Definition of Crisis

In light of the dramatic consequences of the various forms of disruption of critical infrastructure noted above, one can safely argue that by definition, such disruption has the potential to spawn crises.⁴ However, there may be differences in terms of the specifics of the direct and cascading effects. In fact, the adjective critical and the noun crisis are related; one standard meaning of critical refers to being at or on the verge of crisis.⁵

From the perspective of policy decision-making, a crisis may be usefully defined in terms of three subjective criteria perceived by strategic leaders (as well as by those for whom they are responsible): threat, uncertainty, and urgency (cf. Hermann, 1963; Rosenthal et al., 1989; Stern, 2003).⁶ These criteria are not only helpful in distinguishing crises from other types of situations but also provide a means for probing and preparing to act during them. The components of crisis will be introduced from a general perspective in the paragraphs below and then applied to the domain of critical infrastructure disruption in the following section.

First, crises are associated with threats to (and often potential opportunities to promote) the core values cherished by decision makers and/or their constituencies. These include, among others, human life, public health and welfare, democracy, civil liberties and rule of law, political autonomy, economic viability, and public confidence in leaders and institutions. Leaders must also be prepared to cope with conflicts among such values (Farnham, 2000). The threat of terrorism, for example, entails potential conflicts between security considerations and civil liberties, as demonstrated by the post-9/11 debates on the Patriot Act and Guantanamo Bay, and, more recently, regarding electronic surveillance. Corresponding tensions can emerge with regard to potential public safety and health measures associated with other contingencies as well, such as quarantine restrictions in epidemics and mandatory evacuation orders in the face of hazards such as storms, wildfires, and toxic spills.

Second, crises are associated with high degrees of uncertainty regarding the nature of the threat (i.e., the known and unknown unknowns), the contours of an appropriate response, and the possible ramifications of various courses of action. For example, the causes and manner of contagion of sudden acute respiratory syndrome (SARS) were not known during the initial outbreak in 2003 (Kleinman & Watson, 2006). It was difficult for Chinese and Canadian authorities to deal with the public health challenges—and the political, social, and economic consequences—of this disease in the absence of such knowledge (Olsson et al., 2011). Some analysts distinguish between “familiar” and “novel” contingencies when it comes to crises. All else being equal, the more familiar the contingency (and the more it conforms to scenarios used prior to the crisis as a basis for planning, training, and exercising), the more likely it is that crisis managers will face moderate levels of uncertainty and be working in the domain of structured problem-solving. The more unexpected and novel the event, the greater the uncertainty and the more ill-structured the domain in which crisis managers must operate. Coping with such novel contingencies and the cascading shocks associated with them makes the already difficult challenges of crisis sense and decision-making even more demanding (cf. Baubion, 2013).

Third, crises are associated with a sense of urgency. Events are perceived as moving quickly, and there are fleeting windows of opportunity to influence their course. Effective, proactive intervention can minimize vulnerability (e.g., by getting citizens or mobile assets out of harm’s way before a storm hits) and help to prevent or mitigate the impact of a potential threat (e.g., disrupting a terror plot or isolating carriers of a highly infectious disease). Additional time pressure stems from the relentless pace of the 24-hour news cycle. Strategic decision makers and their organizations must cultivate the capacity to diagnose situations and formulate responses under severe time pressure. Thus, crises force decision makers to make some of the most consequential decisions in public life under extremely trying circumstances.

