Benjamin D. Trump
Igor Linkov
Ready for the Next Crisis? Investing in Health System Resilience
OECD Health Policy Studies
2. Investigating resilience in other sectors
Abstract
Although relatively new to health systems, the resilience methodology has decades of application in other fields. This chapter outlines the common features of resilience across various sectors. It discusses how resilience has been applied in infrastructure, finance, the environment, cybersecurity, disaster response, and medicine and public health. Common features and lessons for health systems, and common methods of testing for resilience are identified. Improving health system resilience should use the lessons learnt from its application in other sectors, with a focus on systems adapting to shocks and an emphasis on a multi-systems view.
Key findings
Resilience has decades of application in disciplines beyond health. All resilience analysis requires stakeholders to adopt a perspective that recognises the connections between systems – interdependency – and the need to go beyond risk to absorb, recover from and adapt to disruption.
Resilience is a rapidly developing field of endeavour. Many different approaches are used to analyse and improve resilience of systems, from scorecards and tabletop exercises to simulations using artificial intelligence. More advanced approaches require both more data and more sophisticated approaches to modelling the systems of interest.
Multiple sub-systems must work together for any system to be resilient. A generic conceptualisation of the sub-systems includes, but is not necessarily limited to, the following domains:
physical – equipment and facilities
information – data
cognitive – understanding and decision making
social – interactions between actors.
Resilience is a function of system performance. The capacity for systems to recover from and adapt to disruption cannot be engineered in silos – it should be undertaken holistically. Longer time horizons and the complex, adaptive nature of the environment and population that the system serves should be considered. Improving one domain without the others may not improve resilience. For example, over-reliance on physical infrastructure (such as intensive care unit beds) without considering social behaviour (for example, adherence to containment and mitigation policies) will not produce resilience in response to a pandemic. Resilient systems should foster risk prevention and mitigation to specific and well-characterised threats, but must also possess a “risk-agnostic” or “all-hazards” capability to manoeuvre in the face of disruption.
In the infrastructure sector, disruptions that place transportation networks or energy facilities offline cannot be effectively engineered against. A focus on maximum efficiency can leave systems vulnerable to disruption. This has meant that resilience efforts focus not only on mitigating risk and preparing for shocks, but on recovering from and adapting to them.
In the finance sector, the global financial crisis (2007-08) was the impetus for resilience thinking and implementation. Financial systems have operationalised resilience by trying to prepare for and absorb shocks through stress testing. The purpose of these tests is to identify areas of financial systems that should be bolstered. The results of these tests are published and can be brought into the public domain.
In the environmental sector, systems still require resilience, although they are formed rather than designed. For example, the Environmental Protection Agency of the United States has explored qualitative approaches to evaluating resilience. Resilience to the impacts of climate change is also a growing field.
Further, experiences in cybersecurity and digital systems show the potential for feedback loops and interdependencies between systems. For example, cyber attacks that disrupt the provision of essential goods and services – such as fuel – can result in hoarding behaviour, further exacerbating shortages.
Experiences in the health sector, including outbreaks of the Ebola virus, also demonstrate the importance of considering resilience at different levels: the individual, the local community, institutions and the whole system. These are linked with each other and the socio‑economic determinants of health.
The key lesson learnt from analysing other sectors is that health systems should use the basic principles of resilience: preparing for, absorbing, recovering from and adapting to adverse events. For health systems, which are frequently confronted by challenging events, developing the capacity to recover from and adapt to emergent conditions is necessary to prevent collapse and allow health care to continue.
The question that follows – which the remainder of this report explores – is “What components of the health system should be improved to strengthen health system resilience to be ready for the next crisis?”
2.1. Introducing resilience
Resilience has decades of application in a range of fields, including psychology, military operations, and civil and environmental engineering. Synonyms commonly used for resilience include “elasticity” and “toughness”. Resilience or resilient behaviour in systems is focused on these ideas, but more modern applications have analysed how systems rebound from disruption. The many fields that apply resilience have derived various methods and tools by which resilience is analysed and implemented.
Resilience as a philosophy and methodological practice underscores absorption of and recovery from a disruption. From a philosophical lens, this mind-set is grounded in maintaining system survival while accepting risks will inevitably materialise. From a methodological lens, practitioners of resilience seek to optimise available resources to safeguard their systems from a host of possible and even unknown threats, while acknowledging threats are inevitable. This can be contrasted with the more traditional approach of risk management and assessment, which focuses on systemic threats, employing risk science on a per-threat basis to quantify how individual threats can exploit a system’s vulnerabilities.
Figure 2.1. The disruption cycle: the four stages of a response over time
Sources: Adapted from OECD (2019[1]), “Resilience‑based Strategies and Policies to Address Systemic Risks”, https://www.oecd.org/naec/averting-systemic-collapse/SG-NAEC(2019)5_Resilience_strategies.pdf, and National Research Council (2012[2]), “Disaster resilience: A national imperative”, https://doi.org/10.17226/13457.
Ultimately, applications of resilience are defined by two components: time and space. Considering time, a system’s resilience is not a singular, temporal phenomenon; its application is the process of understanding how a system absorbs and overcomes disruption. Considering space, analysts of resilience must determine the interconnections in systems once disruption occurs – keeping in mind that those connections may be either easily identifiable or hidden.
Resilience has been adopted by governments and international actors – such as the OECD, the Group of Twenty (G20), the International Risk Government Council – as a term to address various risks that could cause cascading, negative effects through systems that are interdependent over time and scale. The OECD adopted resilience as an international strategy after the financial crisis in 2007‑08, and it has been adopted increasingly to mitigate and address disruption. The G20 published a Note on Resilience Principles in G20 Economies in 2017, which detailed advice to “strengthen macroeconomic foundations and policy frameworks to reap the benefits of openness to trade and international capital flows” (Nienaber, 2017[3]). International actors, including the OECD, have promoted resilience in addition to conventional use of risk assessment as a method to get systems ready to recover and overcome disruptive shocks efficiently, providing a softer landing for the entire system and its individual units (Linkov, Trump and Fox-Lent, 2016[4]). Resilience methodology has proved useful for increasingly intricate and interdependent systems through identifying the vulnerabilities that disruptors present as well as to strengthen system-wide capability despite the magnitude and likelihood of shocks.
Alberts and Hayes (2003[5]) characterise four network-centric operation domains essential for system flexibility, which they define as “the ability to successfully effect, cope with, and/or exploit changes in circumstances”. This process of “resilience thinking” makes its users consider the vast array of decisions that influence a system’s performance. Domains are affected in different ways, and one domain’s success in defence against disruption does not guarantee success in others. Additionally, resilience is at its maximum strength for a system when all its domains consider a resilience approach. These network-centric operation domains are as follows (Hayes, 2004[6]; Alberts, 2007[7]):
physical: sensors, facilities, equipment, system states and capabilities
information: creation, manipulation and storage of data
cognitive: understanding, mental models, preconceptions, biases and values
social: interaction, collaboration and self-synchronisation between individuals and entities.
The domains can be applicable to any complex system, but are targeted toward resilience of systems (Roege et al., 2014[8]; Collier and Linkov, 2014[9]).
The physical domain is typically the most obviously affected, as this is where the disruption meets the physical environment. This includes infrastructure – ranging from transportation systems (highways, streets, railways, airports, etc.) to cyber networks and energy infrastructure that delivers goods and services to populations (DiMase et al., 2015[10]). These areas of impact are typically those most visibly disrupted, but other domains can also be affected because of disruption to the physical sector. Infrastructure threats include environmental disruption (i.e. disasters and natural hazards) and anthropological disruption (i.e. socio-political factors). The physical domain’s objective for resilient behaviour is to restore infrastructural systems to full integrity, post disruption.
