu AD, Salvador R (2008) When and how do differences matter? An exploration of perceived similarity in teams. Organ. Behav. Human Decision Processes 107(1):41–59.CrossrefGoogle Scholar

Mark R. DesJardine is an associate professor of strategy and management at the Tuck School of Business at Dartmouth College and a senior fellow of the Wharton ESG Analytics Laboratory. With a focus on investor and stakeholder relations, his research resides at the intersection of strategy, sustainability, and finance. He is a CFA Charterholder and received his PhD in business administration from the Ivey Business School.

Wei Shi is a professor of management and a Cesarano Faculty Scholar at Miami Herbert Business School, University of Miami. His current research interests lie at the intersection of shareholder governance, stakeholder governance, and strategic decisions. He received his PhD in business administration from Rice University and recently coauthored the book entitled Understand and Managing Strategic Governance.

Emilio Marti is an assistant professor at the Rotterdam School of Management, Erasmus University and a core faculty member of the Erasmus Initiative “Dynamics of Inclusive Prosperity.” His research focuses on how different types of shareholders influence corporate sustainability, with a particular interest in sustainable investing. He received his PhD in business administration from the University of Zurich.

INFORMS site uses cookies to store information on your computer. Some are essential to make our site work; Others help us improve the user experience. By using this site, you consent to the placement of these cookies. Please read our Privacy Statement to learn more.