Strategic leadership in organizational crises: A review and research agenda Long Range Planning

Of all actors involved in managing an organizational crisis, strategic leaders play a particularly central role. However, the growing scholarship on the impact of strategic leaders in crisis situations is characterized by a high degree of fragmentation, considerably hindering the generation of parsimonious theory and practically useful insights. To address this issue, we conduct a systematic multidisciplinary literature review that spans the research streams on strategic leadership and organizational crises. For each type of strategic leader — Chief Executive Officer (CEO), top management team, and board of directors — we identify the different applied theoretical lenses and highlight commonalities and differences between studies and their insights. We use our review to derive an integrative conceptual framework that guides future research. Our exploratory review unveils that, while each type of strategic leader plays a significant role in a crisis context, the perspectives taken and the resulting evidence vary: as for the CEO, research focuses on social evaluations — for instance, based on the CEO ’ s appearance — as well as agency-theoretic considerations — particularly, financial incentives. Regarding the top management team, research mostly adopts a managerial and organizational cognition lens, focusing on characteristics such as personality and human capital. Lastly, for the board of directors, agency-theoretic considerations again dominate the scholarly conversation, especially studies of board independence. Overall, we review and organize a rich but patchy research land- scape, and we derive ample opportunities for novel theoretical and empirical inquiries into strategic leaders and their role in managing organizational (Sohn and Lariscy, 2012) and when CEOs match their crisis communication style with evaluators ’ image of their personality (Kim et al., 2016). Regarding communication content, research suggests that an apologetic CEO crisis response positively affects both external (Ver ˇ ci ˇ c et al., 2019) and internal stakeholders (Gillespie et al., 2014). At first sight, these findings seem to contradict general crisis communications research, which cautions against apologies as a default strategy (Coombs and Holladay, 2008, p. 252). However, these studies all investigate internally-induced crises for which assuming responsibility, as implicated by an apology, seems indeed warranted (Pace et al., 2010).


Introduction
One of the most challenging episodes for a firm is an organizational crisis (Pearson and Clair, 1998). An organizational crisis is an event perceived by organization members and stakeholders to be highly salient and unexpected, and a potentially disruptive threat to the focal organization and its stakeholders (Bundy et al., 2017;. Organizational crises are especially challenging not only because of their acute and public nature but also because they are typically difficult to resolve. Thus, although they are construed as events-i.e., episodes with a beginning and an end-they can turn into a chronical state (Jaques, 2009), for instance, when underlying causes are particularly systemic such that the crisis unfolds as a series of events (Milinski et al., 2016;Seeger et al., 1998).
Crises can emanate from both within the organization and from external events (Grewal et al., 2007). The corporate fraud scandal at Wirecard that led to the firm's insolvency (Davies, 2020), the severe quality data falsification at Kobe Steel that affected the global supply chain (Soble and Boudette, 2017), and Boeing's struggles after two crashes of its 737 MAX airplanes (Sucher, 2019) are prominent examples of internally-induced organizational crises. The current COVID-19 pandemic is an acute external cause of widespread organizational crises. For example, Christopher Nassetta, the CEO of Hilton Worldwide, noted: "Hilton has been around 100 years-we've never closed a hotel that wasn't going to be demolished for rebuilding. The bulk of our hotels in the major cities are closing, as we speak" (Thomaselli, 2020).
Scholars agree that of all actors involved in an organizational crisis, strategic leaders such as Nassetta-i.e., the CEO and other top management team (TMT) members, as well as members of the board of directors (BOD) (Finkelstein et al., 2009)-play a particularly central role (König et al., 2020b). In fact, while strategic leaders' characteristics generally influence organizations (Hambrick and Mason, 1984), they are likely to do so especially in the context of an organizational crisis. First, crisis management includes central tasks of strategic leadership, e.g., signal detection (Pearson and Clair, 1998), resource allocation under high levels of uncertainty and time pressure (Bower and Gilbert, 2005), and communication with internal and external stakeholders (Coombs, 2007;Waller and Younger, 2017). Second, as suggested by threat-rigidity theory (Staw et al., 1981), firms concentrate authority at the organizational apex in crisis situations (Dutton, 1986;Elsbach, 1994), rendering strategic leaders especially influential. Third, the decisions of strategic leaders in crisis situations will be extraordinarily impacted by their personal dispositions (Brockner and James, 2008;Dutton, 1986;Milburn et al., 1983). This is because crises are characterized by extraordinary ambiguity and time pressure (Pearson and Clair, 1998) and thus, particularly require strategic leaders to take mental shortcuts in decision-making which, in turn, are highly affected by personal dispositions (Hambrick, 2007).
Given the crucial importance of strategic leaders, it is not surprising that their impact in crisis situations has garnered substantial scholarly attention. As for strategic leaders' personality, for example, scholars have highlighted the potentially double-edged nature of trait empathy on crisis management effectiveness (König et al., 2020b). Regarding strategic leaders' financial incentives, researchers have shown that well-established incentive alignment mechanisms, such as CEO stock options, can increase the probability of product-harm crises (Wowak et al., 2015). Scholars have also illuminated crisis communication activities and suggest-contrary to most prior research on impression management-that in crisis situations, CEOs may favor sincere disclosure over impression management tactics (Patelli and Pedrini, 2014).
However, perhaps linked to the increasing variety of empirical findings, research on the role of strategic leaders in crisis contexts is characterized by a high degree of fragmentation as it is spread across various disciplines. In fact, influential studies on this topic come from a wide array of business-related fields such as management (Andreou et al., 2017), finance (Ho et al., 2016), accounting (Beasley, 1996), marketing (Kashmiri and Brower, 2016), and business ethics (Seeger and Ulmer, 2001). In addition, scholarly interest in strategic leaders' role in organizational crises extends to non-business research, including-but not limited to-communication science (Turk et al., 2012) and psychology (Gomulya et al., 2017).
Surprisingly, so far, a comprehensive in-depth review of this highly relevant but fragmented literature does not exist. Recent reviews on organizational crises acknowledge the importance of strategic leaders in crisis situations (Bavik et al., 2021;Bundy et al., 2017;James et al., 2011); yet, these reviews address the topic only peripherally or adopt a narrow scope and focus only on specific facets. In addition, with only a few recent exceptions (e.g., König et al., 2020b), the two literature streams of strategic leadership and organizational crisis have developed in isolation, with different theoretical underpinnings and assumptions. The complexity of the theoretical landscape makes it hard to fully grasp the topic and negatively impacts the normative and predictive validity of extant knowledge. In sum, the current, disintegrated state of research hinders the generation of new, parsimonious theory and practically useful insights.
We address this fragmentation with a systematic, exploratory literature review that comprehensively spans the research streams on strategic leadership (Finkelstein et al., 2009) and organizational crises (Pearson et al., 2007). In addition, we use the insights gained in the process to derive an integrative conceptual framework that combines elements of the upper echelons literature and research on organizational crises and guides future research.

Method and baseline model
In our search for relevant literature, we cast a particularly wide net, especially in view of the variety of scholarly disciplines dealing with strategic leaders in crisis contexts. To identify relevant journals, we proceeded in three steps. First, we included all peer-reviewed journals from the FT 50 and UT Dallas rankings. Second, we added all journals included in the scopes of recent major reviews of research on crisis (Bundy et al., 2017) and strategic leadership (Wowak et al., 2017). Third, we included key journals from relevant disciplines beyond management, such as psychology and sociology (following Ireland and Webb, 2007), as well as communication and crisis research (prompted by expert interviews with the authors of Bundy et al., 2017). Finally, we added Long Range Planning to the sample. This entire procedure yielded 72 journals (see Appendix A).
We used different combinations of synonyms for crises and for strategic leaders as keywords. To systematically identify relevant synonyms, we followed David and Han (2004) and scanned the text of core literature-the first chapter of Finkelstein et al. (2009) for strategic leaders and James et al. (2011) for crises-for synonyms. In addition, we complemented the results with synonyms from a thesaurus and from expert interviews with other researchers specializing in strategic management and organizational crises. The complete list of keywords used can be found in Appendix B. We focused our search on articles published from 1984-the publication year of Hambrick and Mason's seminal work-through 2019. We used Web of Science to perform our initial search. This search yielded 1212 articles.
In addition to this primary search, we aimed to identify relevant, highly cited articles published in journals other than the 72 core journals mentioned above. To do so, first, we defined 'highly cited' as the 10th percentile of the average yearly citations from the articles in our initial sample (i.e., 8.1 average annual citations). Second, to focus our search on relevant disciplines only, we determined which Web of Science subject categories to include. In this quest, the first two authors independently went through all subject categories and coded them as 'include' or 'exclude' candidates, respectively. This initial coding procedure resulted in substantial agreement (Cohen's Kappa = 0.75); the remaining cases of disagreement were resolved through discussion (see Appendix C for final list of subject categories). Again, we used Web of Science to perform this additional search: we applied the same keywords and inclusion criteria as in our initial, primary search, removed the restriction on selected journals, and set the filter to include relevant subject categories only. This yielded 377 additional articles, leading to a total sample of 1589 articles. We read the abstracts of all articles and applied our crisis definition-whose main components relate to the definition of Bundy et al. (2017) which, in turn, is based on insights from an analysis of multiple crisis definitions of the past 20 years-to each article to determine whether it is in scope. After completing this screening and additionally including articles found via backward and forward research, we retained 74 articles.
We systematically coded these 74 articles, including methodological aspects and content-related aspects, e.g., theoretical lenses. To identify relevant lenses, we followed an iterative inductive approach. First, to assign articles to the respective lenses, the first author examined the theory part of each article. She assigned an article to a lens if the theory section included an explicit mention of a lens (e. g., MOC) and/or referred to a lens's seminal articles (e.g., Walsh, 1995) or if it referred to a sub-category of a lens (e.g., upper echelons theory). In case more than one lens was evident, she chose the most relevant one. If an article did not explicitly mention any specific lens or sub-category, the authors jointly drew upon their collective expertise to code the article. The few cases of disagreement between the authors were resolved through discussion. Fig. 1 shows the number of articles included in our review by year of publication. Clearly, scholarly interest in strategic leaders' role in organizational crises has increased substantially over time: more than three quarters of articles in our sample have been published since 2008. Notably, this development seems driven by articles that use major organizational fraud scandals in the 2000s or the global financial crisis of 2007-2008 as their empirical setting. Fig. 2 depicts our baseline model which we used as a structure and roadmap for our review. The model is based on the general underlying notion in the upper echelons literature that strategic leaders' characteristics affect organizational outcomes (Hambrick and Mason, 1984). The model comprises different elements, which are anchored in the strategic management and crisis literatures as follows. At the center, first, there are the different types of strategic leaders and their characteristics. Second-as a link between leaders' characteristics and organizational-crisis-related outcomes-the model includes crisis management activities, in line with the work of leading crisis scholars who identify crisis management as a key aspect of crisis studies (e.g., James et al., 2011;Shrivastava, 1993). Third, the right-hand side of the model features organizational-crisis-related outcomes, as frequently discussed in the crisis literature (Bundy et al., 2017;Pearson and Clair, 1998). Fourth, at the top, the model allows for contingency factors, as the literature highlights their influence on strategic decision-making (Shepherd and Rudd, 2014), an activity directly and particularly linked to addressing organizational crises (Dutton, 1986). Finally, as indicated by the surrounding frame, the model reflects our expectation that prior research has adopted different theoretical lenses, each likely to contribute unique insights.