Applying the Crisis Definition to Critical Infrastructure Disruptions

The infrastructure systems of concern are labeled critical for a reason. These are systems vital to sustaining life, health, welfare, and functionality in highly developed societies, and disruptions to them tend to have a dramatic and cascading impact on core societal values. Furthermore, policy decisions regarding managing the consequences of disruption and priorities as well as resource allocations regarding system restoration involve (often competing) core values for societies and organizations. Political leaders and infrastructure operators face typical dilemmas relating to the protective shutdown of systems. Proactively shutting down systems in the face of threat or hazard warnings can help reduce exposure and avoid damage. For example, authorities shut down the U.S. civilian air transportation system completely in the wake of the 9/11 attacks (Gordon et al., 2007). Similar preventive measures—with dramatic impacts on air transportation—and indirectly ground transportation—initially reduced the risk to air traffic stemming from volcanic dust during the eruption of an Icelandic volcano in 2010 (Alexander, 2013; Brooker, 2010). However, it should be noted that such protective measures can cause severe disruption and post-decision controversy in and of themselves—especially if the potential threat triggering the measures does not materialize. Conversely, failure to take protective measures in the event of a major attack or destructive storm can expose government leaders and infrastructure operators to even more devastating liability and criticism. To take another example, for example, failure of the power grid can cause life and health-threatening consequences through impacts on food storage and preparation, hospital and critical care functionality, traffic management, temperature (heating or cooling of indoor spaces), and so on. Failure to maintain or rapidly restore service can negatively affect trust in government and in critical infrastructure operators alike (Newlove et al., 2003). To take another example, recent cyberattacks have caused severe financial and reputational damage to individuals and institutions as indicated by the cyberattacks on the 2016 U.S. election, the Equifax data breach, the NotPetya data destruction attack, and breaches targeting major retailers. Furthermore, as so-called cyberphysical systems become more common and cyberweapons become more sophisticated, cyberattacks are liable to cause material, as well less tangible, forms of damage and pose severe threats to human life as well as information security.⁷ The security firm Thales frames this change in the potential consequences of cyberattacks as a move from attacks on business information technology systems that result in mostly “financial and reputational risks” to attacks on industrial control systems that can also result in “safety and operational risks” (Thales, 2013). Finally, it should be noted that many disruptions also raise significant issues of justice, fairness, and liability with regard to the distribution of responsibility for costs or other negative consequences stemming from the direct and indirect impacts of critical infrastructure disruption.

Core Crisis Management Tasks

Several decades of intensive empirical research on crisis management shows that leaders face recurring challenges when confronted with (the prospect of) community (or organizational/national/international) crises (Boin et al., 2017).¹⁰ These are:

These tasks are germane to leaders and other crisis managers across sectors and are central not only to effective crisis leadership in a particular incident but also to creating better preconditions for future incidents and resilient adaptation to changing environmental conditions over the longer term. Hannah et al. (2009), like Leonard and Howitt (2012), suggest that different forms of leadership may be needed in different phases of a disaster or crisis. Helsloot et al. (2012), with their focus on “mega-crises,” describe leadership challenges across crises of differing size and scope. The following conceptualization identifies crisis leadership tasks likely to arise during a variety of extreme events.

Sense-making in crisis refers to the challenging task of developing an adequate interpretation of what are often complex, dynamic, and ambiguous situations (cf. Stern, 2015; Weick, 1988, 1993). This entails developing not only a picture of what is happening but also an understanding of the implications of the situation from one’s own vantage point and that of other salient stakeholders. As Alberts and Hayes (2003) put it: “Sense-making is much more than sharing information and identifying patterns. It goes beyond what is happening and what may happen to what can be done about it” (p. 102).

Making sense of critical infrastructure crises is a challenging task. As this discussion has shown, uncertainties, ambiguities regarding threats, hazards, latent vulnerabilities as well as complex interdependencies within and among critical infrastructure systems make sense-making extremely difficult. Even when warnings are forthcoming—for example, with regard to natural disasters such as hurricanes and flooding where monitoring and detection systems have improved significantly in recent years in many parts of the world—it is often very difficult to predict with confidence how particular critical infrastructures will be affected or what the cascading effects of disruptions might be. This task is even more difficult with relation to novel, relatively unanticipated crises in which information and data-sharing networks and expert communities of practice may be underdeveloped and much of the sense-making work must take place on a relatively improvised basis. This was the case, for example, to a considerable extent during the volcanic ash cloud crisis sparked by volcanic activity in Iceland in 2010 bringing commercial air traffic to a halt over much of Europe (Parker, 2014). In such cases, difficult scientific and technical assessments with regard to “exotic” issues must be produced and communicated to governmental and other key decision-makers under crisis conditions. In many countries this still occurs largely on an ad hoc basis, although others such as the United Kingdom have developed (or are in the process of developing) Scientific Advisory Groups for Emergencies [SAGE].¹¹

A first step in making sense of a critical infrastructure crises is to identify key public, private, and nonprofit actors and stakeholders and gather information via various forms of (social and technical) networks associated with or connecting across critical infrastructure sectors. Many highly developed countries have or are in the process of developing strategies and collaborative fora to improve critical infrastructure protection and resilience. Such efforts, hopefully undertaken well before the onset of a crisis can improve the flow of information, contribute to improved situational awareness, and facilitate “heedful interrelating” and other forms of crisis management cooperation among diverse sets of actors involved in governing, operating, or using critical infrastructures.¹² Making sense of critical infrastructure crises is, of course, facilitated by deeper and more dynamic understanding of systems and cross-system interdependencies. As a result, efforts to improve technologies for mapping, modeling, simulation, and visualization of such systems can contribute not only to system design improvements to improve resilience and reliability but also to capacity for crisis management and rapid restoration.¹³ This issue will be discussed further in the final section of this paper, Preparing for Critical Infrastructure Crisis Management.

