1.2.1 Policy Entrepreneurs and Policy Entrepreneurship

Policy entrepreneurs have been theorized extensively in the policy studies scholarship, and their importance as a heuristic for agency has been recognized not only in theories of the policy process but also as a stand-alone concept (Petridou, Reference Petridou2014). All entrepreneurs in the public sphere perform three functions: they discover unmet needs and select suitable solutions for them; they bear the risk (personal, political, or economic) associated with introducing these solutions, and they build teams of individuals willing to work together towards the realization of the proposed solutions (Schneider and Teske, Reference Schneider and Teske1992; Schneider et al., Reference Schneider, Teske and Mintrom1995). Additionally, policy entrepreneurs are creative and strategic in that they must be able to think ahead how their proposals may influence the policy debate not only in the short run but also in the long run. These actors are socially competent and able to read cues so that they may anticipate how others will receive their proposed solutions, for which they must be able to argue persuasively in well-maintained and broad networks. Finally, policy entrepreneurs are expected to be competent leaders of the coalitions they assemble (Mintrom, Reference Mintrom2000). Necessary requirements for the success of policy entrepreneurial action include (i) telling a persuasive story which frames the problem in a way that is attractive to policymakers, (ii) ensuring the preferred solution is available prior to entrepreneurs drawing attention to a problem, and (iii) exploiting a propitious moment in time, a window of opportunity, during which policymakers are willing to listen in order to realize their policy solution (Cairney, Reference Cairney2018).

Less scholarly attention has been paid to the concept of policy entrepreneurship. This is not necessarily surprising, given that the policy entrepreneur has widely been used as a heuristic for agency, which in turn is often treated as a residual variable explaining change when structural variables fail to do so (Capano and Galanti, Reference Capano and Galanti2018). In other words, public policy scholars are interested in the entrepreneur as an actor, not in entrepreneurship as a process. Yet, understanding entrepreneurship can further the understanding of the entrepreneur. We follow Kirzner (Reference Kirzner1973), who viewed entrepreneurship as a process rather than an end state. We define policy entrepreneurship as the process that enables entrepreneurs to discern the most expedient instance to act in order to achieve their goal of affecting change.

1.2.2 What Is a Crisis?

A crisis is an unusual situation that presents some extraordinary challenges for those directly affected (Almond et al., Reference Almond, Flanagan and Mundt1973). Early political science literature drawing from international relations, defined a crisis as:

Εven in the contemporary crisis landscape, where threats include a host of non-state actors and implications that are broader than the engagement in military hostilities, the basic tenets of this definition still hold. Moreover, while there are examples of definitions that allude to the idea that crises are not entirely bad – including the notion crises present opportunities (Dror, Reference Dror1993; Stranks, Reference Stranks1994), that crises contain an element of duality (Drennan et al., Reference Drennan, McConnell and Stark2015) consisting of a historical period of rupture (from the business-as-usual) and a framed event (Gotham and Greenberg, Reference Gotham and Greenberg2014), or that they constitute a turning point or critical juncture (Petridou and Sparf, Reference Petridou and Sparf2017), the idea of crisis as a predominantly negative situation prevails (Shaluf et al., Reference Shaluf, Ahmadun and Mat Said2003). Broadly, the term implies an undesirable and unexpected situation that possesses current or latent harm to people, organizations, or society, engendering feelings of panic, fear, danger, or shock (Darling, Reference Darling1994).

In the crisis management literature, a crisis constitutes “a serious threat to the basic structures or the fundamental values and norms of a system, which under time pressure and highly uncertain circumstances necessitates making vital decisions” (Rosenthal et al., Reference Rosenthal, Charles and Hart1989, p. 10). Boin et al. (Reference Boin, ’t Hart, Stern and Sundelius2017) draw from this definition to conceptualize crises as “critical junctures in the lives of systems – times at which their ability to function can no longer be taken for granted” (Boin et al., Reference Boin, ’t Hart, Stern and Sundelius2017, p. 5). Boin et al. (Reference Boin, ’t Hart, Stern and Sundelius2017) go on to say that during a crisis, a community, organization, region, or country (all of which constitute a system) experiences an urgent threat to the values that they consider fundamental for their existence. This threat necessarily contains a host of unknown contingencies and necessitates urgent action.

The definitions above are not dissimilar to each other in that they all contain the key elements of a crisis: the threat it poses to a system, the time pressure that crisis managers experience when dealing with a crisis, and the uncertainty and the concomitant ambiguity regarding both the causes of the crisis and the “best” ways to manage and contain it in a way that limits harm done. The extent of harm done – in other words, the magnitude of the framed-as-crisis-event – depends on the existing landscape of risk and resilience in a given society, which in turn is informed by power differentials and degree of inequality of environmental protections (Gotham and Greenberg, Reference Gotham and Greenberg2014). Failure to successfully manage the crisis raises important questions about the competence of the governing coalition and can bring governments down. The so-called Partygate Scandal in 2020–2021, for example, eventually led to the resignation of UK Prime Minister Boris Johnson and considerable leadership turmoil in the governing party in 2023. In the United States, the poor response to Hurricane Katrina in 2005 led to significant criticism of the Bush administration and greatly weakened its second term.

By definition, the crisis management literature offers guidance regarding how disasters should be managed – that is, how their adverse effects can be prevented, mitigated, or controlled. Successful crisis management is generally good for society because it can guide society away from harm. Still, successful crisis management is often framed in terms of its value to the entity managing the crisis. The historical trajectory of crisis management spans an extensive and intricate timeline characterized by the gradual evolution of strategies and practices employed by civilizations to navigate and mitigate a variety of crisis events. The origins of crisis management can be traced as far back as in ancient civilizations such as the Egyptians, Greeks, and Romans (Rainbird, Reference Rainbird1976), and success or failure reflected accordingly on the ruling entity governing the crisis management efforts.

In the contemporary era, crisis management has seen significant advancements in dealing with emerging complexities such as terrorism, cybersecurity vulnerabilities, and pandemics. National and international institutions and private sector organizations have produced comprehensive crisis management plans and strategies. Over time, the field of crisis management has demonstrated a capacity to evolve in response to the dynamic characteristics of crises, resulting in the establishment of more formalized and specialized approaches. Currently, the management of crises encompasses not only governmental entities but also international organizations such as the European Union Emergency Response Coordination Centre (ERCC), which is the heart of the EU Civil Protection Mechanism, non-governmental organizations, and the business sector, fostering collaborative efforts to alleviate, address, and recover from diverse forms of emergencies (Drennan et al., Reference Drennan, McConnell and Stark2015). This is not surprising in an era where the state has essentially declared its inability to keep citizens safe (Evans and Reid, Reference Evans and Reid2014) and calls upon the entire society to contribute to its own resilience and safety.