Strategic leadership in organizational crises as a multi-faceted research field
Our review is organized along two dimensions: first, focal actors; and second, theoretical lenses. This distinction underlines the integrative nature of our review as it helps us to trace back which actor and/or theory accounts for which building blocks in our final conceptual framework. Also, this distinction helps us to reveal promising future avenues in a structured and transparent manner. As focal actors, we include CEO, TMT, and BOD. All of them exert a substantial influence on strategic decisions and are therefore critical in crisis situations (Dutton, 1986;Finkelstein et al., 2009). Evidently, the CEO is the primary strategic decision maker (Busenbark et al., 2016). Yet, relying on the CEO alone seems barely sufficient, particularly because TMT members also contribute, often greatly, to strategic decision-making with their diverse experiences and abilities (Nielsen and Nielsen, 2011). In fact, TMT characteristics might be especially important for coping with a context as high in complexity as an organizational crisis (Edmondson et al., 2003). In including the BOD, we follow prior research which suggests that the role of the BOD becomes particularly pronounced in a crisis context (Chatterjee and Harrison, 2001;Lorsch and MacIver, 1989). Notably, we explicitly distinguish between different types of strategic leaders instead of considering them as one 'composite' actor. After all, each of these strategic leadership roles comes with specific entitlements and duties which makes it hard to aggregate the respective insights in a meaningful manner. In this vein, we proceed similarly to several previous reviews on strategic leaders that either consider the types of actors separately or focus on one type of strategic leader from the beginning (see, for example, Bromiley and Rau, 2015;Busenbark et al., 2016;Carpenter et al., 2004;Kurzhals et al., 2020;Menz, 2012). Table 1 shows the primary theoretical lenses we identified as being referred to in research on strategic leadership in organizational crises: agency theory (Jensen and Meckling, 1976), managerial and organizational cognition (MOC) (Walsh, 1995), social evaluations (Bundy and Pfarrer, 2015), executive ethical leadership (Treviño et al., 2003), and resource dependency theory (Pfeffer and Salancik, 1978).
The agency theory lens examines issues resulting from agency relationships, i.e., relationships where one actor-the principal, e.g.,

Fig. 2.
Baseline model for strategic leadership in organizational crises.

Table 1
Main theoretical lenses on strategic leadership in organizational crises. Jensen and Meckling (1976); Eisenhardt (1989) Strategic leaders' pursuit of own interests-which can go as far as explicit managerial misconduct-can trigger or exacerbate organizational crises. This tendency might be curbed by effective incentive schemes and monitoring measures.

Managerial and organizational cognition
Strategic leaders act under the regime of bounded rationality. Hence, their personal dispositions, such as their knowledge structures, influence and bias their information-processing and decision-making processes.
Walsh (1995) Handling crisis situations is an exceedingly complex task. To simplify this task, strategic leaders often unconsciously rely on their personal dispositions when it comes to addressing crises.

Social evaluations
Different stakeholders', i.e., evaluators', perceptions and sensemaking of an organization and its members' behavior is based on both analytical and affective processes. Important organizational outcomes of these processes include approval, legitimacy, and reputation. Bundy and Pfarrer (2015) Strategic leaders are often seen as the "public face" ( Gomulya et al., 2017, p. 617) of a firm and are thus likely to influence evaluators' perceptions-especially in a crisis context, since it is often characterized by high ambiguity, which increases evaluators' reliance on easily accessible cues.

Executive ethical leadership
Strategic leaders are pivotal for the ethical conduct of the organization by setting an appropriate "tone at the top" which trickles down to all organizational ranks (Soltani 2014, p. 255). Strategic leaders need to be both a "moral person"-i.e., possess normatively appropriate characteristics-and a "moral manager"-i.e., actively promote ethical conduct in the organization (Brown and Treviño, 2006 p. 597) Treviño et al. (2003) Many organizational crises, particularly internallyinduced ones, can be traced back to unethical conduct prevailing throughout the organization. Strategic leaders' "tone at the top" (Soltani 2014, p. 255) is decisive when it comes to addressing organizational crises or avoiding such crises in the first place.

Resource dependency theory
The organization is seen as an open system that is influenced by external factors. Strategic leaders can decrease environmental uncertainty and dependence, for instance by providing important resources such as expertise or access to information to the focal organization. Pfeffer and Salancik (1978); Hillman et al., 2009 Crisis contexts are often unique and require particular resources to be addressed adequately. Strategic leaders' expertise and connections to other organizations can constitute one decisive resource in such particularly challenging situations.
company owners-delegates a task to another actor who performs that task-the agent, e.g., the CEO. Of particular interest in this domain is solving the agency problem that arises when the goals of agent and principal are in conflict and it is difficult for the principal to monitor the agent's behavior (Eisenhardt, 1989;Jensen and Meckling, 1976). This perspective is especially relevant in a crisis context, as strategic leaders' pursuit of own interests can entail managerial misconduct and hence trigger or exacerbate organizational crises (Kulik, 2005). At the same time, strategic leaders' tendency to pursue their own objectives rather than that of the principal might be influenced by financial incentives and the behavior of other focal actors, particularly the BOD (Eisenhardt, 1989). The MOC lens emphasizes the role of strategic leaders' individual cognition and perceptions in recognizing and resolving strategic issues-crucial processes in the context of organizational crises (Dutton, 1986). This perspective emphasizes that personal dispositions shape how decision makers simplify complex tasks as well the potential biases in this process (Hodgkinson and Healey, 2008;Walsh, 1995). It incorporates the notions of bounded rationality (Simon, 2013) and cognitive structures (Thomas and Porac, 2002) and it is an important lens for strategic leaders' role in crises because it conceptualizes managerial sensemaking along all crisis stages (Maitlis and Sonenshein, 2010;Weick, 1995). The MOC lens traditionally centers on the TMT (Thomas and Porac, 2002) and inherently explores less visible leader characteristics, e.g., personality, which are often proxied by demographic variables (Hambrick, 2007).
The social evaluations lens focuses on how different stakeholders perceive and evaluate an organization and its members. In many cases, the person of interest is the CEO, who is often perceived as the "public 'face' of the company" (Men, 2012, p. 171, p. 171)-and whose visibility, e.g., as spokesperson, becomes particularly important in a crisis situation (Laufer et al., 2018). The resulting social evaluations are reflected in organizational outcomes such as approval, legitimacy, and reputation (Bundy and Pfarrer, 2015;Pollock et al., 2019). Hence, the social evaluations lens contributes to a richer view on both strategic leaders' characteristics-above all, appearance-and organizational outcomes, especially non-financial ones.
The executive ethical leadership lens is based on the seminal work by Treviño et al. (2003). It emerged in the wake of major corporate scandals in the early 2000s and is thus relatively novel when compared to the other lenses. This lens enriches this review as it emphasizes the "tone at the top" (Soltani, 2014, p. 255), i.e., the importance of strategic leaders and their values for the ethical conduct of the whole organization (Treviño et al., 2003).
Finally, the resource dependency lens on strategic leaders acknowledges the ability of top executives, especially directors, to provide important resources, such as expertise and access to resources to their organizations (Hillman et al., 2009;Pfeffer and Salancik, 1978). Having such "resource-rich" individuals who possess particular amounts of human and social capital as strategic leaders is crucial in situations characterized by a high degree of strategic uncertainty, such as organizational crises (Boyd, 1990, p. 428). This lens is often used complementary to agency theory, especially when it comes to the role of the BOD (Hillman and Dalziel, 2003). Table 2 summarizes the insights from the coding procedure and clusters the identified articles along the two dimensions of theoretical lenses and focal actor(s). The respective cells contain the number of studies for each category. These results highlight the fragmented nature of the literature and the broad variety of underlying theoretical approaches. Overall, the most frequent lens is agency theory, driven by studies on the CEO and/or the BOD, followed by the MOC and social evaluations lenses. Table 2 also reveals   Table 2 Theoretical lenses on strategic leaders in organizational crises by category of focal actor.

Table 3a
Literature on strategic leadership in organizational crises-focal actor CEO.

Reference
Main theoretical lens

Research design
Findings/propositions Reference to framework Harris and Bromiley (2007) Agency theory Quantitative The share of CEO pay based on stock options increases the probability of crises resulting from accounting restatements.
Financial incentives Wowak et al. (2015) Agency theory Quantitative The share of CEO pay based on stock options is positively related to the likelihood of product-harm crises. These effects are especially strong in the early periods of CEO tenure and among non-founder CEOs. Financial incentives; Firmlevel context Zhang et al. (2008) Agency theory Quantitative CEO out-of-the-money options induce earnings manipulation, particularly for long-tenured CEOs; in-the-money options do not exert a significant effect. Low firm performance is related to higher crisis likelihood when CEOs own less stock.

Financial incentives;
Firm-level context Liu et al. (2016) Agency theory Quantitative In case of a product recall, full remedy is more likely in case of more CEO equity incentives. Also, in case of a product recall, a longer-tenured CEO is less likely to opt for a full remedy of affected customers to manage the crisis than a CEO who has been holding this position for a shorter period of time.  In a public-relations crisis, CEOs with a babyface are perceived as more honest than CEOs with a mature face and, thus, attitudes towards companies led by baby-faced CEOs are more favorable; increasing crisis severity attenuates this effect.

Quantitative
Organizations that appoint a new CEO after an internally-induced crisis tend to opt for a successor whose facial appearance conveys integrity, which leads to more positive analyst forecasts and mitigates negative press coverage. gaps in the research landscape, such as an ethical leadership perspective on the BOD's role in organizational crises, indicating intriguing potential future research opportunities which will be discussed in the final section of this paper.