The information domain houses data and knowledge. It also incorporates how data are changed and shared. Examples include public or private databases, which have increasingly become targets for attack (Zhao and Zhao, 2010[11]; Osawa, 2011[12]). Furthermore, information domain risks have increasingly attacks on online communications. These attacks range in impact from individual humiliation to state‑wide security risks (Berghel, 2015[13]; Petrie and Roth, 2015[14]). Such attacks are growing and inevitable in the Information Age (Kaur, Sharma and Singh, 2015[15]), which necessitates protection against these risks and bolstering of the information domain for public and private companies (Lino, 2014[16]). This domain’s objectives for resilient behaviour are to plan and prepare individual and system assets for any range of attacks, while ensuring that the system can react quickly. Risk preparation, absorption, recovery, and adaptation in the information domain entails unique tools and datapoints to monitor, and are particularly critical for governments and businesses (Björck et al., 2015[17]; Collier et al., 2014[18]; Linkov et al., 2013[19]).
The cognitive domain comprises the beliefs, perceptions, levels of awareness and values that inform individual and system decision-making (Eisenberg et al., 2014[20]). Parallel to the social domain, the cognitive domain is regarded as the “locus of meaning, where people make sense of the data accessed from the information domain” (Fox-Lent, Bates and Linkov, 2015[21]). These factors can be overlooked or even dismissed as a result of common reliance on physical systems to facilitate public response to disruption, yet the tenets of the cognitive domain are invaluable to how a system’s resilient operations are undertaken (Wood et al., 2012[22]). Without policy recommendations that factor in the beliefs and perceptions of individuals and the system, even robust resilience plans may fail. This is particularly important when a disconnect exists between policy makers and the public – for example, with international infrastructure development projects geared toward health-based interventions. In these cases, sensible policy may refute common opinion or the belief of local people, causing discord in the system’s resilience.
The social domain incorporates the interactions both between and within the entities involved in the system. Careful attention must be paid in the social domain to building community resilience. As social aspects of societies can affect physical health (Ebi and Semenza, 2008[23]), communities with strong social cohesion can recover better from disruption, such as from a pandemic. Both the social and information domains require trust in information sources. If the community does not trust the provider of information, a delay can occur until the content of the information is relayed and considered trustworthy (Longstaff, 2005[24]). This has a notable impact on resilience plans: if social actors do not implement a resilience plan for any reason, the resilience of an entire system can diminish (see the chapter on containment and mitigation).
This chapter discusses how resilience is characterised in other sectors to identify instances of resonance and dissonance for health systems. By taking the lessons learnt from applying resilience in other sectors, health systems have the potential to be at the forefront of resilience by learning how to prepare for, absorb, recover from and adapt to disruptions.
2.2. Infrastructure
Resilient infrastructure systems are a target of the United Nations Sustainable Development Goals and Sendai Framework (UNDRR, 2015[25]). While this aim is subjective, the vision of the United Nations remains clear: settle the disparity between developing and developed countries’ infrastructure systems. The picture of what resilience entails for infrastructure systems, however, remains unidentified, since infrastructure is a broad “umbrella” term encompassing vastly different services; and the connections between infrastructure systems (and thus their failures) are noticeable, but difficult to quantify or predict.
Infrastructure systems are diverse and numerous; thus, the discussion around characterising infrastructure resilience is similar. From power generation, transportation and water to telecommunications and beyond, resilience in infrastructure systems typically refers to the infrastructure’s ability to “bounce back” from any disruption (Bocchini et al., 2014[26]; Labaka, Hernantes and Sarriegi, 2016[27]; Linkov et al., 2014[28]; Panteli et al., 2017[29]; Vugrin et al., 2010[30]). While the need to prevent and absorb disruption to infrastructure is discussed (Kumar et al., 2021[31]), the inevitability of disruption that places infrastructure systems offline warrants discussion of how to improve resilience for infrastructure (Francis and Bekera, 2014[32]). For example, as the COVID‑19 pandemic surged and air travel decreased due to travel restrictions, the first response of airlines was to maintain profitability and efficiency by reducing staffing (Sun, Wandelt and Zhang, 2022[33]). Such decisions highlighted the interdependencies within supply chains, which challenged airlines to crafter newer, improved models for operation (Reeves and Varadarajan, 2020[34]). As seen through real-time data, the pandemic and its consequences for cancelled flights revealed and compounded the brittleness of an overburdened transportation system.
To begin measuring resilience for infrastructure, existing tools can be categorised based on the Linkov et al. (2018[35]) tiered methodology for resilience assessment.
Tier 1 assessments are coarser in scope yet more readily available, operating at the qualitative level – including scorecards and tabletop exercises.
Tier 2 assessments uphold elements of decision making and analysis to make semi-quantitative analyses, such as resilience matrices and stress testing.
Tier 3 assessments provide granular assessments of system resilience at the link-and-node level of systems through artificial intelligence (AI) and simulations.
Scorecards have been used to provide a snapshot in time of any nation’s infrastructure, such as the Report Card for America’s Infrastructure (ASCE, n.d.[36]) or the UN’s Disaster Resilience Scorecard (UNDRR, 2017[37]). While these provide valuable exercises to begin the discussion of resilience in infrastructure, the outcomes can be somewhat hazy, since generalisation is a requirement for evaluation. Furthermore, scorecards analyse infrastructure using a piecemeal approach. Connections between infrastructure systems are typically not explored, although these interconnections can result in cascading failures (for example, the supply chain challenges that occurred during the first year of the COVID‑19 pandemic, see the chapter on securing supply chains). Since these tools are, however, highly visible and accessible to the public, they act as a vehicle to direct attention to the need for resilience. Scorecards are generally approachable for stakeholders with limited means or within small communities to foster a full resilience study (Ludin and Arbon, 2017[38]; Sachinthana, Chandana and Shehara, 2022[39]).
Tabletop exercises are another form of subjective, qualitative exercise to evaluate the resilience of infrastructure systems. The US Cybersecurity and Infrastructure Security Agency (CISA) provides one example of tabletop exercises, including a host of “what-if” scenarios that decision makers can discuss to form conclusions about where gaps lie in their infrastructure systems (CISA, n.d.[40]). The scenarios present a range of disruptions that might happen, such as earthquakes, wildfires, tsunamis, hurricanes and socio-political threats. Connections between infrastructure systems become apparent as the tabletop exercises guide decision makers with questions about how these systems may be co-dependent or mutually affected.
Resilience matrices have been developed by Fox-Lent, Bates and Linkov (2015[21]) to represent resilience temporally and spatially. From a temporal perspective, the matrices utilise the elements of resilience to consider a system before and after disruption, using preparation, absorption, recovery and adaptation as temporal indicators (the disruption cycle). Spatially, the system is broken down across domains, namely the infrastructure’s physical characteristics, cyber components, social linkages and cognitive requirements. At least 16 unique metrics are created for the 4×4 matrix to operate resilience as a function of space and time. This enables decision makers and policy makers to identify where the system is most likely to fail.
Discussion of stress testing is also prevalent in the literature on critical infrastructure, such as with the nuclear industry. The term “critical infrastructure”1 is a broad one that generally covers any transportation, utility or robust engineered system (Chopra and Khanna, 2015[41]). Evaluation of such a wide range of physical infrastructure and modes of engineering has led to many applications of stress testing for critical infrastructure. For example, unique modes of stress testing exist for water, transportation, telecommunications and energy (Environmental Protection Agency (EPA), 2015[42]; Jovanovic and Auerkari, 2016[43]; Pitilakis et al., 2016[44]; Samoylenko, Panychev and Panychev, 2017[45]). Creation of a combined approach to evaluate all sectors has been increasingly considered over the past decade (Comes, Bertsch and French, 2013[46]; Galbusera et al., 2018[47]; Tsionis et al., 2016[48]).
Tier 3 analyses use robust data analysis to provide resilience considerations for a system with optimal fidelity. These tend, however, to be costly (even prohibitively so for some end-users) and are intended for audiences that have the capability to understand advanced mathematics, AI and modelling through simulations. For example, Galbusera et al. (2018[47]) discuss how network science at the link-and-node level can be used to improve comprehension of heavy infrastructure systems. Through a deep understanding of this type of science, decision makers can optimise individual components of a system.