An important threshold question in this literature concerns the extent to which a crisis is endogenous – that is, an organizational crisis brought on by failures, shortcomings, and malevolent behavior in a particular organization – or whether it is an exogenous shock to a political or governance system that should be controlled, but that is not credibly caused by the entity seeking to manage the crisis. As Seeger et al. (Reference Seeger, Sellnow and Ulmer1998, p. 233) argue, “organizational crises are conceptually distinct from disasters, which are usually defined in the research literature as non-organizationally based events generated by natural or mass technological forces”.

This distinction may be naive. Organizations are normatively expected to be prepared to protect people from a dizzying array of financial, technological, and environmental risks (Beck, Reference Beck1992; Birkland and Nath, Reference Birkland and Nath2000; Perrow, Reference Perrow1999). The legitimacy of the responsible organizations or the social, political, and economic system of which these organizations are a part may be eroded if these organizations fail to address a crisis effectively, particularly if the crisis is either caused, somehow, by poor decisions made by an organization or whether, once confronted with a crisis they did not cause, they fail to respond effectively or fail to even imagine that the event could happen. For example, a food poisoning crisis may implicate the supplier of food to a restaurant, the restaurant that prepared the food, and the government organization that regulates food safety. In such a case, while the restaurant may not have caused the food to be tainted, they should have anticipated the possibility that food could be unsanitary and should have systems in place to prevent their serving contaminated food, even though the fact that the food was tainted was the fault of a supplier (Birkland and Nath, Reference Birkland and Nath2000). The September 11 terror attacks in the United States were a crisis because of the physical harm they did, the increased sensitivity to the possibility of mass-casualty terrorism, and the Bush Administration’s failure to detect and thwart the plot in the face of a growing realization that terrorism poses a threat to the United States. In the contemporary policy environment characterized by networked governance and in a complex risk environment, efforts to claim that a crisis event is unforeseeable and uncontrollable for which an organization cannot be expected to prepare is, at best, unrealistic. As Darling (Reference Darling1994) notes, successful firms anticipate potential crises and seek to manage risks before they develop into crises. This is true of any organization, including the public sector.

The crisis management literature also implicitly recognizes that labeling an event a crisis is a political act (Boin et al., Reference Boin, ’t Hart, Stern and Sundelius2017; Edelman, Reference Edelman1977), as contestants in the policy process engage in framing contests in an effort to elevate or block issues from reaching the agenda. Participants may claim that a current situation – such as the large number of migrants at the Mexico-US border – is a current crisis, or more forward-thinking actors anticipate policy problems before they arise (DeLeo, Reference DeLeo2016).

1.2.3 Crises as Focusing Events

Crises can be characterized as focusing events or as having been triggered by such an event. A potential focusing event is an event that is “sudden, relatively rare, can be reasonably defined as harmful or revealing the possibility of potentially greater future harms, inflicts harms or suggests potential harms that are or could be concentrated on a definable geographical area or community of interest, and that is known to policymakers and the public virtually simultaneously” (Birkland, Reference Birkland1997, p. 22). We cannot know in advance whether a particular event will be a focusing event; that is, we do not know that the event will have a substantial effect on the political agenda. The logic of MSF suggests that a focusing event is an event that opens up the agenda to new contestants in policy debates, and those contestants – often as policy entrepreneurs – find that the event is an opportunity to couple solutions with problems. Kingdon (Reference Kingdon2011) identified focusing events as one way in which issues gain greater attention, along with changes in indicators and spillovers from other domains, though he never provided a definition of focusing events as narrowly focused as Birkland. But regardless of how we conceive of focusing events, they typically reach the agenda because they reveal policy failures and may, therefore, trigger policy learning (Birkland, Reference Birkland2006). Disasters create learning opportunities because they create “cognitive openness” to change (Stern, Reference Stern1997) on the part of actors in the policy subsystem. As Birkland argued, “focusing events can be used to demonstrate the existence of policy failure; that is, participants in policymaking can reasonably argue that a focusing event would not have happened or would not have been so severe ‘if only’ something had been done” (Birkland, Reference Birkland2004, p. 343).

In MSF, the interaction between the policy entrepreneur and crises is articulated in detail. By coining the term focusing event, Kingdon (Reference Kingdon2011) pointed to the window of opportunity for policy entrepreneurs to link policy prescriptions to the new appreciation of a problem that is revealed by a focusing event. On the other post-event side, the conceptualization has had implications for policy learning. To simplify the conceptualization, we can assume that individuals are the agents of learning. Busenberg (Reference Busenberg2001, p. 173) defines learning as “a process in which individuals apply new information and ideas to policy decisions.” The information need not be new to the policy domain or subsystem – indeed, a great deal might already be known in the problem stream – but the problems revealed by the event may be new to the most important decision-makers within a policy community. Peter May argues specifically that policy failure can yield three different kinds of learning: instrumental policy learning, social policy learning, and political learning. In all three types of learning, policy failure – politically and socially defined – provides a stimulus for learning about how to make better policy (May, Reference May1992) and may be used by a policy entrepreneur in their quest for change.

Instrumental policy learning concerns learning about the “viability of policy interventions or implementation designs.” This sort of learning centers on policy tools, such as the use of subsidies, sanctions, or incentives. If an event reveals that something about the implementation of an otherwise sound policy has failed, instrumental learning will yield improved policies. For example, after the September 11 attacks in the United States, legislation that removed passenger screening at airports from the private sector to the federal government is a change in the policy tool intended to prevent hijackings. Conversely, social policy learning goes beyond changes to policy instruments and seeks to better understand the problem itself. It can result in a better understanding of the causes of public problems, which can yield “policy redefinition entailing changes in policy goals or scope – e.g., policy direction, target groups, rights bestowed by the policy” (May, Reference May1992, p. 336). These first two types of learning are not likely to lead to innovative policies because they emerge as a result of incremental change.

Finally, a third type of learning is political learning, which is learning about “strategy for advocating a given policy idea or problem,” potentially leading to “more sophisticated advocacy of a policy idea or problem.” In simplest terms, political learning is manifest when advocates for particular policy prescriptions learn how to make better arguments for adopting those policies. Policy entrepreneurs can be such advocates as they seek to realize dynamic change.

2.2.1 Crises from a Policy Perspective

When does a crisis become a crisis? By 2023, the scale of opioid addiction and the casualties and misery it has caused in the US had been declared a crisis. However, for the preceding three decades, physicians had prescribed such medications in large quantities with seeming abandon, while addiction was framed as an individual character flaw rather than the manifestation of institutional failure that is currently recognized to be. Though crises, from a societal perspective, often refer to disastrous events, such as earthquakes, hurricanes, or wildfires, most crises are more or less socially constructed, some of them completely so (Boin et al., Reference Boin, ’t Hart, Stern and Sundelius2017). McConnell (Reference McConnell2003, p. 393) goes as far as to say that “[w]hat constitutes a crisis is a matter of judgement” rather than a matter of fact, and that crisis is a word ascribed to a particular set of circumstances that are outside the scope of the ordinary. This is in line with ‘t Hart’s (Reference t Hart1993) early treatment of the term, defined as “a breakdown of familiar symbolic framework legitimizing the pre-existing sociopolitical order” (p. 39), a definition that leans on an understanding of crisis as a situation of fundamental ambiguity, inducing collective stress and create space for the possibility of de-legitimizing existing ways of doing things (‘t Hart, Reference t Hart1993). We adopt a critical realist approach acknowledging that objective facts constitute a component of crises, but the delineation of these events in time and space as crisis events is based on perceptions and sensemaking in the sense that the consequences of these events and the narratives surrounding them include interpretation (Stone, Reference Stone2002) and to a certain extent are socially constructed (Drennan et al., Reference Drennan, McConnell and Stark2015).