Another promising development—to the extent that communications and mobile data networks remain available—for improving situational awareness in critical infrastructure crisis is citizen crowdsourcing. Armed with smartphones and web and geographic information system (GIS)-based platforms for aggregating, analyzing, integrating and transposing data, citizens as well as various forms of organized official or unofficial responders can act as sensors (Akhgar et al., 2017; Stern, 2017b).

Decision-making refers to the fact that crises tend to be experienced by leaders (and those who follow them) as a series of “what do we do now” problems triggered by the flow of events. These decision occasions emerge simultaneously or in succession over the course of the crisis (Newlove et al., 2003; Stern, 1999; Stern et al., 2014). Coping with critical infrastructure disruptions and managing the restoration process tends to require an interdependent series of crucial decisions and to be taken in a timely fashion under very difficult conditions by a variety of public, private, and nonprofit sector actors and stakeholders. Of note, governance arrangements may vary significantly among highly developed countries regarding the extent of public control, regulation, and ownership of critical infrastructures.

It is important to keep in mind that decisions in critical infrastructure crises will, by definition, be taken under highly stressful conditions of considerably uncertainty, time pressure, and threat to core societal and organizational values, as noted above. Furthermore, conflicts among values are to be expected, contributing further to the decisional stress load facing decision-makers (cf. Lebow & Stein, 2019).

It may be helpful to consider typical examples of the kinds of decisions that are likely to arise in the context of a critical infrastructure crisis. Taking a threat/disruption to electric power infrastructure as an example, the following are just some of the very difficult crisis decisions that may be required (Newlove et al., 2003; Nye, 2010; Parker et al., 2009):

Thus, decision-making in critical infrastructure crises, as in other types of crisis events, is extremely complex, difficult and demanding.

Meaning-making (and crisis communication) refers to the fact that leaders—from public, private, and nonprofit sectors alike—must attend not only to the operational crisis communications challenges associated with a contingency, but also to the ways in which various stakeholders and constituencies perceive and understand them. Because of the emotional charge associated with disruptive events, followers look to leaders to help them to understand the meaning of what has happened and place it a broader perspective. By their words and deeds, leaders can convey images of competence, control, stability, sincerity, decisiveness, and vision—or their polar opposites.

Critical infrastructure crises can pose unique challenges with regard to crisis communication in general and meaning-making in particular. First of all, information and communications technology—and critical infrastructures which support them—may be directly affected by the impact of accidents, natural hazards or attacks associated with the crisis. Extreme earth (windstorms, floods) and space weather (e.g., solar storms) events as well as physical and cyberattacks can disrupt the capacity to communicate via both wired and wireless communications networks for shorter or longer periods. In addition, many types of events (again including both deliberate attacks, major accidents, and natural hazard–based events) may be associated with increased citizen and responder communications traffic to the extent that vulnerable communications networks may fail or have severe degradations regarding access, reliability, or quality of communications. As a result, crisis managers may have difficulty in sending and receiving messages and sharing data when ability to do so matters most.

For example, during the prolonged summertime power outage in Auckland, New Zealand in 1998, the normal press briefing facilities (among other key functions and coordination centers for city government) were unuseable because of high temperatures and lack of functioning elevators. As a result, Mayor Les Mills ended up conducting much of his business and media interviews from his car and mobile phone from various locations around the city, which complicated the coordination of operations and messaging. The power company, Mercury Energy, which was responsible for the outage, was kind enough to offer the mayor the use of their press briefing facilities, which had functioning power and communications. Unfortunately for the mayor, this also entailed giving his briefings from a podium boldly displaying the logo of what was at the time a very unpopular critical infrastructure operator. This served to symbolically associate the mayor with the power company at a time when a degree of critical distance would have served him better politically. It should be noted that he lost the next election, in part due to what some regarded as a subpar performance in critical infrastructure crisis management (Newlove et al., 2003).