2.2.1. Everyday road transportation efficiency and resilience
Roadway design targets efficient movement of vehicles through a road transportation network (Samuelsson and Tilanus, 1997[49]; Kuhn, 2010[50]; Hoogendoorn, Van Arem and Hoogendoorn, 2014[51]; Sami, Pascal and Younes, 2013[52]). The designation “efficient” from this standpoint is conferred as a level of service in civil engineering, which is a subjective evaluation to identify congestion in the road network. Often, this type of efficiency sets the goal for roadway officials to maintain a given level of service while using only a certain amount of resources (Chang and Nojima, 2001[53]; D’Este, Zito and Taylor, 1999[54]; Yan et al., 2006[55]; Yamashita, Izumi and Kurumatani, 2004[56]). As one metric, the Texas A&M Transportation Institute uses and assesses the yearly delay spent in traffic per driver when it reports levels of traffic congestion in urban areas (Schrank et al., 2015[57]). Similarly, other studies measure driver delay but at an individual level (D’Este, Zito and Taylor, 1999[54])or by average travel time between commuted trip ends in the roadway network (Allen, Liu and Singer, 1993[58]). These metrics of travel efficiency can be affected by several variables beyond the design of the roadway – such as congestion, weather, construction, or special events (e.g. a parade). Resultant delays from these factors lead to wasted resources including time, money, fuel and emissions (Çolak, Lima and González, 2016[59]; Turnbull, 2016[60]). However, for many roadway transportation systems, this evaluation under normal or typical circumstances leaves out important information regarding the system’s ability to cope with stress and less-than-optimal conditions (Caliendo, Russo and Genovese, 2022[61]).
Resilient infrastructure systems can adapt to atypical situations, both of relatively commonplace (vehicle accidents, road closures, or severe weather) to uncommon (hazardous material spills, mass casualty events, etc.). Since transportation systems play a vital role in emergency response, economic well-being and essential services, road networks have garnered increased policy making attention. Scholars have not yet, however, derived a common definition of resilience for transportation systems that could guide the design lifecycle of roadway network, as it is a multidimensional concept within different fields. For example, in comparison to many fields of engineering – including civil engineering – resilience is defined as a system’s ability to prepare for, absorb, recover from and adapt to disruptions. Transportation resilience has often emphasised the importance of consistent service delivery. Disruption beyond an accepted range of delays (e.g. daily vehicle traffic) must be resolved, and the transportation system’s ability to return to an acceptable service level as quickly and inexpensively as possible. More simple descriptors relate to a system’s ability to minimise or reduce operational loss in transportation services. Colloquially, resilience in this field has also taken on synonyms such as robustness, redundancy, reliability or overcoming vulnerability, which complicates the focus upon system recovery and adaptation (Galaitsi et al., 2020[62]).
Contemporary research on transportation resilience has developed frameworks and measurement methods for resilience, including factors such as total traffic delay, economic loss, post-disaster maximum flow and autonomous system components (Lambert et al., 2012[63]). Since this research is driven by empirical methods, practical concerns are attributed to its effectiveness. It often dismisses indicators that cannot be quantified, and can be affected by heuristics, rules of thumb and subjective evaluation. Other methods of applying resilience in transportation systems include modelling of traffic networks to optimise locations for critical services (e.g. fire stations and hospitals), reducing travel distances and minimising the travel time necessary across the system. These network approaches can be data intensive, requiring information that can be difficult to track down. Furthermore, quantifying resilience for transportation networks typically only serves the system of interest and are not generalisable or easily changed. Since disruption is inevitable, resilient transportation systems must be characterised by their ability to withstand these stresses. As a result, concepts of transportation efficiency and resilience are often not implemented.
2.2.2. Case study: airport resilience and severe weather
Airports, especially international airports, are critical elements of transportation infrastructure that affect urban areas and their economies (Stevens, Baker and Freestone, 2010[64]). Airports (including its infrastructure, energy, communications and labour) are the nucleus of the supply chain for airline travel, including compliance for risk and security, sustainable operations and maintenance, and the production and sale of goods and services. This makes the capacity of airports to recover from and adapt to possible disruptions critical to passenger and worker safety, and to delivering socio‑economic benefits to the areas they serve (Pickard and Gençsü, 2021[65]).
The focus of this case study – Dallas – Fort Worth (DFW) International Airport – ranks second in terms of passengers and third in aircraft movements (operations) globally.2 Due to its heavy traffic, DFW has considerable societal benefits and obligations – both to maintain operations regardless of threats, as well as to mitigate environmental externalities such as carbon output. Lessening disruption effects while maintaining operational capacity has been challenging when combined with the need to mitigate greenhouse gas reductions.
In February 2011, minimum temperatures decreased at DFW between ‑10 to ‑20 (°C), prompting de‑icing to meet necessary traction abilities for runway operations. As a result, only 440 flights could land and take‑off daily, prior to recovery. Even though this event lasted five days, only three were needed for DFW and the airport’s partners to bounce back to relatively normal operation standards. The lack of the airport’s absorption ability was reflected in the steep decline in operations following the storm, but its recovery was demonstrated by its swift return to near-normal capacity. To adapt to future storms, DFW and its partner airlines collaborated to strengthen method for de‑icing as well as the runway infrastructure itself to result in an “all weather airport” status in regard to airport equipment, procedures for winterisation and adaptive monitoring of traction on the runway (Federal Aviation Administration, 2020[66]).
Jumping to February 2021, a stronger winter storm blasted most of the state of Texas with both minimum and maximum below-freezing temperatures lasting eight days. Taking heed from the winter event that occurred in 2011, de‑icing strategies and runway clearances were operated and maintained at normal levels. However, cascading disruptions from breaks in the water line, equipment affected by the low temperatures and reduced ability for staff to commute to their posts eventually led to a systemwide failure akin to what occurred in 2011 (Doss-Gollin et al., 2021[67]). Operations recovery was delayed (recovery took seven days as opposed to three days), meaning that post‑2011 lessons learned did not lead to maintained system-wide resilience.
However, DFW has made post‑2011 improvements. By establishing its Integrated Operations Center in 2021, DFW harmonised essential functions from key stakeholders into a 24/7 operating platform to strengthen collaboration, while decentralising decision making. Due to this, emergency response times to disruptions became faster. For example, when the arrival of fuel for snow and ice equipment was delayed due to the inclement weather in 2021, DFW provided alternative diesel quickly. Airport staff in the Integrated Operations Center received quicker communications, which supported real-time co‑ordination in redirecting diesel, typically used for fuelling its fleet at onsite fuel stations, to snow- and ice‑clearing equipment. DFW had also invested in owning and operating nearby hotels. As the 2021 event was underway, these assets proved advantageous because hotel rooms could be reserved for the occupancy of essential airport workers, helping recovery.
While DFW took a longer time to recover from the 2021 event compared to the 2011 event, this may be due to system-wide complexity. Air and landside operations as well as communities are entangled in networks, which benefit from or harmed by information flows in and out of the system. Although airport infrastructure is a complex web with many interdependencies, enhancing de‑icing capabilities was found as the most critical method to mitigate risk from the more advanced risk analysis techniques.
Regardless of the sunk costs in economic investments for DFW in 2011, disruptions increased in severity over time. Critical system assets have failed from various forms of disruption, due to risk-based efficiency practices. Resilience adoption was a shortcoming, with resilience as the ability to plan for, absorb, recover from and adapt to disruptive events.
Improving resilience by design and intervention at airports
Resilience places emphasis on how the system performs after a disruption (resource requirements and recovery times) and not traditional threat prevention, mitigation and absorption (limiting the magnitude of the initial lost function). This means that the social domain of airport infrastructure operations should connect with infrastructure domains (like airline scheduling at the airports).