The dual nature of crises rests on the fact that they constitute both opportunities for policy change or organizational reform and real events with adverse consequences for the livelihoods of the people that they affect. This duality is reflected in two perspectives of crisis management, the political and the managerial (Rosenthal and Kouzmin, Reference Rosenthal and Kouzmin1997; ‘t Hart and Boin, Reference t Hart, Boin, Rosenthal, Boin and Comfort2001). Thus, studies of crisis management, in essence, may be conducted at two levels of analysis: the strategic/political level and the operational/managerial level. The former refers to the “overall direction of crisis responses and the political process surrounding these responses” (Boin et al., Reference Boin, ’t Hart, Stern and Sundelius2017, p. 12), which entails a lot of the social construction of crises mentioned above, while the latter comprises the actions of people “who directly experience and respond to a critical contingency (Boin et al., Reference Boin, ’t Hart and McConnell2008, p. 7). This Element views crises and their management from a policy angle; this necessarily involves both perspectives. More specifically, the strategic/political perspective corresponds to the decision-making process that results in the adoption of a crisis response, whereas the operational/managerial perspective corresponds to policy implementation. We argue that the partial primacy of perceptions revolving around a real event opens up space for policy entrepreneurial action in both perspectives. And yet, the theorization of policy entrepreneurs in crises, crisis management, and risk management tends to be fuzzy and lacking specificity.

Concomitantly, crises are complex events with varying spatial, temporal, and ontological characteristics, even as they unfold at the strategic/political and operational/managerial levels. To add nuance and richness to our analysis, we consider two types of crises: fast-burning and creeping crises (‘t Hart and Boin, Reference t Hart, Boin, Rosenthal, Boin and Comfort2001) based on the rate of development and the pace of termination of crisis events. In binary terms, crises are differentiated based on whether they are fast-burning or slow-burning. Fast-burning crises constitute what comes to the mind of most people when they think of a crisis – a clearly defined event with sharp edges, such as the 9/11 terrorist attacks (Boin et al., Reference Boin, Ekengren and Rhinard2020). This type of crisis has been explored in the crisis management literature, but also in public administration (see, e.g., Stillman, Reference Stillman2005). Boin and colleagues argue that slow-burning crises (reconceptualized as creeping crises (see Boin et al., Reference Boin, Ekengren and Rhinard2020), though under-researched, constitute the type of crisis most often occurring in contemporary societies. Creeping crises have both spatial and temporal dimensions, with ambiguous causes developing over time as a result of transboundary processes, most often overlooked by decision-makers until it is too costly for them to turn a blind eye to them. The authors go on to say that climate change, the COVID-19 pandemic, and the increased refugee flows are some salient examples of creeping crises.

Choosing this broad binary typology of crises lends analytical leverage to the case studies examined in this Element. From a theoretical perspective, inserting policy entrepreneurs in fast-burning crises can provide nuance in the existing literature by yielding insights as to when and how much entrepreneurs are involved in the development of a crisis, what drives them, and what strategies they use. On the other hand, explicitly involving policy entrepreneurs in creeping crises has the potential to increase our understanding of the management of the most relevant crises of our time and the public policies that are introduced (or not introduced) as a result.

Theories of the policy process do not differentiate between extraordinary and normal policymaking. Indicatively, crises are conceptualized as focusing events in the MSF and external shocks in the ACF. Conceptualized this way, crises become a moving part of the theory rather than altering the policymaking process altogether. A recent paper posited that policy theories do not adequately account for the complexity, urgency, ambiguity, and uncertainty characterizing crises. Drawing from theories of the policy process and theories of policy output, Wenzelburger et al. (Reference Wenzelburger, König and Wolf2019) constructed a framework that models agenda-setting and decision-making during crises, in which policy entrepreneurs feature prominently, though perhaps not differently than in normal policymaking processes. Policy entrepreneurs are crucial in the agenda-setting stage because of their ability to sway public opinion and provide competing frames for the unfolding crisis. Equally, policy entrepreneurs are key to convincing policymakers that their preferred solution is the most suitable one. If the issue behind the crisis is hyper-politicized, then the process leads to dynamic change; if not, the result is incremental policy change (Wenzelburger et al., Reference Wenzelburger, König and Wolf2019).

2.2.2 Policy Actors Involved in Crisis Management

The examples above notwithstanding, theories of the policy process are very good at explaining the role of institutions in the policy process but less good at specifying agency in a consistent manner. They are also relatively silent when it comes to specifying the actions of leaders. Capano and Galanti (Reference Capano and Galanti2018) attribute this deficit to the conflation of leaders and entrepreneurs; indeed, entrepreneurs possess leadership qualities that enable them to build coalitions and provide ideational leadership.

In order to anchor our discussion on policy entrepreneurs from a policy perspective in a crisis context, we must first tease out the differences between them and leaders, who have so far shown to be the most salient actors in the politics of crisis management. Isolating policy entrepreneurs in the landscape of actors involved in the policy process is not always a straightforward exercise. Theories of the policy process, having (mostly) developed independently of each other (see Nowlin, Reference Nowlin2011; Petridou, Reference Petridou2014), treat actors inconsistently, comparatively speaking, in terms of focus and substance. For example, brokers are key actors in the ACF; they mitigate conflict and promote policy learning within and across coalitions (Ingold, Reference Ingold2011). In contrast, they are much less developed in the MSF, where most agent actions are performed by policy entrepreneurs in the interest of coupling the policy, politics, and problem streams during windows of opportunity. These actions include brokerage and leadership, rendering brokers and leaders opaque in MSF, which is the one theory of the policy process that most integrates the notion of crises through the concept of focusing events. Conversely, agency in the NPF is conceptualized through policy narrators and policy marketers. Hand et al. (Reference Hand, Morris and Rai2023) and Petridou and Mintrom (Reference Petridou and Mintrom2021) have imagined the policy entrepreneur as the narrator and the policy marketer, respectively.

Entrepreneurship has been defined as “a particular form of leadership” with an emphasis on problem-solving and the synthesis of already existing processes in forming an innovative and effective solution (Ostrom, Reference Ostrom2005, p. 1). Entrepreneurs are motivated by benefits, including delivering improved services to their communities and sharing the burden in return for increased benefits, the excitement of innovation, and receiving monetary income as well as respect from their positions as public officials (Mintrom in Ostrom, Reference Ostrom2005).