The role of strategic leaders in organizational crises
Below, we discuss the key insights of the articles in our sample, including direct effects of strategic leaders' characteristics on organizational-crisis-related outcomes as well as contingency factors (i.e., crisis attributes and context) that influence these effects. For each focal actor, we present the theoretical lenses, sorted by frequency of occurrence in the literature. Within each of the lenses, we discuss the respective identified strategic leader characteristics that are part of the conceptual framework that we derive as a central outcome of our review.

Effects of CEO characteristics
As shown in Table 3a, most articles in our sample explicitly focus on the CEO. This is interesting given that research on strategic leaders has long recognized the need to also focus on the TMT as a group (Carpenter et al., 2004;Hambrick and Mason, 1984). The large scholarly interest in the CEO in a crisis context seems to be related to the general perception of the CEO as the primary strategic decision maker in the firm (Busenbark et al., 2016;Quigley et al., 2017). Scholars primarily apply social evaluations, agency theory, and MOC as theoretical lenses. Some, but substantially fewer, studies also refer to executive ethical leadership. Notably, none of the studies on the CEO in our sample adopt a resource dependency lens.
Agency theory. Given the emphasis of agency theory on executive compensation (Eisenhardt, 1989), the reviewed studies mainly focus on CEOs' financial incentives as a determinant of organizational-crisis-related outcomes. Agency theory promotes the use of stock-based compensation to align the CEO's and shareholders' respective interests and to motivate CEOs to take greater risks (Hall and Liebman, 1998). The underlying assumption is that shareholders tend to be relatively risk-neutral as they typically hold a diversified portfolio; conversely, CEOs typically prefer less risky strategies-although those probably come with lower returns-because much of CEOs' financial and reputational fate is tied to the focal organization (Sanders and Hambrick, 2007;Wowak et al., 2015).
However, with only a few exceptions (e.g., Liu et al., 2016), most of the studies in our sample (Grove et al., 2011;Harris and Bromiley, 2007;Soltani, 2014;Wowak et al., 2015) illustrate that the reliance on stock-based compensation does not necessarily promote "smart", i.e., shareholder-wealth maximizing, risk taking Hambrick, 2007, p. 1073). For instance, research demonstrates that this type of CEO pay is linked to severe financial scandals (Harris and Bromiley, 2007;Soltani, 2014) and product safety issues (Wowak et al., 2015), both triggers for organizational crises.
To explain such anomalies to standard agency theory, studies refer to Wiseman and Gomez-Mejia's (1998) model of behavioral agency, which suggests that managers do not so much avoid risk as they avoid loss (Kahneman and Tversky, 1979). Consequently, the motivational effects of specific compensation vehicles seem to depend on strategic leaders' evaluation of the situation (i.e., gain versus loss context). In this vein, Zhang et al. (2008) differentiate between different types of stock options. They show, for instance, that out-of-the-money options (with current market price of stock below strike price) put CEOs in a situation of perceived losses and thereby induce aggressive behavior such as earnings manipulation.
Another factor investigated from an agency perspective-although not as extensively as financial incentives-is CEO power. Generally, in strategic management research, power refers to "the capacity of individual actors to exert their will" and can originate from different sources, for example top managers' expertise or formal position in the organization (Finkelstein, 1992, p. 506). Paradoxically, CEO power seems to be unfavorably linked to crisis occurrence (Agrawal and Chadha, 2005; Grove et al., 2011) but, at Given that a crisis is of global nature, CEOs tend to engage in sincere disclosure, since there is less incentive to sugarcoat information.
Crisis communication the same time, favorably (Mueller and Barker, 1997), or at least neither favorably nor unfavorably (Tang and Crossan, 2017), to outcomes related to crisis resolution, e.g., firm performance. A possible explanation might be that, on the one hand, power makes CEOs focus on the potential rewards of their actions rather than on the potential threats (Lewellyn and Muller-Kahle, 2012), making crises more likely. On the other hand, powerful CEOs can make and implement strategic decisions more quickly-a pivotal factor to effectively battle an acute crisis (Dutton, 1986;Pearson and Clair, 1998). The diverging findings of Mueller and Barker (1997) and Tang and Crossan (2017)-i.e., whether CEO power is linked to crisis resolution in the first place-might be explained by the different operationalizations of power, and by the fact that Tang and Crossan (2017), in contrast to Mueller and Barker (1997), focused on newly appointed CEOs, who might lack the knowledge and networks to execute their power to effectively solve the crisis at hand (Virany et al., 1992).
Researchers have also made first advances regarding the impact of CEOs' human capital. In this regard, research particularly highlights the beneficial effect of CEOs that belong to the owning family (Minichilli et al., 2016) or that have a relatively short tenure (Liu et al., 2016). The main underlying reasoning behind these studies is that such CEOs are more long-term focused, which positively impacts organizational-crisis-related outcomes.
Managerial and organizational cognition (MOC). In this domain, scholars have focused primarily on the sociocognitive implications of CEO personality for crisis management (König et al., 2020b). In particular, the literature we reviewed reflects the recent shift in strategic leadership studies towards a more nuanced portrayal of executive personality (Gerstner et al., 2013)-i.e., these studies show that any given personality trait can have both positive and negative effects when it comes to CEOs' impact on organizational crises.
In this regard, with only a few exceptions (König et al., 2020b), most studies focus on the roles of the related traits of CEO narcissism and overconfidence and highlight both positive (Patel and Cooper, 2014) and negative (Buyl et al., 2019;Ho et al., 2016;Kashmiri et al., 2017) effects on crisis-related outcomes. For instance, Patel and Cooper (2014) refer to studies in psychology (e.g., Foster and Trimm, 2008) which suggest that narcissistic CEOs tend to strive for desirable outcomes ('approach-motivation') and eschew undesirable outcomes ('avoidance-motivation'). These scholars find that, as a result, firms led by more narcissistic CEOs are more likely to take risks and invest aggressively, and ultimately, to resurge after a crisis. Buyl et al. (2019) also refer to avoidance-approach motivation theory but draw different conclusions. Specifically, they conclude that narcissistic CEOs deplete their companies' pre-crisis resources and, in so doing, undermine their capabilities to recover from the crisis. Interestingly, Buyl et al. (2019) mainly attribute the conflict between their own and Patel and Cooper's (2014) findings to a difference in empirical settings. While the latter examine manufacturing companies, Buyl et al. (2019) analyze banking companies which suffered from a tarnished image after the global financial crisis of 2007-2008, hence reducing their CEOs' discretion to take risks and resulting in problems to raise fresh capital to finance their post-crisis recovery.
Other studies that adopt a MOC lens focus on elements not directly tied to personality, such as human capital, values, and cognition. Remarkably, all these studies use proxies-either CEOs' demographic data or communication-to illuminate the proposed mechanisms. Regarding CEOs' human capital and values, respectively, research indicates a beneficial effect for organizations led by exmilitary (Koch-Bayram and Wernicke, 2018) or female CEOs (Palvia et al., 2015); supposedly, because the former are more likely to follow rules diligently and because the latter display higher conservatism-both success factors for navigating an acute crisis or avoiding a crisis from occurring in the first place. Other research goes one step further and investigates the fit of the CEO's human capital to the specific crisis situation, showing that fit is pivotal to successfully emerging from a crisis (Chen and Hambrick, 2012).
As for CEO cognition, research illustrates potentially detrimental effects. For instance, CEOs' failure to assume responsibility for the crisis-as indicated by their sensemaking efforts towards the public-seems to inhibit their individual learning and eventually also organizational learning, as CEOs' sensemaking trickles down and ultimately permeates the entire organization (Tourish and Hargie, 2012).
Social evaluations. In this domain, the extant literature focuses on the role of readily observable cues, such as CEOs' facial appearance or crisis communication. This is in line with the general nature of social evaluations as perceptions of external evaluators (Bundy et al., 2017). As for CEO appearance, studies find that CEOs with faces that convey traits like integrity and honesty have a beneficial effect on stakeholder perceptions in crisis situations (Gomulya et al., 2017;Gorn et al., 2008). For instance, Gorn et al. (2008) find that, during a public-relations crisis, attitudes towards companies led by baby-faced CEOs are more favorable since such CEOs are perceived as more honest than CEOs with a mature face. This is interesting in light of previous research which suggests that in non-crisis times, having a mature face might be more desirable for a strategic leader, particularly because people perceive others with mature-looking faces as more competent (Zebrowitz and Montepare, 2005) and perceptions of CEO maturity are associated with higher company profits (Rule and Ambady, 2008).
Scholars have also investigated the ramifications of CEOs' crisis communication. In this regard, with only one notable exception (Einwiller et al., 2017), research focuses on CEO presence and communication style. Generally, the CEO's presence in crisis communication, e.g., as a spokesperson, makes consumers more willing to buy the company's products and enhances their attitudes towards the company (Laufer et al., 2018;Turk et al., 2012). Yet, other studies suggest that a CEO's presence alone is barely enough to explain the positive effect on social evaluations. Indeed, they demonstrate that constituents make positive social evaluations when CEOs have gained a favorable reputation before the crisis (Sohn and Lariscy, 2012) and when CEOs match their crisis communication style with evaluators' image of their personality (Kim et al., 2016). Regarding communication content, research suggests that an apologetic CEO crisis response positively affects both external (Verčič et al., 2019) and internal stakeholders (Gillespie et al., 2014). At first sight, these findings seem to contradict general crisis communications research, which cautions against apologies as a default strategy (Coombs and Holladay, 2008, p. 252). However, these studies all investigate internally-induced crises for which assuming responsibility, as implicated by an apology, seems indeed warranted (Pace et al., 2010).
Another, rather rarely explored, CEO-level variable is social capital, which comprises the valuable social relationships of strategic leaders (Adler and Kwon, 2002). The extant literature finds that CEOs' social capital is often beneficial during crises and can assume various forms, such as credibility among creditors (Andreou et al., 2017) or affiliations with organizations that have a positive image (Lungeanu et al., 2018). However, social capital is not universally beneficial. O'Rourke (2004), for instance, investigates the case of the media and merchandising firm Martha Stewart and her CEO of the same name who triggered a crisis through misconduct, highlighting the CEO's celebrity status as detrimental social capital because it arguably amplified the impact of the misconduct on the company's stock performance and reputation.
Executive ethical leadership. Similar to research that applies an agency theory lens, studies that adopt an executive ethical leadership perspective mainly focus on CEOs' role in triggering a crisis. Generally, these studies find that internally-induced organizational crises can be caused by a failure to establish an appropriate "tone at the top" (Soltani, 2014, p. 255) which emphasizes aspects such as an ethical corporate culture, effective compliance, and a code of good conduct. Specifically, this literature highlights certain CEO personality traits as conducive to the emergence of such crises. Such results add to insights based on the MOC lens which links narcissism and related personality traits mostly-but not always-to undesirable organizational outcomes. Specifically, the extant executive ethical leadership literature pays special attention to the detrimental impact of CEOs' hubris and narcissism, which can both result in severe organizational crises (Chen, 2010;Fassin and Gosselin, 2011).
Likewise, the literature emphasizes that CEOs' moral values-or rather, a lack thereof-can contribute to crisis occurrence (Zona et al., 2013). Conversely, adequate CEO values can mitigate crises, as Seeger and Ulmer (2001) illustrate with case studies of catastrophic fires at two manufacturing companies. More specifically, these scholars show that a set of CEOs' virtuous values-for instance, appreciation of employees and focus on opportunities arising from the crisis-can help in battling an acute crisis. These scholars highlight that if these values are deeply ingrained in CEOs, this allows them to signal authenticity and commitment, which contributes to the relief of employees and to the restoration of business operations.
In addition, executive ethical leadership research has examined CEO power and arrives at conclusions similar to that of agency theory. Soltani (2014), for instance, analyzes high-profile corporate scandals and demonstrates that powerful CEOs-characterized by an extraordinarily influential position-may see this power as both a means and an end. This scholar argues that, as power is such a central concept for these CEOs, they desire to further increase their power. This desire, in turn, can make them become entangled in a self-reinforcing vicious circle, ultimately leading to corporate misconduct, a common trigger of crises.
There is only little research from an ethical leadership perspective regarding another factor, crisis communication. Notably, Patelli and Pedrini (2014) examine CEOs' letters to shareholders regarding sincerity, a major principle of discourse ethics (Yuthas et al., 2002). These scholars find that in the wake of the global financial crisis of 2007-2008, CEOs tended to engage in sincere disclosure-conceptualized as the congruency between the level of optimism displayed in CEOs' letters to shareholders and company performance. In order to explain these results, Patelli and Pedrini (2014) argue that in case of a global crisis, external stakeholders expect negative results and demand transparent communication about it, both reducing CEOs' incentives to sugarcoat information. Hence, these findings provide first evidence that during global crises, impression management tactics might be less common.