In practice and for this purpose, resilience can be achieved by design – in essence, allowing airports to have the ability to absorb disruptions – and by intervention – which involves the security of exogenous resources which strengthen the airport’s operations in the face of disruption (Linkov et al., 2021[68]; Kott et al., 2021[69]).
Resilience by design can be improved on the airside of operations. Furthermore, investments in infrastructure also uphold resilience by design (e.g. building necessary redundancies) in physical, social, or information domains. For example, the Integrated Operations Center at DFW Airport (see the case study above) provides the sharing of data with analytics to inform decision making (DHS, 2017[70]). Other metrics for resilience prompt decision makers at the airport to forecast threats, benefitting real-time operations.
Resilience by intervention in the context of airport operations is reliance on community and/or landside‑based relationships. Resilience by intervention leverages operations and efforts toward environmental mitigation within the airport to meet or even exceed regulatory requirements. Regulatory compliance and cost savings are balanced through resilience by intervention. One example drawn from the case study on DFW Airport (outlined above) was the airport’s investment in hotel infrastructure. This strengthened the capacity of the community near the airport to receive and provide resources, contracts, and facilities, thereby lessening the impact of disruptions.
2.2.3. Case study: Texas deep freeze – energy systems and severe weather
The February 2021 polar vortex resulted in a shock in supply and demand in energy. In Texas, disruptions to energy contributed to residential energy blackouts and electricity price increases of 100 times for those still capable of operating household electricity (Jin et al., 2021[71]). Limited energy access and blackouts were triggered by temperatures far colder than the operating conditions of the available infrastructure.
Designed for maximum efficiency to reduce producer and consumer price burdens during expected operating conditions, the Texas electricity grid does not typically maintain backup storage (Douglas, 2021[72]). While limited natural gas reserve capacity is not uncommon for thermal power generation plants, consumption patterns in Texas placed the state in a more precarious condition – it consumes 14% of the total US natural gas but has only 8% of the total US storage field. With state‑wide natural gas production declining by over 10 billion cubic feet per day during the February 2021 polar vortex, the lack of redundant capacity along with limited recovery planning for anomalous severe weather events disrupted energy provision to homes and businesses. It also required substantial time to fix what would otherwise not be experienced in other parts of the United States. The loss of available generating capacity shifted the Electric Reliability Council of Texas into emergency operating conditions, cutting power to 2 million homes during the freezing temperatures.
Complicating matter was the configuration of the Texas electricity network. Self-enclosed, the state electricity grid system is disconnected from other energy networks in the United States. While this allows the Texas electricity system to fluctuate operations and reduce costs to customers, it left it with fewer options to rely upon power generation and transfer from unaffected states within the same energy provider region. Likewise, the self-contained structure of the Texas energy network reduced federal regulatory oversight, which may identify possible disruptions and/or co‑ordinate recovery when disruptions occur.
The brittle, isolated, and self-contained nature of the Texas energy grid had drawn warnings from the United States’ National Academy of Sciences and other experts, particularly due to its vulnerability to cold weather and the limited adaptive capacity during emergencies at the local level (Jin et al., 2021[71]). The deregulated Texas energy market envisaged energy companies competing to deliver power to consumers through the common electricity grid. The idea was that this would facilitate new market entrants, enhance competition, and incentivise innovation to further benefit consumers. The state was supposed to intermediate between producers and consumers, with prices rising in the event of high demand. Unfortunately, electricity demand is inelastic – it does not respond to changes in price but does respond to changes in weather (Jin et al., 2021[71]).
The severe demand shock during tfhe polar vortex, coupled with the isolation of the electricity grid, non-winterisation of critical facilities and supply chain failures, meant that the state of Texas and other system stakeholders could not respond quickly to the disruption. Although the Texas electricity grid was minutes away from irreparable network damage, the decision to alleviate 5 gigawatts minimised strain on the system (Blunt and Gold, 2021[73]). This decision, however, came at a steep cost. Limited winterisation strategies and a lack of planning for resilience against weather anomalies resulted in fatalities, disrupted transit, and substantial local and state‑wide direct and indirect economic losses.
Implementing strategies to improve the resilience of energy systems
A reliance upon extreme efficiency in energy network design and operations is not unique to Texas (see the case study above). Other US and European energy and utility systems have also experienced the consequences of weather disruption. Non-weather-related energy network disruptions also occur, including the software and operator errors that sparked the 2003 United States’ Northeast blackout, and the synchronised multi-target cyberattack that caused power outages in Ukraine in 2015. As disruptions to interconnected networks and energy stakeholders are expected to increase, application of resilience in energy systems is paramount to ensure this critical function is delivered.
Efficiency is an incentive towards energy generation, transit, and consumption. It has been a guiding engineering principle in energy systems – redundancy costs money and resources that, when not being used, have less perceived value than if they had been deployed elsewhere. Supposedly efficient systems are, however, subject to increasing disruptions spurred by climate change, the potential risks of digitalisation, and growing interconnections between natural and human-made systems.
Efficiency, however, must also accommodate anomalous events that could degrade or destroy an infrastructure network. Resilience emphasises the capacity of a system to recovery and adapt to disruption. Risk management, in contrast, emphasises the planning and absorption of threats, placing less emphasis on cyclical recovery and post-disruption transformation. Resilience internalises the notion that, because disruptions are inevitable, a system should be designed not only to mitigate the risk of a disruption, but to recover and adapt when disruptions occur. Thus, resilience extends efficiency-based thinking from optimising resources for normal operations to optimising resources for anticipated and unanticipated disruptions, leveraging the capabilities of either regulated or deregulated energy networks.
Although operators, owners and other stakeholders favour efficiency in energy systems, resilience strategies have been implemented successfully, even if in an ad hoc manner. Regions that experience more frequent or regular disruptions have backup natural gas, coal or other supplies for power generation. Grid managers may forecast system maintenance in co‑ordination with other external and internal drivers such as anticipated system demand, weather events or geopolitical events (e.g. Brexit, new pipelines, solar technology, etc.) that may affect supply, distribution and production. Fundamentally, the energy system design itself – through smart microgrids, modular systems for reorientation and localisation of disruptions, and adoption of advanced automation by distribution utilities – provides ample opportunity to leverage resilience strategies.
Resilience thinking leverages both stress testing and network science of the full system to allow for effective decision making to maximise resilience and efficiency to the greatest benefit to society (Linkov et al., 2022[74]). A network model of the entire system of the energy grid can be crafted from known data points and/or inferred through machine learning and AI techniques where system visibility is lacking. Using models, resilience analysis identifies weak points in the system by stress testing the network and the intricacies, complexities and interdependencies of the system. The results can be used to determine necessary corrective actions and policies from within the specific energy system to prevent degradation of critical functions, post disruption.
2.3. Finance
Research on financial systems primarily focuses the discussion of resilience on preparation for and absorption of shocks. As with infrastructure, the financial sector discusses resilience from multiple perspectives and disciplines, such as supply chain management, organisational management and economics (Anderies, Janssen and Ostrom, 2004[75]; DesJardine, Bansal and Yang, 2019[76]; Jüttner and Maklan, 2011[77]; Plummer and Armitage, 2007[78]). While these perspectives may deviate from each other, the concept of resilience is maintained: resilience is a strategy of thinking through a system’s dynamics to prepare for and absorb a financial loss.
Measuring resilience of financial systems stems from the Dodd-Frank Wall Street Reform and Consumer Protection Act (US Congress, 2010[79]). In response to global financial crisis of 2007‑08, these tools aimed to operationalise resilient behaviour through better planning and absorption via stress testing. A stress test for financial systems is a host of “what-if?” scenarios, where critical variables are hypothetically tracked for performance in response to theoretical stimuli. The lessons learnt from the stress test are published to provide a status of the health of the financial system, and to announce changes that need to be made.