Similarly, Capano and Howlett (Reference Capano, Howlett, Capano and Howlett2009) consider policy entrepreneurship to be a component of leadership. As Petridou et al. (Reference Petridou, Narbutaité Aflaki, Miles, Narbutaité Aflaki, Petridou and Miles2015) have pointed out, there have been few conceptual distinctions between the entrepreneur and the leader in the literature (though see Capano and Galanti, Reference Capano and Galanti2018, Reference Capano and Galanti2021). Having said this, Timmermans et al. (Reference Timmermans, van der Heiden and Born2014) add granularity to entrepreneurship by suggesting that policy entrepreneurs score high in transformational leadership as opposed to transactional leadership. This implies that when it comes to leadership skills, policy entrepreneurs exhibit leadership behaviors described as “visionary, change-oriented and non-conservative” (p. 102). This type of leader, accordingly, is generally more effective under socially stressful situations and crises (Petridou, Reference Petridou2017).

In summary, the utility and function of policy entrepreneurs are fairly well developed (if inconsistently) in theories of the policy process, while the specificity of leaders is less developed in these theories, though theorized considerably in the crisis management literature. Policy entrepreneurs have leadership qualities, but at a fundamental level, leaders govern, and policy entrepreneurs promote policy innovation and change (Capano and Galanti, Reference Capano and Galanti2018; see also Petridou, Reference Petridou2017). In the section that follows, we elaborate further on policy entrepreneurs.

3.2.1 How Do Crisis Challenges Affect Policy Entrepreneurial Strategies?

The September 11 attacks may be an unusual case for policy entrepreneurs because the attacks seemed to strike so suddenly and with such grave effect that it became very clear that something would need to change after the attacks. The ability of well-placed, opportunity-seeking policy entrepreneurs to influence the course of policymaking after the attacks was also due to the fact that policy entrepreneurs in the homeland security regime address highly technical “policies without publics” (May, Reference May1991), which offer limited opportunities for mass mobilization. Instead of inciting widespread action, as seen in environmental disasters, these domains rely on the internal mobilization of individuals with exceptional expertise or legal authority. Baumgartner and Jones (Reference Baumgartner and Jones2009) term this “Downsian” mobilization, involving actors within the policy regime responding to the “alarmed discovery” of the crisis (Downs, Reference Downs1972), rather than a “Schattschneider mobilization” (Schattschneider, Reference Schattschneider1975) that mobilizes typically unseen actors to seek access to decision-making venues (Pralle, Reference Pralle2003). This is not to say that there was no popular support for taking some sort of action after the terrorist attacks; rather, this means that the substance of the policies to be enacted after the attacks would be shaped by policy entrepreneurs.

According to MSF theorization, policy entrepreneurs draw on existing ideas in the policy stream to match to new appreciations of problems in the problem stream. This implies, as Kingdon posited, that there are seldom entirely novel ideas in these policy domains activated during a crisis. Instead, policy entrepreneurs adapt existing ideas to the newfound awareness of problems. Utilizing preexisting ideas is often a more effective strategy than attempting innovation, as policy windows, even those as momentous as the September 11 attacks, remain open for a brief period. Additionally, focusing events often stimulate attention across multiple policy issues or domains (Lawrence and Birkland, Reference Birkland2004). Despite the initial surge of national unity and determination post-attacks, the window to capitalize on this sentiment was brief. The institutional structures that impede change or significantly alter its course did not undergo fundamental shifts after September 11.

Two of the central policy entrepreneurs in the aftermath of the September 11 case were Attorney General John Ashcroft and Senator Joseph Lieberman of Connecticut. Of the main strategies that are key to policy entrepreneurs’ success, the most important was in framing the problem. Policy entrepreneurs pressing for rapid change did not frame the September 11 attacks as crimes, even though federal law defines terrorism and hijacking as serious crimes. Rather, these attacks were framed as threats to the national security of the United States. National security threats are usually met by bold measures to respond to the threats and to prevent a recurrence of an attack. Furthermore, Senator Lieberman and his supporters framed the September 11 attacks as a policy failure that could only be met with creating an agency like the Department of Homeland Security, in the most sweeping change to the organization of American government since the late 1940s.

Policy entrepreneurs were also able to work with advocacy coalitions, although the role of such coalitions was perhaps not as great as it might be in other domains. So much of the post-9/11 reforms were driven by processes internal to government that the policy entrepreneurs were inside operators, seeking to balance the forces inside government that would have a role in the post-attack world. This means that, while the September 11 attacks were framed as a policy failure, the agencies most implicated in this failure – the FBI and the CIA, in particular – were not made a part of the new Department of Homeland Security, while agencies such as the Border Patrol or the Federal Emergency Management Agency were swept up into DHS. This is likely due to the much stronger support for administrative autonomy enjoyed by the FBI and CIA compared with other agencies. Policy entrepreneurs therefore had to pick their battles and were willing to accept partial success in the reorganization of government.

In the case of John Ashcroft, his concern was less with the organization of government than it was with the creation of legal authorities that would grant the federal government greater power to pursue terror suspects and their organizations. But these efforts were not novel and were clearly an example of joining new appreciations of problems with existing solutions – the so called “wish list” of criminal justice reforms. Attorney General Ashcroft was thereby able to frame the terrorist attacks in a way that would accomplish long-standing goals that would, to be sure, be useful in the pursuit of terrorists but would also, as it turned out, be even more useful in the more typical sorts of crimes prosecuted at the federal level, such as drug offenses.

3.2.2 How Do Policy Entrepreneurs Shape Policy Outcomes?

Policy responses after the September 11 attacks unfolded in three phases. Initially, measures were taken to address revealed policy deficiencies in legal authorities and aviation security. The second phase involved a debate on government organization in the wake of the attacks, culminating in the establishment of the DHS. The third phase saw the formation of congressional committees and the 9/11 Commission, which issued reports to identify and rectify the lapses that allowed the events to occur, leading to intelligence reforms in the United States.

Attorney General John Ashcroft played a prominent early role as a policy entrepreneur, advocating for substantial changes to law enforcement through the enactment of the “USA Patriot Act” on October 26, 2001. Some viewed this as the Bush Administration’s “wish list” for criminal justice and intelligence reform, made possible by leveraging the attacks as a catalyst for change. Ashcroft and supporters, such as Senator Orrin Hatch, contended that the Patriot Act was not merely a wish list but was the realization of ideas previously introduced in Congress, yet stalled for various reasons. Subsequent experience demonstrated that the Patriot Act’s provisions were more effective in prosecuting drugs and other crimes than in terrorism detection and prevention (Doherty, Reference Doherty2021). Nevertheless, it serves as a classic example of policy entrepreneurs adapting existing ideas to address new problems within a window of opportunity.

The Aviation and Transportation Security Act (ATSA) was adopted on November 19, 2001, and tackled known deficiencies in the passenger screening system (Birkland, Reference Birkland2004). The stakeholders included airlines responsible for contracting and paying private firms for passenger screening, the Federal Aviation Administration overseeing security, airport operators, and concerned members of Congress. Disagreements over enhanced security measures and who should bear the costs had persisted for years before the attacks. The events of September 11th removed obstacles to federalizing the screening function, leading to the creation of the Transportation Security Administration. The primary opportunity for policy entrepreneurship lies not in shaping the federal role in passenger screening but in promoting various technology solutions like passenger and baggage scanners. Technology companies and defense contractors advocated for the adoption of their products in congressional hearings across different aspects of homeland security (Tierney, Reference Tierney2005).