Contingency factors regarding the effect of CEO characteristics
To date, scholars have uncovered a broad range of contingency factors regarding the CEO's role in organizational crises. These include both crisis attributes and aspects of the firm-level and environment-level context. Notwithstanding its breadth, the depth of the extant research regarding these factors is limited and most factors are only examined by one study each.
Crisis attributes. One important crisis attribute is (perceived) crisis origin. Related research findings imply that if a crisis is seen as preventable, the CEO is well-advised to communicate in a manner that displays a concern for others and use apologies (Maiorescu, 2016), which further corroborates research on the direct effects of apologies in the context of internally-induced crises (Gillespie et al., 2014;Verčič et al., 2019).
Crisis severity is another condition the literature discusses. Most notably, Gorn et al. (2008) argue and empirically demonstrate that if external stakeholders are aware of situational cues-such as crisis severity-that have greater diagnostic value than strategic leaders' characteristics, they tend to use these cues to correct their possibly inaccurate inferences regarding the crisis and any involved strategic leaders. Overall, this observation illustrates the need to put the direct effects of easily readable cues (e.g., Gomulya et al., 2017) into perspective.
In addition, Sohn and Lariscy (2012) underline the importance of congruency between crisis type and CEO characteristics. These scholars argue that crises are highly ambiguous situations, prompting external evaluators to use easily observable cues (Pearson and Clair, 1998) when making inferences about unobservable organizational abilities to resolve a crisis (Roest and Rindfleisch, 2010). Specifically, Sohn and Lariscy (2012) suggest that, in a crisis setting, a CEO's reputation exerts a halo effect such that a company benefits greatly when it actively manages the public image of its CEO-at least when the CEO's reputation (e.g., known for managerial excellence) matches the type of crisis (e.g., crisis due to lack of corporate abilities).
Context. Prior studies have found both firm-level and environment-level context to be influential factors, although research is rather sparse, especially on the latter. With regard to firm-level context, the majority of prior research has focused on internally-induced crises and emphasizes corporate governance and firm performance as particularly important aspects.
First, regarding corporate governance, research mostly investigates the influence of the BOD as well as CEO power (Fogarty et al., 2009;Fralich and Papadopoulos, 2018;Koch-Bayram and Wernicke, 2018;O'Connor et al., 2006). The results underscore the general notion that corporate governance aspects should not be considered in isolation (García-Castro et al., 2013;Misangyi and Acharya, 2014). For instance, O' Connor et al. (2006) demonstrate that in case the CEO holds high levels of stock options, internally-induced crises are less likely when the BOD does not hold any options and when the CEO is not chairman of the board. Arguably, in this

MOC Conceptual
Top managers are likely to perceive a crisis as a threat, which makes them likely to react more emotionally and to show more restricted efforts in resolving the crisis.

Cognition
Vuori and Huy (2016) MOC Qualitative Attention structures in the organization can foster shared fears among top managers, which is likely to lead them to exert pressure on their subordinates and to interpret information in biased ways. These mechanisms can foster or exacerbate an upcoming crisis, as they hinder innovation and increase myopia. Cognition Greening and Johnson (1996) MOC Quantitative TMT level of education, organization tenure heterogeneity, and functional heterogeneity are associated with less humaninduced crises; organization tenure is associated with more crises.
Human capital Hambrick and D'Aveni (1992) MOC Quantitative There is a vicious circle between firm performance deterioration and declining TMT functional expertise as top managers exit the organization, finally resulting in firm failure.
Human capital Amatucci and Grant (1993) MOC Qualitative Organizations with TMTs characterized by homogenous functional and experiential backgrounds can be more prone to inertia and, ultimately, organizational crises.

Human capital
Ham et al. (2017) MOC Quantitative CFO narcissism is related to a higher likelihood of crises triggered by accounting restatements.

Personality
Carmeli and Sheaffer

MOC Quantitative
There is a positive association between both top managers' risk aversion and self-centeredness and crisis occurrence.

Personality
Brockner and James

MOC Conceptual
There are certain top manager personality traits that might foster the perception of a crisis as an opportunity; such traits can interact with factors related to the crisis or the focal organization. Specifically, crisis opportunities might be perceived as more attainable, the less the organization is deemed responsible for the crisis, the less severe the crisis is or if the organization's belief system promotes a "can-do" mentality.
Social evaluations

Quantitative
In case a high-status Member of Parliament was involved in a major organizational scandal in British Parliament, this attracted increased media attention, eventually leading to increased turnover among high-status members.

Social capital
Kulich et al.
Social evaluations

Quantitative
In a crisis context, firms mostly opt for a female successor of departing top managers in order to signal change to external stakeholders.
Social capital (continued on next page) scenario, the BOD is more aggressive at monitoring because it has less incentive to let the CEO artificially prop up the company's stock price; in addition, the BOD is relatively powerful, and thus able to curb potentially damaging behavior by the CEO. Moreover, research links two diametrically opposed combinations of firm performance and CEO stock ownership to the occurrence of internal crises: first, firm overvaluation and high stock-option based CEO pay, because it induces CEOs to take excessive risks in the hope of further increasing expected gains (Fogarty et al., 2009); and second, low firm performance and little CEO stock ownership, because CEOs succumb to excessive risk-taking as they think there is not much left to lose anyway in a crisis (Zhang et al., 2008).
For another factor, firm strategy, Zona et al. (2013) induce from their case study of 'Bancopoli'-a major scandal in the Italian banking sector-that an imbalanced strategy that overly focuses on aggressive takeover endeavors may strengthen the relationship between a CEO's lack of moral values and an organizational crisis. These scholars argue that such CEOs might use constant restructuring activities implied by such a strategy to divert stakeholders' attention away from corporate wrongdoing. In addition, if an acute crisis is already ongoing, and coupled with poor corporate reputation or a lack of corporate philanthropic activities, firms tend to establish a particularly great number of new CEO ties to independent foundations' boards-probably in an attempt to recuperate lost social approval (Lungeanu et al., 2018).
The firm's actors and their characteristics constitute further important contingency factors. The relevant studies in our sample mainly use CEOs' characteristics-e.g., founder status or tenure-as indicators of their risk preferences, which moderate the effect of CEO stock option pay on the emergence of internally-induced crises (Wowak et al., 2015;Zhang et al., 2008). Also, research has made first forays into empirically corroborating the suggestion that the TMT influences the effect of CEO characteristics in a crisis context (König et al., 2020b), especially in such a way that CEO and TMT complement each other (Kashmiri et al., 2017).
Moreover, environment-level context, particularly consumers and the macroeconomic situation, influences CEOs' role in a crisis (Laufer et al., 2018;Rule and Tskhay, 2014). For instance, Laufer et al. (2018) examine whether power distance-a value orientation indicating the extent to which a person will tolerate that power is unequally distributed (Hofstede, 2001)-makes consumers more likely to accept and trust authority figures like the CEO. Indeed, these scholars find that consumers' level of power distance reinforces the positive relationship between having a CEO as crisis spokesperson and consumers' future purchase intentions.