Various banking agencies perform stress tests and publish the results (Federal Reserve Bank of Minneapolis, n.d.[80]). While the methodology is often scenario based, variables are classified in an interconnected model to assess overall market risk. In other words, stress testing the financial sector is a strategy for supervising market activity to ensure more resilient behaviour (European Central Bank, 2021[81]; Levy-Carciente et al., 2015[82]). Several strategies have been created by researchers to simplify this process, such as the Basel stress testing scenarios, which were used routinely on the US bank system after the 2008‑09 recession, resulting in several revisions to the methodology over the years (Heyen, 2008[83]; Jokivuolle, Virolainen and Vähämaa, 2008[84]; Miu and Ozdemir, 2008[85]). While most resilience and stress testing exercises emphasise system performance within financial institutions, the underlying assumptions (and scenarios) are often predicated on substantial shifts in individual and community behaviour, spending patterns and other considerations that affect the economic performance of countries.
Linking back to the Linkov et al. (2018[35]) tiered methodology (see Section 2.2), the vast majority of financial resilience assessments through stress testing operate in Tier 3. Advanced mathematics and an array of simulations are used to track system performance. In essence, the network of the financial system is analysed to discover where individual variables must be hardened. As stress tests seek to simulate probable, improbable and even uncharacterised shocks for finance (Berkowitz, 1999[86]; Foglia, 2008[87]; Geithner, 2014[88]), their focus answers the question “which elements of the system must be bolstered?” This correlates with the definition of resilience by identifying how systems can better plan for and absorb a shock, in the same way that a financial system may plan for and absorb an economic downturn.
2.4. Environment
Environmental resilience differs from built infrastructure resilience, in that environmental management stakeholders are not managing a “purposeful” system. Unlike physical infrastructure, which has designed operating requirements and service delivery capabilities, natural systems are formed, adapted and reformed over extensive cycles. They lack a predetermined purpose and achieve local equilibrium based on the balance of organisms living within them (Angeler et al., 2016[89]).
Environmental resilience is often reviewed through the lens of an area (rather than through common theories), such as climate resilience, air and water systems, and various ecosystems (e.g. tropical, oceanic, etc.). For example, the OECD continues to be active in efforts to strengthen the resilience of human and natural systems to the impacts of climate change (OECD, 2021[90]), including outlining a way forward for defining, measuring and mobilising adaptation-aligned finance (Mullan and Ranger, 2022[91]). These areas are beyond the remit of this chapter, meriting separate exploration.
The European Commission began exploring resilience analysis in 2013 when it called for proposals within the Horizon 2020 programme Disaster-resilience: Safeguarding and Securing Society, Including Adapting to Climate Change. This tasked EU Member States with building in principles of resilience analysis with the purpose of managing risks at the system level. Climate change was the primary driver for this resilience initiative, but the Commission also mentioned other societal disruptors, such as terrorism and unforecastable infrastructure threats. The European Commission noted that “a better understanding of critical infrastructure architecture is necessary for defining measures to achieve a better resilience against threats in an integrated manner including natural and human threats/events” (European Commission, 2015[92]). In this respect, the European Commission sought to promote a holistic approach by using resilience to address interdependent systems across Europe, recognising threats that are both purposeful (e.g. terrorism) and accidental (e.g. natural hazards) could affect the continent, if unprepared.
In the United States, the Environmental Protection Agency (EPA) mentions resilience as “the capacity for a system to survive, adapt and flourish in the face of turbulent change” (Fiksel, Goodman and Hecht, 2014[93]). With this strategy, the EPA seeks to lessen risk by increasing the preparedness of systems facing external disruptors, even when in-depth knowledge of such threats may not exist or is incomplete.
Although resilience analysis is still evolving in the environmental field, the EPA discusses its ability to provide “consideration of a system’s capacity to withstand even unforeseen disturbances”, which is necessary for an increasingly complex, globalised world with interdependencies that are both easily seen and hidden (Fiksel, Goodman and Hecht, 2014[93]). The EPA identifies that resilience analysis could be useful in guiding changes to policy in areas with challenges to their systems’ resilience plans. However, it also mentions shortfalls and concerns about the method, including a lack of a governance structure to measure and characterise what resilience means.
The EPA identifies several possible methods of resilience analysis, contending that no “one size fits all” approach can apply to every scenario. It mentions five characteristics for evaluating resilience. These characteristics are diversity – known as “the existence of multiple behaviours within the system”; adaptability – “the capacity of the system to change in response to new pressures”; cohesion – “the strength of unifying forces”; latitude – “the maximum amount of change the system can absorb”; and resistance – “the capacity of the system to maintain its state in the face of disruptions” (Linkov and Trump, 2019[94]). Given these characteristics, the EPA notes that, where information is limited, a more quantitative assessment of resilience could be conducted. A close review of each of the five characteristics may provide input about the overall resilience of the system and its abilities to protect itself from crippling external forces. The EPA concludes that qualitative resilience analysis may be a useful starting point for further understanding the threats and uncertainties of a system’s individual applications of resilience.
While the EPA has not formally adopted resilience analysis, some of its units apply these principles. For example, the EPA’s Office of Water has created a Climate Resilience Evaluation and Awareness to give users the ability to “anticipate potential impacts of climate change to drinking water and wastewater utilities” (Fiksel, Goodman and Hecht, 2014[93]). The EPA’s Gulf of Mexico Programme seeks to guide coastal communities to identify relevant environmental threats and boost their resilience to these disruptors.
To help communities identify the extent of their preparedness for coastal storms, the Sea Grant Mississippi and Alabama offices of the US National Oceanic and Atmospheric Administration (NOAA) built a Community Resilience Index. This Index is meant to identify areas of resilience performance both during and after a coastal storm. This is divided into six sections, asking yes/no questions to gauge resilience (such as services for transportation or even critical infrastructure). Checkmarks are tallied in the Index, resulting in subjective scores that are low, medium or high, pertaining to a jurisdiction’s ability to bounce back from a coastal storm. Using this approach, leaders in communities can easily identify areas of weakness that need to be improved. The NOAA has sought to integrate stakeholder feedback in these analyses, inviting community leaders and members to find holes in the resilience plan (Emmer et al., 2010[95]). This co‑operation serves as an example of how a government can co‑operate with local areas to address long-term disruptions and potential risks to environmental resilience (Murphy et al., 2014[96]).
2.5. Cybersecurity and digital systems
In a world increasingly affected by the Internet of Things, cybersecurity and digital systems are intertwined with the resilience of every other sector mentioned in this chapter (see the chapter on digital foundations). How might cyber systems not only be used but required for the successful operation of this sector? For example, what might happen to a city’s electricity grid or energy supply if malicious actors hack into the digital system?
These are not theoretical questions. For example, in 2021, a major gas pipeline on the east coast of the United States had its digital systems hacked and a feedback loop was created in the gas supply chain that caused prices to surge, leading to the Colonial Pipeline gas shortage (Jin et al., 2021[71]). As the malicious actors behind the hack in the system requested pay-outs, consumers began panic-buying gasoline as the Colonial Pipeline stopped distribution for several days. Gasoline/petrol stations went dry up the eastern seaboard of the United States, inflating prices significantly. This example shows why resilience is necessary in cybersecurity and digital systems.
The surrounding literature explores all stages of resilience: preparing for, absorbing, recovering from and adapting to disruption. For example, Kott and Linkov (2019[97]) and Linkov et al. (2013[19]) discuss the need to use this holistic approach to resilience for cyber systems. Depending on the application, however, other instances focus more on the earlier stages of resilience (preparing for or absorbing shocks) or later stages (recovering from or adapting to shocks).
Research into cyber systems that focuses on preparing for and absorbing disruption connects to exogenous systems that are “too big to fail” – in other words, where disruption can cause catastrophic consequences based on geography, the socio-political context and other variables. For example, researchers note cyber resilience as a property of systems to overcome adverse events (Arghandeh et al., 2016[98]; Björck et al., 2015[17]; Harris and Impelluso, 2008[99]) or intelligent systems, energy grids and financial systems. The goal of higher-order planning and absorption is to blunt or even avoid disruption entirely, since the recovery and adaptation stages still place consumers or users in a state of distress.