Hurricane Maria in Puerto Rico in the fall of 2017 is another excellent example. Vulnerable communications and electrical power networks, already battered by Hurricane Irma, were devastated by Hurricane Maria. As a result, much of the response and early recovery effort took place under particularly challenging conditions and large portions of the stricken island remained without power and electricity as not only days but weeks and months passed. Modular cellular technologies (so-called cells on wheels, or COWS) were helpful in restoring communications in certain urban or other particularly sensitive areas but were insufficient to restore communications capacity for the bulk of the island.¹⁴ Degraded communications capacity can limit the ability of communities to self-organize and coordinate citizen-based efforts to respond to and recover from disasters using social media and smart communications technologies—an increasingly important dimension of resilience in large-scale events (Akhgar et al., 2017; Stern, 2017b; Stern, 2017a).

Hurricane Maria also contains good examples of the ways that leaders at various levels of government engage in competitive attempts to demonstrate empathy and engagement as well as endeavoring to shape the narratives and public perceptions of the event. While President Trump and acting Secretary of Homeland Security Elaine Dukes suggested that the response to Maria in Puerto Rico was effective and a “good news story,” Carmen Yulin Cruz, the Mayor of San Juan, appeared on CNN wearing a black T-shirt with the text “We are dying,” creating vivid images that went viral via social and traditional media far beyond the island. Similarly, the tone set by President Trump’s combative approach to crisis communication—which included sharp criticism via Twitter of the leadership of Mayor Cruz and allegations that the Puerto Rican locals were not doing their part with regard to response and recovery—proved highly controversial.¹⁵

Ending and Accounting refers to the nontrivial task of finding the appropriate timing and means to end the crisis, manage accountability processes, and return to normalcy. Furthermore, attempting to end a crisis prematurely can endanger or alienate constituencies who may still be in harm’s way, traumatized, or otherwise continue to be politically or emotionally invested in the crisis. Crises may be particularly difficult to terminate if the operational challenges lead to a so-called crisis after the crisis in which serious recriminations—resulting in losses of trust and legitimacy—are launched against those who failed to prevent, respond to, or recover effectively from a negative event.

Critical infrastructure crises, like other forms of crisis, may have different trajectories and different combinations of operational and legitimacy dimensions. Some are “sudden onset” in which a dramatic event such as an earthquake or hurricane causes dramatic damage. Others may demonstrate a “creeping” quality whereby a threat slowly manifests or the functionality of a critical infrastructure slowly degrades for a prolonged period before attracting notice. The water crises that have affected many cities around the world—from places as diverse as Mexico City; Flint, Michigan; and Hoosick Falls (New York)—are good examples of vulnerabilities mounting as a result of environmental change; rapid population growth and development; industrial pollution; or unsafe properties of aging infrastructure installations (use of lead pipe in water distribution systems); and so on.¹⁶

Similarly, some critical infrastructure crises demonstrate rapid and effective restoration of the pre-crisis status quo, whereas others may exhibit prolonged disruptions of functionality. This issue of differing time frames and trajectories is even more troubling when combined with the fact that infrastructure risk experts claim that “Many security risk decisions do not account for the life cycle of a system or an infrastructure” Assante, 2009:4), That is, there is a disconnect between short-term decision making—and the incentives surrounding it—and the very long terms for which many infrastructures are expected to remain operational. Logically, this could both make crises more likely but also suggests that crisis decision-making might be likely to be short-sighted. Even in shorter, more clear-cut cases of infrastructure crisis, public evaluation of leadership and authorities can be mixed or contradictory (Helsloot & Beerens, 2009), in long-term and complex situations such as these, the effect is likely to be similar.

Tied to the trajectories, some disruptions and their responses may morph into legitimacy crises after the initial infrastructure crisis (cf. Boin et al., 2017; Fischbacher-Smith, 2006). When it comes to critical infrastructure crises, such legitimacy crises may focus on questions such as:

The reader should note that these issues are likely to arise in multiple accountability fora: parliamentary bodies, professional or industry associations or governance bodies, regulatory agencies, courts, and the media. Furthermore, crisis experiences and their associated accountability processes may profoundly challenge and in some instances destabilize established practices and patterns of governance (including regulation and divisions of authority and responsibility across sectors and levels of government) with regard to managing critical infrastructure sectors. When critical infrastructure operators and industries are perceived to have been—or determined in a court of law to have been—negligent, strong pressures for change and significant legal and financial liability for operators and other private sector actors may emerge. This can take the form of proposals for increased regulation, incentivizing safety and resilience, heightened operator responsibility for damages, or even withdrawal of the “permissive consensus” that allows potentially hazardous activities such as nuclear power or process industries involving potentially dangerous chemicals to continue. The Fukushima nuclear accident stemming from the great Tohuko earthquake and tsunami is a good example, with profound repercussions for the nuclear power industry even in countries on the other side of the world such as Germany (Bernardi et al., 2018).