Effects of TMT characteristics
Studies on the TMT frequently adopt MOC or agency theory as main theoretical lenses, as indicated by Table 3b. Notably, no study in this section adopts a resource dependency perspective. This section includes both studies that investigate the TMT on a group-level as well as studies on individual top managers other than the CEO that are part of the TMT.
Agency theory. Although the studies in this category illuminate different TMT variables-cognition, financial incentives, and power-they all revolve around the role of top managers' underlying motivation as driver of their actions. First, regarding cognition, Kulik (2005) highlights the negative effect of Enron's top managers' agency reasoning which led them to pursue their own self-interest and to maximize their bonuses, leading to the firm's collapse.
Second, as for top manager financial incentives, and similarly to studies on the CEO, research illustrates the drawbacks of different remuneration vehicles in a crisis context, such as salary (Shi et al., 2016) and share ownership (Berger et al., 2016). For instance, Shi et al. (2016) incorporate considerations from tournament theory (Lazear and Rosen, 1981) and suggest that a high CEO salary makes this position especially desirable for other top managers. As a result, these top managers not only work harder but also tend to engage in misconduct, as Shi et al. (2016) demonstrate by showing increased crisis occurrence resulting from securities class action lawsuits. When it comes to role-specific incentives, research paints a more positive picture. For instance, Kashmiri and Brower (2016) argue and empirically show that top managers, in this case, the Chief Marketing Officer (CMO)-owing to their role and the related incentivization-might act as advocates of specific causes in the TMT, which reduces the likelihood of crises related to these causes, in this case product-harm crises.
Third, Aebi et al. (2012) also take up the notion of top managers as advocates for specific causes, but focus on power. More specifically, they show that banks with Chief Risk Officers (CROs) that are relatively powerful-i.e., with direct reporting line to the BOD-performed significantly better during the 2007-2008 financial crisis. Apparently, a powerful CRO-whose very task it is to manage risk-is better able to balance the CEO's interest to maximize sales and profits with adequate risk strategies, thereby ensuring Conceptual Key crisis antecedents can include top managers' "motive," i.e., the outlook of getting stock options with high monetary and immediate value and a low perceived risk and "means," i.e., a large degree of relative power over the BOD and external agents.
Financial incentives; Power superior firm performance in a crisis. MOC. As for MOC, one theme naturally relates directly to cognition-specifically, top managers' perception of crises as threats versus opportunities. Generally, the literature posits that top managers are more likely to perceive a crisis as a threat rather than as an opportunity and, in turn, to react emotionally and show constrained efforts in resolving crises (James et al., 2011). Such threat rigidities, and ultimately the crisis itself, can be further exacerbated in case top managers focus their attention on the wrong issues, as this can considerably hamper important success factors for crisis resolution, such as effective information processing (Vuori and Huy, 2016) and strategic adaptation (D'Aveni and MacMillan, 1990).
Intriguingly, strategic leaders' general tendency to perceive crises as threats (James et al., 2011) might be attenuated or even overridden by certain top manager personality traits that foster the perception of a crisis as an opportunity. For instance, top managers characterized by a high degree of self-efficacy and learning-orientation might view crises as challenging situations to be overcome and focus on the important lessons they can draw from these episodes, which can ultimately enhance organizational innovation reputation (Brockner and James, 2008).
Other relevant top manager personality traits include narcissism, risk aversion, and self-centeredness. First, and consistent with studies on CEO narcissism (Chen, 2010;Kashmiri et al., 2017), research on top manager narcissism links this trait to the occurrence of internally-induced crises, arguably because such top managers' exorbitant feeling of self-entitlement induces them to engage in actions that seem beneficial to them, even if that means violating rules and running the risk of triggering crises (Ham et al., 2017). Further executive personality factors connected to crisis occurrence are risk aversion and self-centeredness (Carmeli and Sheaffer, 2009). Notably, the results of Ham et al. (2017) and Carmeli and Sheaffer (2009) seem conflicting at first glance, particularly in light of research showing narcissistic top managers to be typically less risk-averse (Buyl et al., 2019). However, Ham et al. (2017) focus on internally-induced crises due to misconduct, which might indeed be related to increased top manager risk-proneness. In contrast, Carmeli and Sheaffer (2009) focus on crises resulting from inertia in response to changes in the firm's environment, which might be related to top managers' risk-aversion.
Research on another theme, TMT human capital, highlights the beneficial effects of TMTs' expertise and cognitive diversity. Notably, and in line with general TMT research (Carpenter et al., 2004), the studies in our sample draw on various demographic measures as proxies for the abovementioned characteristics. Elements of TMT expertise that are particularly relevant for avoiding or overcoming organizational crises include education (Greening and Johnson, 1996) and functional expertise (Hambrick and D'Aveni, 1992), as well as the fresh perspectives and different skill-sets of new TMT members (Mueller and Barker, 1997).
Similar to TMT expertise, TMT cognitive diversity-proxied by TMT heterogeneity and size-is deemed beneficial regarding both crisis occurrence and resolution (Amatucci and Grant, 1993;Greening and Johnson, 1996;Mueller and Barker, 1997). This is interesting given that prior, non-crisis related research has repeatedly highlighted the inconsistent results of managers' cognitive diversity (Finkelstein et al., 2009;Pitcher and Smith, 2001). However, a sociocognitive view on heterogeneity (e.g., Priem, 1990) suggests that, in crisis situations, the effect of heterogeneity seems rather straightforward. According to this view, TMT diversity provides increased "sociocognitive horsepower" (Carpenter, 2002, p. 277), which improves strategic decision-making in high-complexity situations-such as organizational crises.
Social evaluations. Studies that adopt a social evaluations lens on TMTs exclusively focus on the effect of top managers' social capital and mainly investigate TMT prestige, one facet of social capital. For instance, research has found that over the course of an organizational crisis, TMT prestige-as indicated by top manager status characteristics such as prior managerial experience, board connections, or military elite membership-declines because top managers leave the organization, ultimately leading to firm failure (D'Aveni, 1990). Interestingly, not only the absence but also the presence of prestige can imply potential hazards for crisis-stricken organizations. This is illustrated by research based on elite targeting, which happens when external evaluators impose higher standards of conduct on elites and scrutinize them more than non-elites. In particular, research shows that in case a high-status leader is involved in an organizational scandal, this attracts increased media attention, eventually leading to increased turnover among high-status members (Graffin et al., 2013). Intriguingly, in order to signal change to external stakeholders subsequent to such internally-induced crises, organizations often opt for a female successor (Kulich et al., 2015).
Executive ethical leadership. Research that illuminates TMTs in a crisis context from an executive ethical leadership perspective is extremely sparse and focuses, similar to studies on the CEO (e.g., Soltani, 2014), on top managers as potential crisis triggers. A notable comprehensive study in this domain is that by Pendse (2012, p. 265), who suggests that the interplay of top managers' motives, means, and opportunity can create "an ethical hazard" that fosters questionable top manager decisions. This scholar identifies top managers' financial incentives-i.e., stock options with high and immediate value-and power-especially relatively to the BOD and external agents like auditors-as meaningful crisis antecedents.

Contingency factors regarding the effect of TMT characteristics
Analogous to research on direct effects of TMT characteristics in crises, research on contingency factors is less frequent than corresponding literature that focuses on CEOs. Also, extant research on contingency factors is often conceptual and awaits empirical scrutiny.
Crisis attributes. Among the few relevant studies illuminating crisis attributes as contingency factors (e.g., Chng et al., 2015), Brockner and James (2008) provide perhaps the most encompassing theorizing. In particular, they investigate how different crisis attributes-crisis origin, severity, and type-affect executives' crisis perceptions. Regarding (perceived) crisis origin and severity, they suggest that top managers are less likely to perceive attainable opportunity in a crisis in two scenarios: first, the more the organization is seen as responsible for its occurrence; second, in case of a major, all-encompassing crisis. In turn, as for crisis type, these scholars suggest an interactive effect with personality in a way that congruency between top managers' dominant regulatory focus (e.g., promotion focus) and type of opportunity associated with the crisis (e.g., foster positive events in the future) might lead top managers to assign larger value to the opportunity, helping them to perceive the crisis as an opportunity.
Context. Crisis research on top managers that examines the contingency effects of firm-level context covers firm strategy as well as compensation schemes. Although these themes are similar to those in crisis research on CEOs, the underlying theories and specific  (Lazear and Rosen, 1981) to examine the effect of organizations' unrelated diversification, i.e., their engagement in various heterogeneous businesses (Baysinger and Hoskisson, 1989). Specifically, research empirically corroborates that unrelated diversification fosters internal competition among executives which intensifies the impact of the pay gap between the CEO and other top managers on the negative effort the latter exert, i.e., volitional actions that harm shareholders' interests and can eventually lead to internally-induced crises (Shi et al., 2016).
Second, research on the effect of compensation assumes the risk-bearing property perspective, which is based on behavioral agency theory (Wiseman and Gomez-Mejia, 1998). Specifically, this research argues and empirically demonstrates that performance-based pay transfers risk from shareholders to crisis-stricken top managers, which increases their aversion to losses, for instance reputational damage, and hence makes them focus their efforts on impression management activities in the face of a crisis (Chng et al., 2015).
The studies in our sample also include additional theoretical propositions and qualitative findings. For instance, and adding to their propositions on the effect of crisis attributes, Brockner and James (2008) posit that top managers might be more prone to perceive opportunities entailed by a crisis as attainable if their organization promotes a 'can-do' mentality-comparable to a collective feeling of efficacy. As another example-in line with related research on CEOs (Sohn and Lariscy, 2012)- Gillespie et al. (2014) suggest that the 'fit' of newly appointed TMT members (e.g., dedicated turnaround experts) to battle a crisis can amplify the positive effect of new TMT appointments on organizational trust repair.
Similar to comparable research on CEOs, environment-level context remains largely unexplored. An exception is Shi et al. (2016) who show that environmental uncertainty attenuates the positive relationship between vertical pay gaps and the likelihood of internally-induced crises. They argue that high degrees of dynamism and complexity in the environment represent noise when it comes to determining executives' promotion outcomes, decreasing their incentive to become involved in potentially detrimental efforts to close the pay gap.

Effects of BOD characteristics
Research on the BOD most frequently adopts agency theory as theoretical lens, as shown in Table 3c. Particular to research on BODs in crises-and mirroring research on boards in general (Hillman and Dalziel, 2003)-is another theoretical lens, namely resource dependency theory. While the literature mentions the BOD as a contingency factor when examining organizational crises from an executive ethical leadership perspective (e.g., Soltani, 2014), direct crisis-related effects of the BOD remain yet to be investigated from this perspective.
Agency theory. Most literature using this lens emphasizes human and social capital aspects of BOD composition, often focusing on the role of board independence and CEO duality. Regarding board independence, findings suggest a beneficial effect when it comes to avoiding crises (Beasley, 1996;Neville et al., 2018) as well as fostering their resolution (Dowell et al., 2011;Erkens et al., 2012;Mueller and Barker, 1997). These results are in line with one central tenet of agency theory-that independent boards curb problems related to ownership-control separation, such as managerial misconduct (e.g., Eisenhardt, 1989).
Regarding CEO duality, in contrast, the evidence is less clear. When it comes to avoiding crises, research shows that CEO-board chair separation is beneficial (Neville et al., 2018). This in line with the traditional agency-theoretic view which emphasizes that boards with a director other than the CEO are more effective at monitoring (Harris and Helfat, 1998). In contrast, as for crisis resolution, evidence suggests that firms with a CEO who is also board chair fare better (Mueller and Barker, 1997). A possible explanation might be that it is easier for dual CEOs to rapidly implement critical decisions, a key success factor during acute crises (Boyd, 1995).
Research has suggested further relevant board characteristics, such as expertise, age, size, and interlocks (Fogarty et al., 2009). While studies on director expertise (Agrawal and Chadha, 2005) and age (Grove et al., 2011) show a beneficial effect on crisis outcomes, the evidence on BOD size is less clear. Some studies indicate that firms that successfully navigate a crisis have larger BODs than their unsuccessful counterparts (Chaganti et al., 1985); other research reveals that large BODs might be dysfunctional in a crisis (Dowell et al., 2011). These inconsistent results reflect the general discussion in corporate governance research which emphasizes that there is literally no one-size-fits-all solution (Coles et al., 2008). Specifically, additional BOD members imply benefits (e.g., access to more resources) and costs (e.g., coordination problems) (Ning et al., 2010). Grove et al. (2011) have demonstrated this trade-off as an inverted U-shaped relationship of board size and bank performance during the financial crisis. What is more, the optimal board size can differ depending on industry-and firm-specific characteristics (Coles et al., 2008;Ning et al., 2010). This reasoning might also help to explain the divergent results between Dowell et al. (2011) andChaganti et al. (1985). The former investigate young internet firms in the 2000s, a setting characterized by large uncertainty, implying a greater relative value of quick BOD decision-making. The latter, in contrast, examine retail firms in the 1970s, presumably a less turbulent environment, implying a greater relative value of the BOD's ability to make well-founded strategic decisions.
Research on other BOD characteristics is sparse and limited to power. Generally, research in this realm links relatively powerless BODs to an increased occurrence of internally-induced crises (Mellahi, 2005;Nguyen et al., 2015), hence mirroring studies on the CEO that illustrate that powerful CEOs can increase the risk of internal crises (e.g., Agrawal and Chadha, 2005).