Conversely, research that centres discussion of resilience around overcoming and adapting to disruption links to endogenous cybersecurity. When a hack or disruption occurs, such as with the Colonial Pipeline, cyber experts attempt to restore the system’s integrity and defend against similar attacks as quickly as possible. Cyber-based research surrounding cybersecurity focuses on this concept, including the definition of resilience for cyber systems to withstand, recover from and evolve from disruption (Carías et al., 2020[100]; Gisladottir et al., 2017[101]). From this perspective, the importance of defending the system is recognised, but the view is taken that disregarding the potential for malicious actors to outsmart the system is naïve.
Measuring the resilience of cyber systems similarly follows the Linkov et al. (2018[35]) tiered framework that was associated with infrastructure systems. Tier 1 evaluations apply CISA’s tabletop exercise routine to cyber systems. As with infrastructure, these tabletop exercises characterise the system and coarsely identify areas of improvement for its ability to plan for, absorb, recover from and adapt to adverse events. For Tier 2, the cyber domain is a central consideration of the Fox-Lent, Bates and Linkov (2015[21]) resilience matrix, which argues that cyber systems are a requirement to classify the level of resilience for a given system. Tier 3 utilises essences of AI to promote resilience by design. For example, a Tier 3 tool could be antivirus software that has elements of resilience baked into the system by being able to identify potential threats, quarantine suspicious agents and remove malware in tenths of a second.
2.6. Disaster and emergency response
The US National Academy of Sciences has characterised four stages of resilience in disasters (National Research Council, 2012[2]). The first stage (planning/preparations) assesses the operational health of the system to identify where resources and/or services are required in the face of an unknown threat. The second stage (absorption) involves assessing the threat to maintain critical functions while managing the disaster. The third stage (recovery) assesses motivation to get the system back to pre‑disaster operational capacity as quickly as possible. The last stage (adaptation) takes the lessons learnt from the disaster to make necessary changes to the system in the event of future disruptions of a similar class and magnitude.
Disaster resilience analysis often focuses on the first stage – preparation/planning. This may be due to the tendency of resilience analysis experts to think prospectively about theoretical scenarios, prepare for them and analyse how similar events were responded to in the past. This maintains more conventional risk analysis methods, which prepare systems at risk of identified disruptions. Such conventional methods of risk analysis focus on optimising available resources to respond to potential disasters with maximum protective capabilities. More conventional risk analysis lacks the ability to protect against high-consequence, low-probability events that are unforeseen or unidentifiable, complicating the governance requirements around recovery from them.
Research on the second and third stages – absorption and recovery – is increasing in publications and regulatory use because it begins to bridge the gap in overcoming risk present in common methods. The absorption stage focuses on maintaining the integrity of the system structurally and functionally both during and immediately after a disaster occurs. The recovery stage focuses on minimising the time needed to get the system back to being fully online and operational after the disruption.
Absorption is the current focus of governmental agencies in the United States. At least seven (the Department of the Interior, Department of Homeland Security (DHS), National Institute for Science and Technology (NIST), the EPA, US Army Corps of Engineers, NOAA and US Army Environmental Command) have adopted metrics to estimate the ability of resilience thinking to improve risk-based approaches to systems of interest. Tools of interest to review an asset’s ability to absorb disruption include relative and standardised scoring functions, using quantitative and qualitative information, based on the method selected. Qualitative reasoning, however, tends to dominate in this field.
Like absorption, recovery (the third stage) is still underutilised in conventional risk analysis, but it is in nascent stages in the literature and government reports. While absorption focuses on preparing for a disruptive event, recovery focuses on events that occur after the disruption. While traditional risk analysis places some importance on reducing the time to recovery, this is not held in as high regard as it is in resilience thinking (Linkov et al., 2014[28]). Due to this difference from traditional risk management, recovery operations and policies tend to not secure the same level of attention or funding from government agencies as the first two stages. Nonetheless, more focus is being placed on this stage as the field advances.
Lastly, adaptation (the fourth stage) presents the largest difference from more conventional risk analysis methods. This departure is the result of this stage’s intent to change the system from an infrastructural and/or organisational level to improve its ability to absorb and recover from future events. Adaptation is most similar to adaptive management, which is a decision-making process where norms and standards are changed through time to face shortcomings in systems (Linkov et al., 2006[102]; Stankey, Clark and Bormann, 2005[103]). Guiding tools to convey how systems should build for adaptation are rarely available. This stage is also the least reflected on by, for example governmental agencies in the United States – only the EPA, DHS and NIST included it in their resilience efforts in 2015.
Bakkensen et al. (2017[104]) compared five metrics that reported a level of resilience assessment to inform communities about how to recover from various disasters. While the indices they created used similar datasets, they were not always internally valid. For example, if one index reported better resilience for one county than another, others did not always agree. For local stakeholders and decision-makers who need these indices for decisions on policy or investment in the locality, the problem has intensified, contributing to concerns about for whom a resilience‑based approach should be created; and what metrics or system components should be benchmarked as “resilience performance” for affected stakeholders.
In the United States, the Federal Emergency Management Agency (FEMA) is a key player in responding to and recovering from disasters and natural hazards. It is, therefore, crucial to resilience plans. FEMA began efforts to recommend and create certification programmes in the light of the goals for resilience of the DHS. For example, FEMA publications include National Preparedness Goal and National Disaster Recovery Framework (Federal Emergency Management Agency (FEMA), 2011[105]; Federal Emergency Management Agency (FEMA), 2015[106]). In the latter, FEMA creates checklists for before and after disasters for certain stakeholders (such as individuals and families and the non-profit sector). These checklists give recommendations, but not a quantified framework for assessment. FEMA also created a private sector accreditation for preparedness and a certification programme to grant recognition to organisations that apply DHS consensus-based standards for preparedness and best practices in the field. In general, these efforts take a system-level view of resilience.
The National Institute for Science and Technology (NIST) in the United States has a disaster resilience framework (Larkin et al., 2015[107]). This is designed to boost resilience for communities in the face of both natural and man-made disasters. Natural disasters include hurricanes, storms, earthquakes, tornadoes, floods, landslides, wildfires, tsunamis and excessive rain or snow; man-made disasters include vehicular impacts, blasts and many others. This assessment provides metrics for performance that should be attained in a limited amount of time, and lists goals to address disasters before, during and after they occur. These goals can be adapted by community leaders to apply more specifically, ensuring that all members of the community and relevant organisations benefit. The NIST’s framework divides resilience into three stages: the response phase (0‑3 days after the event), the workforce/neighbourhood recovery phase (1‑12 weeks after the event) and the community recovery phase (4 to 36+ months after the event). These complement the absorption, recovery and adaptation stages respectively (Linkov and Trump, 2019[94]; National Research Council, 2012[2]). The response phase focuses on aiding community members before and during a disaster, focusing on the critical needs outlined by the NIST: food and water resources, life safety, health, shelter and situational awareness. The workforce/neighbourhood recovery phase addresses metrics for performance that allow a community to recover from a disaster efficiently and quickly. The community recovery phase entails long-term reconstruction of community infrastructure and organisation, including resilience.
2.7. Medicine and public health
Currently, there is no metric for health resilience, at either an individual or institutional level. Clinicians have focused on incorporation of biomarkers such as musculoskeletal changes, stem cell changes, serum markers, metabolic markers, hormonal changes and new inflammatory markers (Al Saedi et al., 2019[108]). More recently, epigenetics and genetic research is working to provide resilience indicators in health across an individual’s lifespan. Additionally, social determinants of health are used, collected from socio‑economic data as well as self-reported data and surveys for the collection of data specific to resilience. Sample sizes used in these studies are, however, typically rather small, and the surveys used are not the same, with varied questions and outcomes measured (Klasa et al., 2020[109]).