The highly complex systems of critical infrastructure governance entailing various forms and degrees of regulation and legislation, diverse patterns of ownership, different varieties of public-private partnership in place in many countries have profound implications for accountability with regard to critical infrastructure disruption/restoration and crisis management (cf. Dunn-Cavelty & Suter, 2009). Awareness of and outcomes from previous accountability processes should inform preparedness, response, recovery, and postcrisis efforts toward learning and reform (which will be discussed next).

Learning and changing requires an active, critical and reflective process that recreates, analyzes, and evaluates key processes, tactics, techniques, and procedures in order to enhance performance, safety, capability, and so on. The learning process has just begun when a so-called lessons learned document has been produced. To bring the learning process to fruition, change in management/implementation must take place in a fashion that leaves the organization with improved prospects for future success (Boin et al., 2017; Deverell & Olsson, 2009; Stern, 2015; Stern et al., 2014). Note that political and organizational cultural obstacles and various forms of disincentives often prevent effective learning and change from taking place (Stern, 1997).

Critical infrastructure disruptions—like other forms of crisis—provide unique chances for learning and opportunities for change. Latent or unrecognized vulnerabilities and interdependencies manifest themselves in dramatic fashion, providing opportunities for redesign and reform of technical systems, institutional frameworks, tactics, techniques, procedures, and so on. The abrupt manifestation of urgent and difficult problems can stimulate creative improvisation as well as new forms and constellations of collaboration and innovation. However, taking advantage of these opportunities requires reflection, inquiry, leadership, and resources to identify adaptive lessons (cf. Heifetz & Laurie, 2001) and translate them into more resilient, reliable effective and legitimately functioning systems for critical infrastructure and crisis management.

Hurricane Sandy serves as a vivid example. The second-largest Atlantic storm on record, Sandy affected the East Coast of the United States from Florida to Maine, as well as states as far inland as West Virginia, Ohio, and Indiana. The storm made landfall in southern New Jersey on October 29, 2012, battering the densely populated New York and New Jersey region with heavy rains, strong winds, and record storm surges. The storm’s effects were extensive, leaving more than 8.5 million customers without power, causing widespread flooding throughout the region, and contributing to acute fuel shortages in parts of New York and New Jersey. In the United Statesl the storm damaged or destroyed hundreds of thousands of homes, caused tens of billions of dollars in damages, and killed at least 162 people.¹⁷

In addition, Hurricane Sandy caused severe damage and disruption to multiple critical infrastructure systems in New York and New Jersey, including transportation,¹⁸ energy (electric power and fuel distribution), water provision,¹⁹ and healthcare (e.g. hospitals and nursing homes). Numerous points of vulnerability and challenging problems emerged ranging from the vulnerability of the subway system and the cable installations servicing it to flood damage, cascading interdependencies between electric power and various nodes in the fuel distribution network, and fragility of hospital power backup systems in the face of massive storm surge and flooding (e.g., the evacuation of N.Y.U. Langone Medical Cente).²⁰ Following Hurricane Sandy, there have been a large number of after-action reports, inquiries, and research studies about various aspects of the crisis, response, and recovery associated with it. Many of these efforts have focused on or significantly involved critical infrastructure functions and actors, and a variety of important lessons have been learned and changes made.²¹

The experience of the Atlantic hurricanes of 2017 has begun to be systematically processed via a postcrisis multidimensional after action and research effort of the kind that took place for Hurricane Sandy and some previous major Atlantic storms (e.g., Katrina). There are significant indications that important lessons were learned from Sandy and previous cases that informed the proactive crisis management by many governmental actors and critical infrastructure operators regarding response and restoration efforts tied to Hurricanes Harvey and Irma (Federal Emergency Management Agency [FEMA], 2018). However, the much more problematic response and delayed restoration of critical infrastructure services (e.g., power, fuel, telecommunications, water) to Hurricane Maria in the U.S. territory of Puerto Rico is both instructive and troubling. Clearly, resilience and preparedness for critical infrastructure crisis management can develop in an uneven fashion even in highly developed countries, leaving pockets of particular resilience and particular vulnerability. Contextual factors such as geographic centrality/remoteness, regional economic strength, and robustness or fragility of particular infrastructures, functionality of local governance, coincidence with other simultaneous or near simultaneous events, and not least political will can all affect response and recovery. The increased frequency and severity of events not only poses challenges to societal and organizational resilience but also to the ability of organizations to focus on and extract the lessons from individual events. Shifting to “seasonal” rather than individual event-based after-action reviews (as FEMA did regarding the eventful 2017 hurricane season) is understandable but can result in missed learning opportunities associated with particular events (FEMA, 2018).