The Homeland Security Act, a comprehensive overhaul of federal security functions, did not pass as swiftly as the Patriot Act and ATSA due to differing participants in the debate. The immediate push for a cabinet-level Homeland Security department lessened with Tom Ridge’s appointment as director of the White House Office of Homeland Security. Initially, the Bush administration resisted creating a cabinet-level DHS, uncertain of the benefits of merging diverse agencies and seeking to avoid Congressional oversight pressure.

However, there was significant Congressional interest in establishing a DHS for organizational purposes, providing the authority and resources necessary for effective coordination. Senator Joseph Lieberman, as chair of the Senate Committee on Governmental Affairs (later the Committee on Homeland Security and Governmental Affairs), emerged as a leader in this endeavor. Political pressure led the Bush administration to signal its intention to create the DHS legislatively by June 2002. Notably, the FBI and CIA, the agencies associated with the September 11 plot, were not included in the DHS. Their bureaucratic autonomy and Congressional influence prevented their integration or substantial mission alteration.

In the rush to act after the attacks, policymakers acted before a formal investigation commission was appointed (Rubin et al., Reference Rubin, Cumming, Tanali and Birkland2003). This lack of a comprehensive review, akin to the one following the attack on Pearl Harbor in 1941, was recognized as a significant gap in understanding the attack. Although initially opposed by the Bush Administration, mounting pressure led to its acquiescence to creating an investigating commission (Birkland, Reference Birkland2009). The 9/11 Commission, officially known as the National Commission on Terrorist Attacks upon the United States, was established to probe events leading up to and following the attacks. It addressed a broad spectrum of security challenges facing the nation. It anticipated policies promoted after September 11, including the call for a “National Homeland Security Agency” and significant White House reorganization to tackle new-century strategic challenges under a previous commission charged with studying terrorism known as the Hart-Rudman commission, which called for the creation of a homeland security agency before the 9/11 attacks (Ellis, Reference Ellis2004). Co-chairs Senators Gary Hart and Warren Rudman publicly discussed their work during and after the 9/11 commission’s tenure. It is noteworthy that former Representative Lee Hamilton, a member of the Hart-Rudman Commission, later co-chaired the 9/11 Commission.

Compared to the Hart-Rudman and Gilmore commissions preceding it, the 9/11 Commission very actively sought to promote its work. Its meetings were prominently publicized, addressing a wide spectrum of issues from emergency responses to intelligence lapses. Its scrutiny of intelligence failures significantly shaped the establishment of the Office of the Director of National Intelligence (ODNI), headed by the Director of National Intelligence. This role was meant to centralize intelligence information from various civilian and military sources, though its full potential has not been realized.

After the 9/11 Commission’s official mandate ended in August 2004, the members of the commission launched the Public Discourse Project; the commissioners served as its board. Its main aim was to publicly advocate for the implementation of its recommendations. Operating for about two years on private donations, they held discussions on “The Unfinished Agenda” in the summer of 2005. They also issued “report cards” on the implementation of recommendations. Members testified before Congress, gaining more visibility than usual unique commitment to ensuring their report was broadly considered sets them apart as a significant example of policy entrepreneurship.

Think tanks were also prominent in the post-attack debate. The RAND corporation, in particular, had close ties to the defense and security establishment and conducted research and provided policy recommendations on issues related to national security, counterterrorism, and foreign policy. Numerous academics appeared in congressional hearings and other forums, but their influence was not widespread and was variable, so relatively few academics stood out as policy entrepreneurs.

The victims of the September 11 attacks held significant moral authority. An external interest group, the Family Steering Committee, gained attention, co-founded by the daughter of a passenger on one of the hijacked planes. This group pressured the Bush Administration to establish the 9/11 Commission, which was done with reluctance. They also lobbied Speaker of the House Dennis Hastert to extend the commission’s final report deadline (Jost, 2011). While the victims’ groups had limited influence, they strategically applied pressure to address the failures that led to the 9/11 attacks. In this way, they occasionally served as policy entrepreneurs.

Finally, traditional interest groups were relatively absent from the debates that followed the September 11 attacks, and when they were active, they were generally on the defensive. Civil liberties groups were particularly troubled by the changes to the criminal and domestic surveillance laws. But in the early days after the September 11 attacks, their concerns were often downplayed, although once the passions of the immediate aftermath of the attacks cooled, their voices became more influential.

3.3.1 Sri Lanka

The tragic consequences of the tsunami in Sri Lanka unfolded in the shadow of the civil war, which has been going on, with the exception of brief periods of ceasefire, since the early 1980s. This is important because the war has deepened existing cleavages along ethnic lines and has revealed new socioeconomic asymmetries among populations, resulting in the creation of new vulnerabilities (Blaikie, Reference Blaikie2009). In addition to the large-scale loss of life and environmental degradation that Sri Lanka suffered, the tsunami touched many aspects of society. Most affected sectors include the social sector, specifically housing and private sector assets, as well as sectors that directly affect the prosperity of coastal communities, including fisheries, agriculture, tourism, and transportation (Birkmann et al., Reference Birkmann, Buckle and Jaeger2010). Following the tsunami, international efforts focused on establishing an integrated early warning system for coastal areas in the Indian Ocean. The Indian Ocean Tsunami Warning and Mitigation System (IOTWMS) is an international institution that became fully operational in 2013. The twenty-eight member state group aims to promote communication, exchange of information, and tsunami research, monitor seismic and other activity that may result in a tsunami, and ensure that a tsunami early warning system is always up and running (UNESCO, 2019).

The disaster attracted international attention and concomitant reconstruction funds to the area. Birkmann et al. (Reference Birkmann, Buckle and Jaeger2010) estimate that Sri Lanka received approximately USD4b in aid, though the reconstruction was hampered by governance deficits, including centralized decision-making, an absence of a set of standard operating procedures for all stakeholders, lack of capacity coordination, lack of know-how and lack of action, lack of trust, negligence and indifference, lack of feedback from training and simulation exercises, high turnout of appointed officers and over politicization on transfers and appointments of public servants (Sakalasuriya et al., Reference Sakalasuriya, Haigh and Hettige2020).