MOC.
A further-although less frequently employed-lens applied to understand the crisis-related effects of BOD characteristics is MOC. In that regard, the extant research focuses on director cognition. Scholars have placed special emphasis on directors' informationprocessing which can be impaired by cognitive biases before and amidst a crisis (Pirson and Turnbull, 2011;Rost and Osterloh, 2010). For instance, and quite paradoxically, Rost and Osterloh (2010) find a negative association between directors' expertise and firm performance during crises, suggesting that experts tend to fall victim to biased information-processing as they are likely to overestimate the correctness of their decisions in a familiar field under uncertainty.
In addition, and similar to research on top managers (Kulich et al., 2015), research on BODs reveals a tendency towards hiring female strategic leaders during crisis episodes and concludes that female directors can broaden the board's perspective and thus improve board-level decision-making. This helps to seize opportunities in the aftermath of a crisis, leading to a stronger recovery in the later stages of a crisis (Sun et al., 2015).
Not only director appointments, but also director exits are vital in a crisis. Hambrick and D'Aveni (1992) show that firms which do not successfully navigate a crisis lose more and more of their outside directors as the crisis unfolds-an effect that accelerates over the course of the crisis. These scholars suggest two main underlying mechanisms: first, talented top managers leave the 'sinking ship,' and potential successors are hard to attract due to limited resources; second, the remaining executives feel increasingly pressured, leading to more strategic errors.
Social evaluations. Research that investigates boards in a crisis context from a social evaluations perspective is extremely sparse and focuses on the potentially harmful effect of social capital in terms of directors' ties to other organizations. The underlying concept is diffusion, suggesting that directors who serve on multiple companies' BODs will transfer managerial practices from one company to the next (Davis, 1991). Janney and Gove (2017) empirically examine such diffusion in the context of the 2006/2007 US stock option backdating scandal and find that firms with director interlocks with another firm that experiences an internally-induced crisis suffer subsequent decreases in stock market performance.
Resource dependency theory. Exclusive to crisis research on the BOD is the resource dependency theory lens. Like in the agencytheory based studies in this review, the main themes are social and human capital. Generally, research in this domain highlights the beneficial effect of the linkages of directors to other firms on crisis-related outcomes. Particularly, research shows that BODs with high advising capabilities-proxied by the number of directors of other firms to which the board is collectively connected-prevent internally-induced crises, as such BODs can help the CEO make better-informed decisions that comply with rules and regulations (Nguyen et al., 2015). Likewise, the number of board interlocks is linked to increased probability of firm survival in a crisis-stricken industry, probably because directors can use these links to secure funding and access valuable restructuring expertise (Filatotchev and Toms, 2003). One key aspect here is directors' actual involvement. Notably, if interlocked directors show effort-in terms of attending board meetings-this helps the BOD to accumulate and process relevant information, which is conducive to successfully overcoming an acute crisis (Arora, 2018).
In sum, these results complement the findings based on the social evaluations perspective. While evidence from the latter suggests that under certain circumstances-e.g., if firms are affiliated with an organization in crisis-linkages may indeed become harmful (Janney and Gove, 2017), evidence from resource-dependency-based studies points towards a generally beneficial effect of directors' linkages to other organizations for the focal firm.

Contingency factors regarding the effect of BOD characteristics
Extant research on contingency factors regarding the BOD's role in organizational crises is rather rare and fragmented. Regarding crisis attributes, the literature has examined crisis origin and crisis stage only. As for crisis context, prior studies covered aspects such as corporate governance, directors' identity, and country specifics.
Crisis attributes. One contingency factor discussed in the literature is (perceived) crisis origin-more precisely, perceived locus of crisis causality. Drawing on identity research (van Dick et al., 2004), Withers et al. (2012) argue that perceived locus may influence the likelihood of outside directors voluntarily leaving the organization. For instance, directors might be more likely to exit when they perceive that the crisis is internally-caused and when, at the same time, their salient identity is coupled to being a director. The underlying motivation is to distance themselves from the firm to maintain their reputation and market value as a director.
Further research demonstrates that crisis stage and board composition can interact: as Sun et al. (2015) show, in an early stage of an economic crisis, companies governed by BODs with a larger share of female directors show less investment growth than BODs with a lower share of female directors; in the aftermath of a crisis, the opposite is the case. These scholars argue that more diverse BODs possess broader perspectives, which allows them to make superior strategic-decisions and to act more adequately in later stages of a crisis.
Context. Regarding firm-level context, the studies in our sample focus on contingency factors related to corporate governance mechanisms (Fogarty et al., 2009;Janney and Gove, 2017) and individual directors (Withers et al., 2012). Most notably, Janney and Gove (2017) investigate whether investors perceive that directors and professional service firms act as conduits of organizational malpractice. Indeed, they empirically show that firms experience particularly severe decreases in performance in case they are dually linked, first, via a BOD interlock, and second, via having a law firm partner as advisor to another firm that suffers an internally-induced crisis. The same holds true in case of multiple interlocks at the same time.
Empirical evidence on the impact of environment-level context on strategic leaders' role in crises is almost nonexistent. A notable exception is a study on the impact of country-level corruption. Building on organizational behavior research (e.g., Ashforth and Anand, 2003;Misangyi et al., 2008), Neville et al. (2018) argue that corruption is systemic in nature, trickles down from the national to the organizational level, and persists as corrupt practices become routinized, habitual, and institutionalized within corporate governance policies. As predicted, these scholars find that the degree of corruption in a country weakens the negative association between board independence and the occurrence of internally-induced crises.

An integrative framework for strategic leadership in crises
The results from our systematic review allow us to develop our initial baseline model into a comprehensive conceptual framework that integrates the findings from several theoretical lenses to consolidate the current research landscape. At the same time, they serve as valuable starting point for future research. Our final conceptual framework is shown in Fig. 3.
Central to the framework-and much in line with our baseline model from Fig. 2-are the different types of strategic leaders and their characteristics. Our review reveals that these characteristics cover the whole range from easily readable cues to more hidden ones, thereby echoing Hambrick and Mason, who emphasize the importance of both "observable" and "psychological" characteristics (1984, p. 198). Notably, for each type of strategic leader, different theoretical lenses and characteristics examined prevail. Unique to the CEO, for instance, and consistent with the notion of the CEO as the "public 'face' of the company" (Men, 2012, p. 171), is the focus on appearance. This is underpinned by social evaluations theory, which is, at the same time, the most frequently employed theoretical lens in studies on the CEO.
Research on the TMT is heavily influenced by the MOC perspective, which is also reflected in the main themes-cognition and personality. This focus also resonates with voices in the MOC literature which highlight the TMT as a particularly relevant actor when it comes to influencing organizational outcomes (Hambrick, 2007). There are two potential shortcomings regarding TMT-related studies. First, many studies in this sub-sample, particularly those of conceptual nature, do not clearly specify which individual top manager positions they refer to, leading to ambiguity and difficulty in generalizing results. Second, consistent with a general tendency in MOC-based studies (Neely et al., 2020), the extant literature mainly focuses on demographic factors aggregated to the group-level as proxies and thereby disregards underlying mechanisms, e.g., social dynamics.
Regarding the BOD, the main theoretical lens employed-agency theory-accounts for the predominant focus on human and social capital, mostly operationalized as board independence, which is one of the key BOD characteristics also in the broader strategic management literature (Dalton et al., 1998). In addition, our framework indicates the relevance of interactions between different types of strategic leaders, although research has only made first advances in that regard (e.g., Kashmiri et al., 2017).
Regarding crisis attributes, the first type of contingency factors, we see that the extant literature covers a considerable variety of attributes, but nonetheless is rather patchy. For instance, the current research focuses, with few exceptions (e.g., Wowak et al., 2015), on crises related to corporate fraud or the global economic crisis 2007-2008. Hence, it does not account for the full diversity of crises. Similarly, regarding context, the second type of contingency factors, the extant literature covers a broad variety of different aspects. However, to date, research fails to investigate these aspects more deeply, and there is no research that looks at multiple aspects of context at the same time. Perhaps, this is because contingency factors are mainly investigated as moderators and seldom the explicit focus of the studies in our sample. For example, Kashmiri et al. (2017) examine the moderating effect of a strong marketing department in the TMT. However, these scholars blend out TMT sub-groups of other potentially relevant departments, such as product Fig. 3. Final conceptual framework for strategic leadership in organizational crises. development, even though they often act as antagonist of the marketing department (Ou et al., 2017).
Third, this review illustrates the myriad of studied organizational-crisis-related outcomes, which can largely be traced back to the influences of the respective theoretical lenses used. For example, studies using the social evaluations or executive ethical leadership lens mostly explain outcomes such as restoration of reputation, whereas research based on agency theory often explains outcomes like crisis occurrence. Interestingly, any reverse effects, i.e., the impact of crises and their outcomes on strategic leaders, largely remain open to investigation. Based on this and other research gaps, the next section will suggest concrete future research opportunities.