Although effective production and adaptive implementation of “lessons learned” is very challenging and requires significant investment of time and resources, failure to learn from experience is a recipe for stagnation and perpetuation of vulnerability. This may be extremely costly in humanitarian, political, and financial terms for the key actors involved in the governance and operation of critical infrastructure systems (Boin & McConnell, 2007; Boin et al., 2017).

Improving International Cooperation and Collaborative Response and Recovery of Critical Infrastructure Systems

As seen previously, contemporary critical infrastructure systems transcend national boundaries in various ways. Disruptions may originate in physical events occurring in or cyberattacks launched from on one or more countries but end up affecting many others. Many natural hazard scenarios—such as extreme terrestrial and space weather events—can have an impact on critical infrastructures across national boundaries and influence global supply chains. Preventing attacks and limiting the damage of disruptions stemming from natural, man-made, or mixed events often requires rapid, even real-time, exchange of intelligence and system status information as well as expert knowledge. In addition, effective coordination of protective, response, and recovery measures is essential to prevent further cascades of disruption and damage and facilitate rapid restoration (e.g., via effective use of scarce repair capabilities and component resources across countries). This is true not only of power and other energy systems, but also of transportation, and telecommunications.

Building Trust and Relationships in Support of Crisis Management Across the Public and Private Nonprofit Divide

A broad set of public, private, and nonprofit players share authority, responsibility, capabilities, information, knowledge, and resources needed to cope with severe critical infrastructure disruptions and their cascading effects. Although these actors are diverse in many respects, they have significant common interests in mitigating disruptions and promoting rapid restoration of critical infrastructures. Whole society/community approaches increasingly in emergency management and resilience-building are highly relevant to improving capacity to cope with critical infrastructure failure as well. Creating opportunities, incentives, and fora for these individuals and organizations to familiarize themselves and cooperate with each other—before, during, and after crisis events—is extremely valuable. Governments are actively supporting this process in many of the most developed countries (e.g., United States, United Kingdom, Sweden, Switzerland²²) via resilience fora, critical infrastructure sector groupings, and other equivalent efforts, although much more can be done.

Understanding, Improving, and Planning Relationships with Internal and External Sources of Technical “Reachback” and Expertise

Although the critical importance of access to technical expertise is perhaps most obvious in the relatively arcane cybersecurity field, parallel needs for technical expertise arise in many other types of infrastructure-related crises. During the Deepwater Horizon spill, there were significant challenges of accessing, coordinating, and integrating different types of expertise into the crisis management effort. (Birkland & DeYoung, 2011; cf. Koraeus & Stern, 2013). Public and private sector crisis managers were dependent upon a small community of experts in offshore drilling and underwater geology from many organizations and sectors (Busch & Givens, 2012). This is, of course, just one among many other types of expertise urgently needed to deal with the cascading environmental, social, and economic consequences of the oil spill.

Similar mobilization of expertise is necessary in many types of infrastructure crises. In cyberattacks or large-scale cyberdisruptions, owners and operators of infrastructure need to be able to assemble rapidly highly qualified teams of cybersecurity, digital forensics, or industrial control system experts. This expertise need not be resident in the owners’ or operators’ company or agency. There certainly could be internal resources such as the personnel at the Security Operations Center (SOC) or incident response team, but would also include contractors and security vendors, industry or sectorwide resources such as Information Sharing and Analysis Organizations (ISAOs) and Information Sharing and Analysis Centers (ISACs), and even regulators and law enforcement. Building the “right” team at the right time is key to sense-making, decision-making, and meaning-making in critical infrastructure crises.