Three formal policy changes occurred against this broad governance background: the establishment of a buffer zone and resettlement; the development of tsunami early warning systems, and the creation of new bureaucratic organizations – the Task Force for Rescue and Relief, the Task Force for Law and Order and Logistics, and the Task Force to Rebuild the Nation. Implementation of the establishment of the buffer zone and the relocation of people at risk was less than straightforward: conflict, lack of transparency, and unequal treatment of populations, even as relocation offered relief from chronic poverty for some families, stymied the process. The governance deficits identified by Sakalasuriya et al. (Reference Sakalasuriya, Haigh and Hettige2020) led to difficulties in the implementation of the remaining two formal solutions. Here we follow Meydani (Reference Meydani, Narbutaité Aflaki, Petridou and Miles2015), who argued that a low level of governability (or non-governability) and a low level of liberal political culture among citizens result in the primacy of short-term considerations among possible policy entrepreneurs in government as well as politicians as the latter seek to maximize their own political capital. “Non-governability” refers to the lack of institutional and administrative capacity to make consistent and stable public policy (Meydani, Reference Meydani, Narbutaité Aflaki, Petridou and Miles2015) resulting from a lack of political legitimacy and a concomitant lack of inclusion of societal stakeholders in the decision-making process. The lack of inclusion of societal stakeholders, coupled with a political culture that is outcome-oriented, leads to calls for short-term solutions (Meydani, Reference Meydani, Narbutaité Aflaki, Petridou and Miles2015). The formal changes in Sri Lanka were long-term, and though enacted into public policy partly due to the influx of foreign aid, their implementation has been largely ineffective. This is mainly due to the non-governmentality characterizing the Sri Lankan political context.

Birkmann et al. (Reference Birkmann, Buckle and Jaeger2010) outline a number of informal changes as a result of the tsunami disaster. We elaborate on an informal change initiated by a policy entrepreneur focused on the publication of a new set of guidelines for reconstruction (Dias et al., Reference Dias, Dissanayake and Chandratilake2006). Ranjith Dissanayake,Footnote ¹ a civil engineer, faculty member at the University of Peradeniya, market entrepreneur, and later high-level civil servant, has had a long-standing interest in disaster risk reduction efforts in Sri Lanka, especially as they relate to the built environment. A survey of the damaged structures revealed that changes in construction guidelines were necessary to build more robust structures that would be less vulnerable to tsunami events. Houses built on stilts and religious structures, for example, performed better because they were built adhering to traditional practices and, in the case of religious structures, stronger. Additionally, Dissanayake pushed for the consistent and uniform training of construction workers, who had been generally treated as unskilled labor and were characterized by high turnover.

3.3.2 Sweden

The tsunami presented an unprecedented challenge for the Swedish authorities, which lacked any kind of coordinating office for serious crises at that time. The unusual autonomy of Swedish public agencies, rooted in its constitution, partly accounts for the lack of such a coordinating body. More specifically, Sweden’s government – the ministries and the Prime Minister’s office comprising the Government Offices – is small. The day-to-day provision of public services and, to a certain extent, the formation of policy beyond broad governmental guidelines are carried out by a large number of autonomous agencies. What is more, Sweden is characterized by the absence of ministerial rule. This means that even though public agencies are placed under a ministry, these agencies and the public servants who work for them have considerable leeway (Larsson and Bäck, Reference Larsson and Bäck2008; Petridou, Reference Petridou2020). This autonomy extends horizontally as well: the Swedish administrative system does not allow for one agency to have operative or strategic power over another. In a 2003 essay, the then Director General of the Swedish Emergency Management Agency explored the question of who is in charge during crises in peacetime in Sweden. She concluded that one coordinating agency would be unconstitutional and instead focus on the need for coordination at the operative level (Eksborg, Reference Eksborg2003). This means that coordination, especially during crisis events, is an activity that must be fostered continuously with network-like mechanisms across departments and agencies.

Such coordination and the ability to upscale proved easier said than done in December 2004. Since this tsunami crisis unfolded abroad, managing it was in the purview of the local consulate offices. However, these offices were not equipped to deal with the enormity of the disaster and the sheer number of people in need of assistance. The Ministry for Foreign Affairs’ response was slow and underwhelming, resulting in thousands of Swedish citizens being stranded without any support from their government or any way to reach the Swedish authorities by phone or otherwise. The fact that the tsunami took place on a public holiday in Sweden when civil servants were off work certainly did not help matters. The failure of the Swedish government to take care of its citizens in their time of need resulted in a decline in the trust citizens place in their leaders, normally quite high in Nordic countries. Characteristically for the Swedish policymaking process, a commission of inquiry was set up in January 2005, named the “Disaster Commission.” Its main assignment was to assess the efforts of the Swedish national public agencies to assist the Swedish citizens in Southeast Asia affected by the tsunami. The Disaster Commission found that (i) the Government Offices lacked a coordination structure to handle serious crises, (ii) the preparedness of disaster medicine was insufficient, (iii) the consular management was slow in relation to the unfolding of the crisis, and that (iv) cross-country collaboration could have been better (SOU, 2005).

The Social Democratic coalition government in power at the time had been unwilling to establish a crisis coordination office in the Government Offices in the proximity of the Prime Minister’s office. The solution offered after the findings of the disaster commission was a token coordination office called the “Unit for Preparedness and Analysis” without any real power. The office had no access to the political leadership and was not part of any security briefing – it was seen as a purely symbolic and “empty” solution (Petridou and Sparf, Reference Petridou and Sparf2017).

The Social Democratic coalition government lost the subsequent election in 2006 to the center-right Alliance, which ran on a platform of safety and security issues. One of the first things the new government did was establish a Crisis Secretariat under the Prime Minister’s office after the completion of yet another commission of inquiry (Petridou and Sparf, Reference Petridou and Sparf2017). Given the preceding discussion, the politics of establishing any form of coordinating entity was precarious, to say the least. The new coordinating office was established with the full support of the newly elected government. However, in order for the Crisis Secretariat to be effective, it needed to be adept at managing the existing power differentials in place inside the Government Offices. In other words, for the Crisis Secretariat to coordinate, it needed access to information that the ministries in the Government Offices had and ensure that they (the ministries) were willing to share it.

Public health in the United States is a shared responsibility between the federal, state, and local governments (DeLeo, Reference DeLeo2010; Gostin et al., Reference Gostin, Friedman and Wetter2020). This means that public health policy in the United States is largely fragmented and is highly variable from state to state and even at the local level (Kettl, Reference Kettl2020; Stivers, Reference Stivers2022). Public health is a low salience issue for most people at most times because most of the time, the public health system works to prevent major disease outbreaks. The federal government’s role is to support state and local efforts to promote public health, particularly in situations where threats to public health are national in scope. The COVID pandemic clearly fit this standard. In this way, the federal government’s role is similar to its role in response to natural disasters, where the federal government supports state and local actions to mitigate to and respond to disasters, particularly when these events outstrip state and local capabilities (Kapucu and Hu, Reference Kapucu and Hu2022; Parker and Stern, Reference Parker and Stern2022).

When the first news of a novel virus originating in China became widely known, an existing public health system was mobilized at the international, national, and subnational level, which all had plans in place to handle such an event. Had the COVID pandemic been managed roughly according to the plans laid out for a pandemic, the CDC would have effectively tracked the spread of the disease across the world and into the United States, it would have developed effective tests and treatment protocols, and the federal government would have coordinated efforts to purchase and distribute supplies and equipment to hard-hit areas based on needs (Kapucu and Hu, Reference Kapucu and Hu2022). These plans developed from experiences such as the SARS outbreak in 2002-2004 and when the H5N1 (avian) influenza became an issue of concern in the early 2002s (DeLeo, Reference DeLeo2010, Reference DeLeo2016).