Future research suggestions
Based on our review and the resulting conceptual framework, we can identify promising research opportunities in each major building block of the framework, i.e., strategic leaders and their characteristics, contingency factors, and organizational-crisis-related outcomes. These research opportunities are summarized in Table 4. Below, we outline these opportunities and develop concise suggestions on how to pursue them conceptually and empirically.

Extend the social evaluations perspective of CEO-focused research
A plethora of research avenues emerge regarding the specific focal actors. As for the CEO, it might be fruitful to dive deeper into the social evaluations perspective, which encompasses three particularly promising opportunities: perceptions of other stakeholder groups, CEOs' management of rhetorical dilemmas, and CEOs' pre-crisis endowment of social approval.
First, even though crisis researchers have made first attempts to investigate the CEO's role in shaping the perceptions of various external stakeholders (Einwiller et al., 2017;Gomulya et al., 2017;Lungeanu et al., 2018), with only a few exceptions (Gillespie et al., 2014), these researchers have so far mostly neglected important other stakeholder perceptions, perhaps most notably that of the focal Table 4 Opportunities for future research and suggested research questions.

Research opportunities (reference to framework)
Main themes (underlined) and suggested research questions

Extend the social evaluations perspective of CEO-focused research (Strategic leaders-CEO)
Other stakeholder groups: What is the CEO's role in shaping perceptions of internal stakeholders (e.g., employees) during crises? Rhetorical dilemmas: Which trade-offs do CEOs face when communicating with heterogenous audiences in a crisis context? How does a crisis-communication strategy chosen by the CEO affect different dimensions of social evaluations? Pre-crisis endowment of social approval: In how far do stakeholders perceive the role of the CEO in resolving crises differently depending on the CEO's pre-crisis endowment?

Deepen insights on the TMT from an MOC view by studying underlying mechanisms (Strategic leaders-TMT)
Decision-making processes: What is the effect of TMT decision-making on organizationalcrisis-related outcomes? Faultlines: In how far do organizational crises trigger faultlines in TMTs? What is the effect of additional, non-traditional faultlines (e.g., membership in the BOD)?

Examine individual top managers' roles from an executive ethical leadership lens (Strategic leaders-TMT)
Newly established top manager roles: What is the impact of newly established top manager positions (e.g., Chief Ethics Officer) on crisis outcomes? Which influence does the importance these top managers place on moral values or the power they possess have?

Expand range of investigated (individual) directors' characteristics using different lenses (Strategic leaders-BOD)
Director personality and values: What is the effect of directors' personality traits (e.g., conscientiousness) and values (e.g., benevolence) on organizational-crisis-related outcomes? Role of individual directors: Are specific types of individual directors (e.g., customers, family directors) helpful in a crisis context?

Shed light on the interplay between strategic leaders (Strategic leaders-Various)
Complementary traits: Do personality traits (e.g., temporal orientation) of different individual actors complement each other? Multiteam systems: Which interaction modes (i.e., interdependence and/or independence) between different leaders are most effective during crises? What is the role of 'boundary spanners' (e.g., the CEO) that link TMT and BOD?

Broaden the exploration of contingency factors (Contingency factors)
Diversity of crisis settings: What is the role of strategic leaders during specific types of crises (e.g., cyber-attacks)? Are there ideal strategic leader 'prototypes' for different crisis scenarios? If yes, what are they? Holistic perspective on contingency factors: How do holistic sets of conditions influence strategic leaders' impact in a crisis context? New contingency factors: What is the effect of contingency factors not investigated so far (e.g., prior crisis experience)? Geographical differences: Do results vary-and if yes, how-across different geographies (e.g., North America vs. Asia)?

Examine the impact of crises on strategic leaders (Organizational-crisis-related outcomes)
Malleable leader characteristics: Do crises trigger a change in strategic leaders' characteristics (e.g., cognition)? Leader succession: How do crises and their consequences affect turnover events (e.g., successive leaders' characteristics)?
organization's employees (Coombs, 1995). This particularly includes high-ranking executives, who naturally have regular direct interactions with the CEO (Georgakakis et al., 2019). Prior research has found that the departure of such executives is linked to adverse crisis-related outcomes (e.g., Hambrick and D'Aveni, 1992) and proposed an effect of strategic leaders' evaluations of the CEO on their decision to leave the firm during acute crisis episodes (Withers et al., 2012). However, we are not aware of any research that investigates these evaluations as underlying driver of executives' behavior in crisis episodes. As traditional data gathering methods, e.g., surveys, seem less suitable due to issues such as notoriously low response rates among top managers (Cycyota and Harrison, 2006) and social desirability bias (Fisher, 1993), future research would be well-advised to follow prior research and draw upon unobtrusive measures such as corporate imagery (Chatterjee and Hambrick, 2007;Ray and Smith, 2011) and texts (Aerts and Yan, 2017;Merkl-Davies et al., 2011). Second, scholars have neglected the rhetorical dilemmas (König et al., 2018) CEOs need to manage in light of substantial audience heterogeneity (Ertug et al., 2016;Kim and Jensen, 2014) and different dimensions of social evaluations (Bundy and Pfarrer, 2015;Pollock et al., 2019). For audience heterogeneity, scholars might compare the trade-offs CEOs face in symbolic communication  vis-à-vis analysts and journalists. For example, analysts might respond highly differently to accommodative crisis communication than journalists. Likewise, any given crisis-communication strategy chosen by the CEO might affect different dimensions of social evaluations-e.g., legitimacy or reputation-differently. Thus, CEOs might face particularly intriguing dilemmas in managing these social evaluations.
Third, studies might also consider the CEO's prior "endowment" (Bundy and Pfarrer, 2015, p. 346) of positive social judgements. For example, stakeholders might perceive organizational crises and the CEO's role in resolving these episodes differently, depending on the CEO's celebrity status . Empirical advances in measuring social evaluations and CEO characteristics-including large longitudinal analyses using quantitative, partially multimodal content analysis (e.g., Graf-Vlachy et al., 2020;Harrison et al., 2019)-might be helpful to test these and other emerging hypotheses.

Deepen insights on the TMT from an MOC view by studying underlying mechanisms
Regarding the TMT, we see particular promise in deepening the MOC perspective. In particular, to shed light on the effect of underlying cognitive and social mechanisms in a crisis context, scholars could draw upon research on TMT decision-making and faultlines. First, the effect of TMT decision-making, which is shaped by TMTs' behavioral integration, on crisis-related outcomes might be worthwhile to investigate, as prior research has highlighted the importance of such TMT characteristics in other critical situations (Carmeli and Schaubroeck, 2006). Second, it might be fruitful to apply the concept of faultlines (Lau and Murnighan, 1998) to extend the understanding of TMT group dynamics. For example, researchers could investigate whether, how, and under which circumstances organizational crises trigger faultlines in TMTs-a phenomenon scholars have already observed for work teams (Meyer et al., 2015). Moreover, scholars could move beyond traditional faultlines like age or sex (Thatcher and Patel, 2012) and look at additional ones, such as membership in the BOD. Future research could also combine these endeavors with the large, but still evolving, research on the effect of cognitive frames such as threat vs. opportunity (König et al., 2020a) on sociocognitive and political processes in TMTs. Suitable empirical approaches might include experiments (Schulz-Hardt et al., 2002), self-report measures (Carmeli, 2008), and ethnographic approaches (Vesa and Vaara, 2014).

Examine individual top managers' roles from an executive ethical leadership lens
To illuminate the role of individual TMT members, future research could refer to the executive ethical leadership lens. Previous research in this regard has this far predominantly focused on the CEO (e.g., Chen, 2010). However, it might be compelling to consider other individual top managers, and at the same time, address the limitations in many current studies that do not explicitly state which specific top manager roles they cover. A particularly interesting approach might be to examine novel top manager positions, such as Chief Ethics Officer (Fombrun and Foss, 2004) or Chief Compliance Officer (Weber and Wasieleski, 2013). In fact, many firms have introduced such new positions as a reaction to a wave of organizational crises that involved a strong ethical component, such as the ones at Enron and WorldCom (Gardberg and Fombrun, 2006;Morf et al., 1999). Yet, notably little, if any, research has illuminated the impact of these new types of executives on crises-related outcomes. As for these managers' characteristics, it might be interesting to investigate the importance they place on moral values (Nystrom, 1990) or how powerful they actually are, e.g., in terms of responsibilities and reporting opportunities (Weber and Wasieleski, 2013). Since scholars have only begun to study individual top managers' role in crises from this theoretical lens, we encourage explorative, grounded approaches, e.g., in-depth interviews (Zona et al., 2013).

Expand range of investigated (individual) directors' characteristics using different lenses
As for the BOD, we see two particularly fruitful future research avenues: first, investigating new, and probably highly relevant BOD characteristics; and second, deepening insights on individual directors. First, future research could move beyond the focus on human and social capital (e.g., Fogarty et al., 2009) and examine the impact of director personality and values. As such research would break new ground, common theoretical lenses-e.g., MOC-as much as less common ones-e.g., executive ethical leadership-might be used. In terms of MOC, research could examine directors' personality. To do so, scholars might look at personality traits used in prior studies on other strategic leaders in a crisis context-e.g., narcissism (Buyl et al., 2019)-or follow general strategic management research (e.g., Torchia et al., 2015) to examine additional personality traits that are particularly important for directors. This could include, for instance, conscientiousness, which has been shown to be associated with directors' monitoring activities (Guerrero et al., 2017). Thus, boards with conscientious directors might be especially valuable in a crisis context, for example to prevent internally-induced crises (Nguyen et al., 2015). As for executive ethical leadership-a lens that has to date not examined the BOD as focal actor-crisis scholars might follow recent research in strategic management (e.g., Nahum and Carmeli, 2020;Olthuis and van den Oever, 2020) and investigate the impact of directors' values. After all, values like benevolence and tradition seem to play an important role in directors' decisions (Adams et al., 2011). A benevolent BOD, per definition, prioritizes in-group members' welfare, and hence, might be particularly effective in crafting crisis responses aimed at internal stakeholders but less effective in addressing external stakeholders. A suitable approach to elucidate directors' personality and values could combine quantitative elements-e.g., questionnaires (Carmeli and Sheaffer, 2009) or proxy measures (McClelland et al., 2010)-and qualitative elements-e.g., in-depth interviews (Zona et al., 2013).
Second, it might be worthwhile to examine the BOD not only on the group-level, but also on the individual-level, analogous to research on TMTs (Menz, 2012). After all, literature on BODs highlights the important role of individual directors for board effectiveness (Hambrick et al., 2015;Petrovic, 2008). In this context, future research could paint a more differentiated picture on human and social capital factors by moving beyond the current focus on board independence. For instance, scholars could follow recent work that emphasizes the importance of customers as BOD members of business-to-business firms (Bommaraju et al., 2019) to examine whether such directors are similarly beneficial in a crisis setting, e.g., during product-harm crises. In addition, and in line with recent work that highlights the relevance of owners as directors (Federo et al., 2020), scholars could illuminate the role of family directors. Although research has linked family representatives on the BOD to low firm performance (e.g., González-Cruz and Cruz-Ros, 2016; Villalonga and Amit, 2009), when it comes to battling an acute crisis, such directors' "survivability capital" (Sirmon and Hitt, 2003, p. 343) could be a decisive factor.