Health, however, is not just a product of individual genetics. Exogenous factors like the perception of race, the physical environment, poverty, and education may all affect outcomes for health of a person. As time continues, these factors can transform to play larger roles, which indicates that they are critical to system stability and service delivery. Five spheres influence an individual: physical activity and active living; individual determinants (genetic and behavioural); the social environment; the built environment and the natural environment (Table 2.1). These spheres directly affect factors that determine health behaviours, including institutional factors, community factors, public policy (i.e. governance and law-making), intra-personal factors and inter-personal processes (Mcleroy et al., 1988[110]). Owing to this complexity, resilience in health cannot completely avoid risk.
Table 2.1. Determinants of health
Resilience is a dynamic process embedded within many systems of interactions – it is not an individual trait or characteristic. An individual is limited in how much they can adapt to a threat because many aspects of well-being are beyond their control. Without interventions to influence social interactions, environmental structures and health resources, resilience cannot be achieved at either the individual or the system level. An individual requires access to health care services, safety, social support and adequate education to optimise their capacities over their lifespan (Hayslip and Smith, 2012[111]). This provides a baseline from which multi-level strategies for health resilience can be designed and implemented.
Aggregating these factors could provide a resilience quantification for a specific profile of a person. A measurement that encompasses all these factors may not, however, adequately characterise the individual’s response to a threat within the different domains (Table 2.2). Health determinants will be affected by disrupting the social environment, such as closing cafes as gathering places, but will not be aided by disaster insurance. Meanwhile, the impact of a disruption to the natural environment, like a cyclone, could disrupt all systems. A person resilient to one type of disruption may be extremely vulnerable to another. Resilience can be measured as resilience to specific events that trigger changes.
Can an individual show overall resilience? If resilience is framed as access to redundant resources, tight social connections can greatly expand an individual’s available resources. A storm that destroys multiple aspects of the built environment and cuts the individual off from social activities, resulting in bodily or cognitive harm, may damage the individual far less when a concerned neighbour with a car takes the time to check in. Social connections are critical to individual resilience. Social connections also are capable of mobilising resources at relevant scales more quickly than the built environment. Thus, resources provided through an expansive social system may amount to redundancies in all other systems.
However, the flexibility that allows quick responses from social support arises because those supports are not structural and maintaining them in the long-term may be beyond the social network capacity (Cohen and Syme, 1985[112]). In studies of resilience to multimorbidity, the time during which support is available matters (Klasa et al., 2020[109]; Wister, Klasa and Linkov, 2022[113]). Different social or structural environments affect resilience over different time periods. This can lead individuals to be resilient to some circumstances but not others, and resilient in the longer term but not in the short term.
Perspectives on health system resilience often change over time. For example, demographic changes worldwide are leading to more adults seeking social support from relatively fewer able‑bodied younger relatives and friends. In Japan, the role of the built environment in ensuring health and welfare is expected to increase, for example, with a focus on in-home smart appliances that evaluate health problems to help elderly people get needed medical support earlier. Whether capability improvements in one determinant of health can decrease the need for or replace another remains to be seen, and such insight would contribute to understanding how overall resilience could be best quantified for individuals.
Table 2.2. Selected definitions of resilience in medicine
Source: Klasa (2020[109]), “Science and Practice of Resilience: Disaster Systems Applications to Aging Resilience”, https://doi.org/10.1007/978-3-030-57089-7_4.
Public health and epidemiological resilience share some commonalities with medical science, although it is less focused on diagnosis and treatment of disease, and more on how social and environmental systems can be safeguarded or adapted to overcome stressors. Fundamentally grounded in a systems-based approach, public health resilience emphasises the interdependent nature of individuals, communities, governments, environments, economies and infrastructure, all of which contribute to and determine health outcomes. While there are many possible applications – such as environmental contaminants, air and water quality and so forth – one salient area of public health resilience is responding to human pathogens.
2.7.1. Case study: Ebola virus (2014)
Human pathogens trigger a dramatic system response whenever they arise. Directly, illness, long-term health debilitation and mass fatality can overwhelm health systems if they not adequately prepared for (see chapters on critical care surge and care continuity). Indirectly, socio‑economic disruptions can result, as government and philanthropic institutions may struggle to deliver various services given increased demand (e.g. food, energy and heating services, housing assistance, education, and many others).
Ebola virus initiates a painful haemorrhagic fever which prompts mortality rates at an average of 50% or greater (Pourrut et al., 2005[126]). There is no remedy beyond oral hydration therapy, leaving many unlikely to survive its cascading health impacts. The virus was first described in 1976 in villages along the Ebola River. Until 2013, the disease was mostly sequestered in sub-Saharan Africa, where approximately 24 recognisable outbreaks contributed to 1 716 cases between 1976 and 2013 (Dixon and Schafer, 2014[127]).
Survivors of Ebola usually encounter challenges with reintegrating into society. In the face of social, health, and financial problems after initial recuperation from the disease, survivors are typically left vulnerable as the attempt to fully recover and stabilise normal life. Typical long-term effects include a range of ailments, such as: muscular pain, liver inflammation, fatigue and long-term weight loss, all placing the survivor’s overall health in question (Magill et al., 2013[128]; Tosh and Sampathkumar, 2014[129]). The World Health Organization (WHO) (2015[130]) notes that survivors require strict monitoring – which is typically unavailable in sub-Saharan Africa – for years after recovery to ensure that other complications do not arise. Long-term problems with health are worsened by economic and social factors that surround the patient’s recovery. For example, economic hardship has been a result for many survivors due to social discrimination preventing occupational pursuits (Lee-Kwan et al., 2014[131]; Levin-Sparenberg et al., 2015[132]; Curson, 2015[133]). In this respect, the Ebola virus indirectly weakens social resilience.
Resilience thinking is needed for the Ebola response. Strong consequences would follow if the Ebola disease spread to larger population centres, in terms of health, social order and trade. While a proliferated Ebola outbreak has yet to happen, the 2013‑16 West African Ebola outbreak exemplified the virus’s strength and the high consequential impacts of failure to maintain disease response and control. As a result of this outbreak, cases reached Europe and the United States from the primary host countries: Guinea, Liberia, Nigeria, and Sierra Leone. Even though fatality rates from this outbreak are estimated to hover around 40%, the real numbers are trickier to ascertain due to inefficient methods in accounting for the disease incidence and because some afflicted individuals refused to receive medical assistance.
Health workers also face an additional challenge with keeping safe from the disease, given that their roles require them to be proximate to infectious bodily fluids. In 2014, an estimated 10% of confirmed Ebola cases derived from health care workers, demonstrating the complexity of ongoing disease monitoring and treatment efforts. Likely causes include few trained staff during the onset of the disease, poor supplies, and a strong reliance of hastily constructed field hospitals. This should not denigrate or overlook the staff that have treated the thousands of cases of Ebola during this time – these professionals should be recognised for their courage in providing medical care in such risky conditions. Rather, this Ebola outbreak serves as an example of how governance for risk and disease in West Africa was inadequate to efficiently address the outbreak. Researchers have also found economic disfunction, social disunity, and a mistrust of public health authorities in the local populations flowed indirectly from this outbreak (Bonwitt et al., 2018[134]; Brown et al., 2017[135]; Massaro et al., 2018[136]).
Risk management would focus on building the individual pieces of how Ebola may propagate through countries. Instead, resilience classifies an observable range of components that are useful to quickly recover and adapt from the outbreak. For this particular disease, improving resilience may include adaptive governance to provide a scaled medical response, as well as to change medical protocols with the focus upon minimising contact between healthy and infected populations. An example includes the provision of adaptive methods for airport security, passenger biocontainment, and air traffic control to prevent the spread of the disease nationally and internationally. The US Center for Disease Control and Prevention (n.d.[137]) has provided guiding measures for medical air transport for affected patients to prevent the spread of the Ebola virus. Massaro et al. (2018[136]) identified that populations susceptible to outbreaks can be modelled using network science to chart the path an outbreak would take as well as its virulence, providing policy recommendations. The authors discovered that, while a risk-based strategy (like shutting down transportation systems between infected countries) seems logical, more flexible strategies to contain and mitigate the outbreak to lessen the impact of it would be medically, economically and socially preferrable.