During the transition between the Obama and Trump administrations, the Obama administration developed a “playbook” for pandemic response that it shared with the incoming Trump administration officials. The new administration largely ignored this plan (Parker and Stern, Reference Parker and Stern2022), which meant that the administration largely treated the pandemic with ad hoc efforts that did little to support local response. Indeed, the mixed messages and bungled response meant that the federal government was remarkably unprepared to help, even after having learned some important lessons about disease and pandemic preparedness with SARS and MERS in the early 2000s. Instead of drawing on this experience, the Trump administration falsely claimed that the pandemic was so unprecedented that it could not have been effectively anticipated (Parker and Stern, Reference Parker and Stern2022). The COVID pandemic was indeed widespread, but not so much so that the plans made in the Bush and Obama administrations would have been useless.

COVID clearly outstripped local public health efforts to track the spread of the COVID virus and to “flatten the curve,” the term used to describe the pressing need to reduce the number of cases before they overwhelmed hospitals and clinics. Along these lines, the federal government was ineffective in helping states secure needed supplies and equipment, such as testing materials and ventilators for patients on intensive care. This meant that, to an extent greater than had been contemplated in pre-pandemic planning, the federal government was particularly poorly prepared to support state and local government.

To a considerable extent, this meant that state and local governments were either left to their own devices (Parker and Stern, Reference Parker and Stern2022), competing with the federal government and with each other to get needed resources (Birkland et al., Reference Birkland, Taylor, Crow and DeLeo2021; Kettl, Reference Kettl2020) or had considerable latitude for experimentation and improvisation (Congleton, Reference Congleton2023), depending on one’s perspective. Adding to this complexity is the high degree of political polarization in the United States and its challenges to federalism (Birkland et al., Reference Birkland, Taylor, Crow and DeLeo2021; Konisky and Nolette, Reference Konisky and Nolette2021).

However, it is important not to overstate the extent to which partisanship and polarization influenced early responses to the pandemic. State governors were compelled to take action in the pandemic because the federal government under Donald Trump did not assert the sort of effective leadership that would be expected in a pandemic. Furthermore, governors were foremost actors in their states because legislatures were not in session or the legislative process was too slow, and governors had considerable power to issue executive orders (Weissert et al., Reference Weissert, Uttermark, Mackie and Artiles2021). For example, between March 19 (California) and April 7, 2020 (South Carolina), 43 states issued some form of form of stay-at-home order that would include, among other things, school closings, closings of non-essential businesses, and restrictions on public gatherings. Beyond stay-at-home orders, some states issued guidance or mask mandates on mask wearing, and later in the pandemic, states developed policies to promote vaccination. But states also rescinded their pandemic measures at different rates. All in all, residents of the states rated their governor higher on trustworthiness compared with President Trump (Weissert et al., Reference Weissert, Uttermark, Mackie and Artiles2021).

The degree to which states adopted similar policies is remarkable. While the substance of policies was more similar than some had assumed, governors made strategic choices on issues like stay-at-home orders or the reopening of businesses to signal to key constituencies the governor’s policy stances. In this way, governors were largely acting politically and strategically. For example, Governors Jay Inslee of Washington State and Gavin Newsom were early and vocal adopters of measures to mitigate the pandemic based on their claims that their decisions were grounded in science. But at the same time, Florida, led by Governor Ron DeSantis, adopted a number of similar measures as other states to slow the spread of the pandemic, although the governor was much less vocal in his promotion of these policies.

Republican governors, and their supporters, by and large, chose to deemphasize the pandemic and emphasized instead the economic costs of business shutdowns, with the result that, later in the pandemic, states led by Republican governors had higher case rates and lower testing rates than did states with Democratic governors (Neelon et al., Reference Neelon, Mutiso, Mueller, Pearce and Benjamin-Neelon2021). Governors of states where Trump was more popular, such as Kim Reynolds of Iowa or Ron DeSantis of Florida, were particularly unwilling to sound any greater alarm about the pandemic than was Trump.

However, while much has been made of the role of partisanship and polarization in the United States in the disparate responses of the states, once one controls for other factors, such as urban or rural residents or the age of the population, these differences fall away. And there were Republican Governors, such as Michael DeWine of Ohio, Larry Hogan of Maryland, and Charlie Baker of Massachusetts, who were also highly public proponents of science-based policies, such as social distancing and vaccination, to respond to the challenge of the pandemic (Gasulla et al., Reference Gasulla, Bel and Mazaira-Font2023).

An example of strategic innovation on the part of governors was the formation of interstate alliances that fell short of formal “interstate compacts” of the sort permitted under the Constitution. For example, Washington, Oregon, and California formed a collaborative alliance, as did “a coalition of seven mostly Northeastern states; and a Midwestern alliance” (Huq, Reference Huq2020) led by the governors of Illinois, Michigan, Ohio, Wisconsin, Minnesota, Indiana and Kentucky (NBC News Chicago, 2020) These alliances shared ideas about how to reopen their states and how they could ensure sufficient health care capacity to deal with a resurgence.

In another example of strategic decision-making, Governor Ron DeSantis of Florida, clearly harboring presidential ambitions, was particularly aggressive about “protecting” the “freedoms” of Floridians by promoting legislation, that, among other things, sought to “permanently [prohibit] COVID-19 vaccine passports in Florida,” or “permanently [prohibit] ‘COVID-19 masking requirements at businesses.’” and banning hiring or firing decisions based on vaccination status. These prohibitions were largely symbolic, given that there was little chance of these policies becoming law in Florida (Office of Governor Ron DeSantis, 2023). Many of Governor DeSantis’s claims were false and misleading, such as his and his Surgeon General’s claims that mRNA vaccines were ineffective (Nehamas and LaFraniere, Reference Nehamas and LaFraniere2023) or even dangerous (Sarkissian, Reference Sarkissian2023). But Governors like Desantis have learned that these sorts of appeals are attractive to a particular segment of the electorate, particularly for leaders who have ambitions for higher office.

This was also true of three governors with a reputation for stringent policy responses to the pandemic. In California, Washington, and New York, governors adopted strict stay-at-home orders, encouraged school closings, and promoted vaccination. Gov. Gavin Newsom of California relied on harnessing public-private partnerships, working to support community-based organizations in California (Office of Governor Gavin Newsom, 2021). He also controversially tasked Blue Shield, a health insurance company, with the distribution of COVID vaccines, relying more heavily on the private sector than did other states because of known shortcomings in California’s public health systems (Colliver, Reference Colliver2021). Indeed, in retrospective of his performance in the pandemic, Gov. Newsom expressed some doubts about the stringency of his state’s response in the face of criticisms from other national figures, such as Gov. DeSantis (Cadelago, Reference Cadelago2023). Governor Jay Inslee of Washington was among the earliest proponents for strict COVID mitigation strategies, and Washington state mandated vaccines for state employees. Washington’s early action was driven the emergence of the nation’s first COVID case in suburban Seattle. There was little support in his state for restricting the emergency powers of the governor (Bernstein, Reference Bernstein2023), in contrast with significant opposition in states like Michigan, where the state Supreme Court ruled unanimously that the governor’s continued emergency declarations ran afoul of state emergency management law (Van Beek, Reference Van Beek2020).