Improving Knowledge and Understanding of Critical Infrastructure Systems Through Mapping, Modeling, and Simulation

As demonstrated in the preceding sections, critical infrastructure systems are highly complex and dynamic. Such systems and the environments in which they operate evolve and change on an ongoing basis. Human ability to understand and predict outcomes with regard to disruptions and protective interventions with regard to such systems is limited (cf. Perrow, 1984; Roe & Schulman, 2016) and must be leveraged by investments in efforts to map, monitor, and develop deeper understanding of them. Significant advancements of methodology and technology may be harnessed to improve mapping (both geographic and functional) to identify pathways to and from failure and toward rapid restoration. Empirically based modeling, simulation, and visualization tools can be used to improve reliability and resilience more generally as well as to facilitate effective response and rapid restoration to critical infrastructure disruptions.²³

Reinvigorating Critical Incident Response and Recovery Planning

Emergency planning in many countries has been criticized for a tendency toward unrealistic planning assumptions regarding environmental and system vulnerabilities, magnitude and duration of disruptions, and availability of capabilities and resources (Clarke, 1999; Eriksson & McConnell, 2011) as well as for neglecting key actors, stakeholders, and particularly vulnerable populations. Conversely, improved planning for critical infrastructure disruptions will:

Training and Exercises Involving Key Actors—at Both Strategic and Operational Levels—From Government, Critical Infrastructure Operators, and Other Key Private and Nonprofit Sector Stakeholders

Government and critical infrastructure operator leaders should take steps to make sure that they and their team members, other key subordinates, and counterparts from essential partners are educated, trained, and exercised in preparation for critical infrastructure crisis management.²⁴ When facing major crises, partnership between strategic leaders and “professionals” inside and outside of government is essential. This means that political leaders who are not “professionals” must be educated as to the nature of critical infrastructure crisis management, informed of what is required of them in scenarios and contingencies involving (the risk of) significant disruption to critical infrastructure systems, familiarized with relevant crisis planning and organization, and equipped to engage effectively in meaningful communicative interaction with others inside and outside of their organizations. Individual and collective crisis management skills are best acquired and honed through hands-on practice in simulated as well as real-world incidents and crises. There are a wide variety of powerful instructional designs and techniques (both traditional and technology enhanced) suitable for critical infrastructure crisis management training and exercises.²⁵ Instructional designs and techniques should be consciously chosen and explicitly adapted to the goals and purposes of a given training or exercise for maximum positive impact.

Improving Accountability and Enabling Organizational and Interorganizational Learning

Critical infrastructure failures raise significant issues of political, legal, and financial accountability. Regulatory regimes, divisions of labor, and responsibility continue to vary greatly across countries and sectors and ambiguities and gaps persist, contributing to suboptimal resilience and suboptimal preparedness for response and rapid restoration. As noted previously, public, private, and nonprofit sector actors participating in the governance and operation of critical infrastructure systems are subject to multiple forms of accountability and must be prepared to answer questions not only in Congress (or Parliament), but also literally in court and in the metaphorical “courts” of (social) media and public opinion.

Disruption of critical infrastructure systems, by definition, may be costly in lives and treasure, as well as in organizational (and in many cases personal) trust and legitimacy. As a result, it is imperative to learn the lessons of experience. Doing so effectively requires fostering a culture of safety and resilience in which information is shared and performance before, during, and after crises that are subjected to benchmarking and critical scrutiny. Experts should resist temptations to withhold “embarrassing” information and whitewash subpar performances in favor of systematic and methodologically informed inquiry and vigorous and forward-looking implementation of measures to address preparedness gaps and shortcomings.

There is good reason to believe that managing infrastructure crises effectively could be improved by a better integration of the scholarship and practice of crisis management and the parallel scholarship and practice of infrastructure management. Achieving this integration will require several things. First and foremost, it requires viewing critical infrastructure disruptions—whether resulting from natural disaster, industrial accident, or adversarial action—as crises that threaten core values and entail decision-making and management under conditions of uncertainty and urgency. If critical infrastructure disruptions can be fruitfully viewed as crises—then good practices from the crisis management literature with regard to sense-making, decision-making, meaning-making, ending and accounting, and learning and changing (Boin et al., 2017; Stern, 2013) promise to be helpful in improving organizational and societal critical infrastructure resilience as well. Finally, if infrastructure disruptions are indeed understood as crises, and models as well as experience-based lessons from crisis management are brought to bear in seeking to understand better and meet such challenges, then both practitioners and scholars of crisis management and infrastructure management will be better equipped to learn from each other (Trainor et al., 2018).