Shed light on the interplay between strategic leaders
Besides looking at individual actors, future research might also investigate the interplay between multiple actors. This might comprise different constellations, such as interactions between individuals (e.g., CEO and CFO) or between two groups (e.g., BOD and TMT). First, and in line with other scholars (König et al., 2020b), we propose to empirically examine whether the personality traits of different individual actors complement each other. For example, future research could build upon the emerging perspective on strategic leaders' temporal orientation (e.g., Levasseur et al., 2020) and examine the effect of the three temporal foci-i.e., past, present, and future (Tang et al., 2020). In a crisis context, an organization whose CEO scores high on present-and on future-focus and whose CFO scores high on past-focus might have a good outlook. This is likely because a present-and future-focused CEO might seek to swiftly overcome the crisis, but-at the same time-might not take enough time to look back and learn from the crisis. A past-focused CFO, in contrast, might put more emphasis on reflecting on the past and on figuring out what led to the crisis, hence stimulating organizational learning.
Second, regarding group interactions, research might further explore the relationship between TMT and BOD by building on the multiteam systems perspective (e.g., Mathieu et al., 2018). Recent work in this regard emphasizes that different settings might require different interaction modes-characterized by different degrees of attention to working interdependently and independently-between TMT and BOD (Luciano et al., 2020). In particular, it might be worthwhile to explore the role of "boundary spanners"-e. g., the CEO-that link TMT and BOD (Luciano et al., 2020, p. 35). The CEO's boundary-spanning function seems especially vital to bridge potential information asymmetries, which are especially pronounced for outsider-dominated boards that inherently lack daily interaction with the firm (Zorn et al., 2017) and become critical in crisis times which require swift and well-grounded decisions (Pearson and Clair, 1998). Viewing a crisis as a special setting, it might also be promising to explore which interaction modes are most effective during such an episode. Luciano et al. (2020), for instance, refer to literature on working modes in group tasks (e.g., van de Ven et al., 1976) to argue that in case collective tasks are difficult, uncertain, and variable-which all applies to crises-a sole focus on working independently seems contraindicated. Hence, a critical factor to address acute crises might be to encourage TMT and BOD to work both interdependently and independently.

Broaden the exploration of contingency factors
As for contingency factors, it might be promising to explore other crisis settings, novel contingency factors, and different geographies. First, perhaps most clearly, future research could consider a larger variety of crisis settings. While the predominant focus on crises related to fraud or the global economic crisis makes sense for each individual study, it limits the generalizability of findings. This is particularly true because the context of a crisis highly influences the severity of a crisis (Milinski et al., 2016) and the accompanying challenges of crisis management (Rhee and Valdez, 2009). A promising future avenue might thus be to extend research to crisis types that have been studied in other research domains, such as catastrophic accidents (Haunschild and Sullivan, 2002), cyber-attacks (Gwebu et al., 2018), and severe pandemics like COVID-19 (Christensen and Laegreid, 2020). In doing so, future research might also draw a more nuanced picture by comparing strategic leaders' role across different crisis settings. Specifically, scholars could investigate whether there are strategic leader characteristics that are universally beneficial across crises or seek to identify ideal strategic leader 'prototypes' for different crisis scenarios.
Second, future research could jointly explore other contingency factors. As summarized in this review, prior research has investigated a variety of contingency factors. However, a concerted effort to empirically examine a holistic set of conditions-e.g., all relevant corporate governance mechanisms both within the focal firm and the external environment (Misangyi and Acharya, 2014)that influence strategic leaders' impact on crisis management activities and crisis-related outcomes is yet to be made. Our integrative framework could serve as a starting point for such an endeavor. This might also include completely new contingency factors, such as the focal firm's prior crisis experience, e.g., the number of past product recalls (Haunschild and Rhee, 2004). At first glance, prior experience might seem beneficial, as it allows learning from the past (Haunschild and Sullivan, 2002). However, being crisis-tested could indeed negatively affect outcomes like stakeholder evaluations and stock price performance since the media or analysts might hold excessive expectations regarding crisis resolution which might be hard to fulfill. To empirically explore these suggestions, scholars could apply a configurational approach, e.g., qualitative comparative analysis (QCA) (Rihoux and Ragin, 2008), which enables researchers to uncover complex patterns or "recipes for success" (Dwivedi et al., 2018, p. 394). Additionally, it might also be valuable to follow new advances in marginal effect analyses (e.g., Oliver et al., 2018) and examine not only if and how, on average, a certain moderator influences a given relationship, but at what level of a given moderator the focal effect is significant, and how large the effect is (Busenbark et al., 2021).
Third, we encourage scholars to extend research efforts to different geographies. In our sample, almost two-thirds of the empirical studies examine a North American context, as depicted in Fig. 4. However, as highlighted in the international management literature (Hofstede, 2001;Trompenaars and Hampden-Turner, 2011), North America differs from other contexts in many dimensions, which might greatly affect strategic leadership, crisis management, and the interplay between the two. To address these concerns, researchers should attempt to replicate and extend (Tsang and Kwan, 1999) existing studies. In this regard, we consider the study by Koch-Bayram and Wernicke (2018) as particularly suitable: it is set in the US, where military service is voluntary; replications might yield different, or at least insignificant, results in countries where military service is generally compulsory (Pollach and Kerbler, 2011). The use of two-stage stratified sampling-i.e., selecting different countries and subsequently selecting firms within each country to create sub-samples-seems promising (Nielsen, 2010).

Examine the impact of crises on strategic leaders
Finally, it also might be of interest to 'close the loop' and examine the impact of organizational crises on both incumbent and newly appointed strategic leaders. Future research could investigate whether a crisis triggers a change in incumbent strategic leaders' malleable characteristics, e.g., their cognitions. Experimental research has already made first forays in this regard and demonstrated that being confronted with a crisis situation benefits leaders' critical thinking and sensemaking abilities (Powley and Taylor, 2014). To examine this in a real-world setting, scholars might use established questionnaires-e.g., the 'Watson-Glaser critical thinking appraisal' instrument (El Hassan and Madhum, 2007)-and administer them to strategic leaders.
Another topic of interest might be strategic leader succession. Prior research has yielded first insights on this matter as it indicates that crisis-stricken firms, adhering to the prevalent "heroic portrait of dominant CEOs," are more likely to appoint such CEOs, surprisingly with no effect on firm performance (Tang and Crossan, 2017, p. 782). This warrants the adoption of a more differentiated perspective, as initiated by Chen and Hambrick (2012), that accounts for the complexity of crises by investigating predecessor and/or successor fit to the crisis situation in particular and to the organization and its environment in general. For instance, scholars might examine if an organizational scandal increases the probability that outside successors are favored over insiders, or if a product-harm crisis makes it more likely that incumbents with a lack of technical expertise are replaced by successors with a technical background, as well as the respective performance implications of such choices. In addition, prior research has pointed towards the tendency to promote women during crises-known as the "glass cliff" effect (Ryan and Haslam, 2005, p. 81, p. 81)-to signal change (Kulich et al., 2015;Sun et al., 2015). In this context, perhaps most obviously, it might be interesting to investigate if, how, and why female CEOs might be more apt to manage certain types of organizational crises. After all, the COVID-19 crisis has highlighted various female leaders-e.g., New Zealand's Jacinda Ardern-as potentially particularly successful (Cherneski, 2020).

Limitations
Just as most research projects, this literature review has limitations. Most importantly, there are limitations inherent to the nature of a systematic literature review and the associated process to identify relevant literature. We expended substantial effort to survey the extant research on strategic leadership in organizational crises in its entirety, e.g., by extending our search well beyond key strategic management journals. Nevertheless, our review might not cover all relevant work because it is a systematic review and the scope of any such literature review is naturally restricted, e.g., by the set of specific keywords used. Also, we limited our review to academic disciplines we consider most relevant for the intersection of strategic leadership and crisis research. Other scholars might sample from different disciplines.
In addition, the necessity to choose definitions for the terms 'organizational crisis' and 'strategic leaders' as inclusion criteria represents a noteworthy boundary condition which restricts our sample and consequently the insights our review can provide. For example, in order to establish whether the investigated incidents actually represent crises, we drew upon the broad crisis definition by Bundy et al. (2017), which unites insights from an analysis of multiple crisis definitions of the past 20 years. Still, choosing a different crisis definition might yield different results. Also, our review only considers strategic leaders who hold an official position in the organization-the CEO, TMT, and BOD (Finkelstein et al., 2009). Yet, there might be strategic leaders that do not or no longer hold an official position-such as a company's retired founder-but still exert a major influence in crisis times (Ling et al., 2007). Future reviews might focus on these aspects to clarify the role of strategic leaders in such particular settings.

Contributions and conclusion
This article aims to provide a comprehensive review on the role of strategic leadership in the context of organizational crises. By doing so, it makes several contributions. First, by identifying and consolidating the studies that make up the rich but fragmented research landscape on strategic leadership in organizational crises, this review provides a holistic overview of the extant literature. It also offers readers a clear structure of the current research landscape as it maps the studies to the respective underlying theoretical lenses, thereby revealing that each lens, with its unique focus, adds a valuable piece to the overall picture. Second, by developing an integrative, conceptual framework, this review not only provides a concise and structured synthesis of the extant literature, but also offers a starting point for future research endeavors-both regarding studies that particularly focus on strategic leaders in organizational crises and studies that investigate upper echelons effects in other strategically relevant contexts. Finally, by explicitly identifying promising avenues for future research and by outlining concrete recommendations on how to pursue them, this review enables and stimulates further research.
We hope that this review, the proposed integrative framework, and the identified future research opportunities can serve as a helpful guide and source of inspiration for fellow scholars who wish to advance this critical field of research. After all, as we are currently experiencing during the global COVID-19 pandemic, crises have formidable repercussions for organizations, their members, and their environments, and they consequently impose great challenges on strategic leaders. As such, it appears prudent and worthwhile to explore the role of strategic leaders in organizational crises in even greater depth and from varied theoretical perspectives.