New York Governor Andrew Cuomo also played a very public role in promoting stringent anti-COVID measures. But being such a public proponent of policies comes with considerable risks if one’s performance is found to fall short of their rhetoric. In New York, evidence emerged that Governor Cuomo, who pressed nursing homes to accept COVID patients when hospitals were full, pressed officials to hide the rates of illness and deaths in nursing homes (Cohrs, Reference Cohrs2021). In covering up the nursing home data, Gov. Cuomo squandered a great deal of political capital and goodwill. He was ultimately forced to resign as governor over other sexual harassment issues, but the coverup of the data combined with his behavior made his position untenable (Ferré-Sadurní and Goodman, Reference Ferré-Sadurní and Goodman2021).

5.1.1  Crisis Policymaking and Policy Entrepreneurs – An Analytical Framework

Much like in normal policymaking, policy entrepreneurs have the potential to emerge during any crisis event. There are plenty of instances, of course, when no entrepreneur is present. In such instances, we would expect the status quo to prevail with no significant changes in public policy. In this section – as in the previous ones – we continue using the analytical distinction of fast-burning and creeping crises. We also differentiate between two kinds of entrepreneurs as outlined in Section 2: proactive policy entrepreneurs, who fit the traditional notion of the opportunity-seeking policy entrepreneur intrinsically driven by focused intentionality, and reactive policy entrepreneurs, who are driven by extrinsic pressures. The result is a two by two analytical framework as visualized in Figure 3. It makes knowledge claims about policymaking based on the two temporally defined crisis types and the two types of policy entrepreneurs described in Section 2.

Figure 3 Crisis policymaking and policy entrepreneurs – an analytical framework

When reactive policy entrepreneurs emerge in a fast-burning crisis event (top left quadrant of Figure 3), it is because this event has created conditions of necessity that act as extrinsic pressures exerted on the entrepreneur, creating the imperative for a satisficing solution within a truncated time period. Political actors may feel they have no choice but to act entrepreneurially (and this feeling of having no choice constitutes necessity) because the perceived impact of the crisis and attendant high levels of uncertainty make action – and through this action providing some solution – imperative. Much like in the market environment, this kind of entrepreneurship is less likely to result in significant change because it is the outcome of a forced choice. We expect the strategy of using and expanding networks to be prominent in the search for a satisficing solution. As shown in the case study of the forest fire in Sweden, the policy entrepreneur tapped his networks for feedback in his search for a solution to a problem he had not considered relevant before – namely, a forest fire of an unprecedented scale.

Conversely, the emergence of a proactive policy entrepreneur in a fast-burning crisis (top right quadrant of Figure 3) is perhaps the textbook case of a focusing event triggering a window of opportunity in MSF terms. These entrepreneurs have a policy solution in mind long before a crisis occurs. They are intrinsically motivated to promote their preferred policy solution and, like their market counterparts, seek and are alert to opportunities. They educate different audiences with material in traditional and social media and are instrumental in softening institutional rigidities. Entrepreneurs are ready to act when an opportunity, such as a fast-burning crisis, presents itself. Two quite different examples from the case studies illustrate this. John Ashcroft and his supporters advocated for the Patriot Act in response to the 9/11 attacks, legislation that its critics claim is an affront to civil liberties. This solution, long advocated by the Bush administration, was presented as one furthering the interest of national security – an opportunity afforded to the policy entrepreneur by the event. Similarly, the Indian Ocean tsunami presented an opportunity to Ranjith Dissanayake, the policy entrepreneur, and civil engineer in Sri Lanka, to mobilize resources for change in building practices. We expect that the most prominent entrepreneurial strategy is problem framing. All entrepreneurs must be adept at problem framing and, in these conditions, perhaps more so when policymakers must be convinced during the acute phase of the crisis during the brief period when the institutional rigidities are less resistant. As with market entrepreneurship driven by opportunity seeking, we expect policy innovation as a result.

Creeping crises are characterized by high levels of ambiguity and a relative lack of urgency. Indeed, Magnus Ekengren, in a 2023 paper, asks why governments seem to be surprised when their countries are hit by the consequences of creeping crises, including the effects of climate change or a pandemic. Despite the abundance of scientific knowledge, the urgency to act is stymied. We argue that during such crises, there exist punctuation points that focus urgency toward action. These instances may constitute conditions of necessity for reactive policy entrepreneurs (bottom left quadrant of Figure 3). Conditions of necessity may translate in practice into top-down political pressure imposed on public servants or elected officials who are called upon to ameliorate the consequences of the creeping crisis. Potential reactive entrepreneurs can be found among the governors and state health officials in the COVID-19 case study in the US outlined in Section 4. The timeframe of a creeping crisis has a longer horizon than that of a fast-burning crisis, and we hypothesize that an important policy entrepreneurial strategy under these conditions is scaling up policy change. Like in the market environment, we expect incremental change as a result of successful policy entrepreneurial action.

Finally, the emergence of proactive policy entrepreneurs in creeping crises constitutes the conditions most similar to normal policymaking (bottom right quadrant of Figure 3). Most challenges of our time, from the climate crisis to deep economic crises and armed conflicts, have been wicked problems spanning over many policy areas, ignored and left to simmer until they erupt. Crisis governance becomes everyday governance, necessitating normal policymaking. This was certainly true during the prolonged economic crisis and period of austerity in Greece, where a salient policy entrepreneurial strategy highlighted coalition building.

5.1.2 Crises in Theories of the Policy Process

Given the discussion above, we argue for the articulated theoretical insertion of a nuanced concept of crisis in theories of the policy process. Given the scope and space limitations of this Element, we focus on the theories that already include policy entrepreneurs and crises in some form, and we propose further avenues for more specificityFootnote ².

5.1.3 A Research Agenda

Preparedness for, the management of, and ultimately the normalization of crises are sociopolitical processes that involve many and diverse actors. They are integral parts of the work actors and institutions in and around government do – or should be doing – to provide societal safety. Crisis management (including preparedness and recovery) includes development and maintaining appropriate technologies, securing political support and agency mandates, managing and curating information, and cultivating broad networks of stakeholders. These actions have traditionally been ascribed to leaders, but as we have shown in this Element, they are performed by policy entrepreneurs as well, though towards a different end; initiating and implementing innovation for the purpose of bringing about change. We propose five avenues of future research: (i) understanding the dynamics of policy entrepreneurship and leadership during crises; (ii) exploring the legitimacy of policy entrepreneurs; (iii) scrutinizing the politicization of policy entrepreneurial action; (iv) examining factors that contribute to the success of policy entrepreneurs, and (v) a focus on epistemological concerns.