Resilience thinking regarding a disease that was previously not determined to spread would also improve response. For example, Ebola’s rare occurrence in West African history was an early difficulty for its identification by medical professionals in the first months of an outbreak (Baize et al., 2014[138]). Resilience to the Ebola virus must use an innovative, robust and interconnected approach that uses all reachable governmental support systems to alleviate future outbreaks.
2.8. Differences and similarities across sectors
There are differences in how sectors characterise resilience. These centre around the purpose and focus of the sector itself. For example, the finance sector is primarily concerned with preparing for and absorbing economic disruption, whereas the infrastructure sector tends to value recovery and adaptation to get systems back operating at capacity. While acknowledging the importance of the holistic definition of resilience, researchers within these sectors use differing narratives to approach resilience.
One key similarity across sectors, including health systems, is vocabulary to describe resilience. At the sector level, resilience is an endogenous property of each system to withstand and overcome disruption. Withstanding disruption tends to be risk-focused, operationalising resilience through risk assessment practices to prepare for and absorb disruption. Overcoming disruption tends to be resilience‑focused, quantifying the potential risk while also attempting to boost the system’s ability to recover from and adapt to the disruption. Preparing, absorbing, recovering and adapting are common terms in use across sectors, even if some focus on specific stages more than others (Table 2.3).
Another similarity across sectors is that there is no “silver bullet” to quantify resilience. While resilience encompasses risk analysis through preparing for and absorbing disruption, even the risk science that researchers use generally does not converge on an optimal method. New methods are still being discussed for recovery from and adaptation to disruption. The literature reveals more divergence than convergence on a single method. Researchers, however, tend to agree that quantifying resilience is complex because all systems are different. Qualitative tools that operate through rules of thumb and discussion prevail. Quantitative tools, such as stress testing, have been developed but are still nascent.
Table 2.3. Breakdown of resilience across sectors
Note: Darker colouring indicates stronger focus.
Furthermore, the methods used to quantify resilience map onto the Linkov et al. (2018[35]) framework of a tiered approach to resilience analysis, with increasing analytical rigour and fidelity of results from Tier 1 to Tier 3. The field of resilience still focuses primarily on Tier 1 assessments of scorecards and tabletop exercises, even though Tier 3 tools have been developed. Tier 1 qualitative assessments are valuable but lack data-driven objectivity. Nonetheless, stricter quantification of resilience is increasingly evident.
A final similarity is the top-down approach in governance to employ resilience for these sectors. National and international actors have responded to shocks, including the COVID‑19 pandemic, by calling for more resilient thinking within relevant systems. These actors include the OECD, the World Health Organization and the World Bank, among others (see Section 2.1).
2.9. Conclusions: Lessons learnt for health system resilience
The key lesson learnt from analysing other sectors is that health systems should apply the same tenets that underlie the characterisation of resilience. The basic principles of resilience are: preparing for, absorbing, recovering from and adapting to adverse events. The question that follows is “what components of resilience are most crucial for health systems?”.
To address this, lessons learnt from both finance and infrastructure sectors can be adopted. In finance, the global financial crisis (2007-08) was the impetus for resilience. As a result, financial systems have operationalised resilience by seeking to prepare for and absorb shocks, especially through stress testing. In the infrastructure sector, disruptions that place transportation networks or energy facilities offline cannot be effectively engineered against. Accordingly, experts focus resilience efforts on recovering from the disruption, and adapting their systems to address future threats.
Health systems can – and should – take a similar approach. In health systems, however, there can be no stage of a disruption cycle left unaddressed. Recovering and adapting from shocks is as important as preparing for and absorbing them. Since each stage of the disruption cycle is equally important and the interactions between these stages are dynamic, one of the outcomes of a resilient health systems should be the movement of critical and scarce supplies and staff for greatest value‑added use.
This report seeks to contribute to this ongoing effort by identifying weaknesses in each of the four stages of the disruption cycle and recommending policy responses (see the chapter on key findings and recommendations). All resilience analysis requires stakeholders to adopt a multi-systems view. In the case of the COVID‑19 response, what appeared to be a sensible policy of reducing elective procedures to preserve capacity for patients with COVID‑19 yielded immense downstream consequences. The health workforce has been left with a heightened burden of care for the foreseeable future, reducing the resilience of health systems (see the chapter on waiting times).
Therefore, health systems should be cognisant of interdependencies when measuring resilience. Tier 2 and Tier 3 tools, such as those used in the finance sector, provide a sound example of the interplay of systems. For example, a financial sector stress test places random stimuli on the system to witness how individual links respond. The health system would benefit from similar analysis of interconnections, such as the links between workforce, available resources and available beds. To make optimal use of an analysis of interconnected variables, however, there must be an effective governance structure around decision making to foster adaptability.
Governance for resilience in other sectors has operated on a top-down basis. Accordingly, health systems should take this into consideration. If policy makers set the vision but do not characterise what achieving the desired outcomes looks like, then noble efforts are for naught. Usually, success in improving resilience requires collaboration between multiple stakeholders, including government, industry, communities and others. Conversely, if a clear vision is not set, objectives are difficult to quantify. Thus, a structured governance agenda for resilience in health systems must be established that sets a clear vision while also bringing together the actors within the system to strive for the shared goal.
References
[121] Ager, A., J. Annan and C. Panter-Brick (2013), Resilience: From conceptualization to effective intervention. Policy Brief for Humanitarian and Development Agencies 1, http://www.resilienceproject.org (accessed on 24 June 2022).
[108] Al Saedi, A. et al. (2019), “Current and emerging biomarkers of frailty in the elderly”, Clinical Interventions in Aging, Vol. 14, p. 389, https://doi.org/10.2147/CIA.S168687.
[7] Alberts, D. (2007), Agility, Focus, and Convergence: The Future of Command and Control, Office of the Assistant Secretary of Defense Washington DC Command and Control Research Program (CCRP), http://www.dodccrp.org (accessed on 24 June 2022).
[5] Alberts, D. and R. Hayes (2003), Power to the Edge: Command and Control in the Information Age, Office of the Assistant Secretary of Defense Washington DC Command and Control Research Program (CCRP).
[58] Allen, W., D. Liu and S. Singer (1993), “Accesibility measures of U.S. metropolitan areas”, Transportation Research Part B: Methodological, Vol. 27/6, pp. 439-449, https://doi.org/10.1016/0191-2615(93)90016-4.
[115] American Psychological Association (n.d.), Resilience, https://www.apa.org/topics/resilience (accessed on 24 June 2022).
[75] Anderies, J., M. Janssen and E. Ostrom (2004), “A Framework to Analyze the Robustness of Social-ecological Systems from an Institutional Perspective”, Ecology and Society, Vol. 9/1, p. 18, http://www.ecologyandsociety.org/vol9/iss1/art18 (accessed on 24 June 2022).
[89] Angeler, D. et al. (2016), “Panarchy use in environmental science for risk and resilience planning”, Environment Systems and Decisions, Vol. 36/3, pp. 225-228, https://doi.org/10.1007/S10669-016-9605-6/FIGURES/1.
[98] Arghandeh, R. et al. (2016), “On the definition of cyber-physical resilience in power systems”, Renewable and Sustainable Energy Reviews, Vol. 58, pp. 1060-1069, https://doi.org/10.1016/J.RSER.2015.12.193.
[141] Argyroudis, S. et al. (2022), “Digital technologies can enhance climate resilience of critical infrastructure”, Climate Risk Management, Vol. 35, p. 100387, https://doi.org/10.1016/j.crm.2021.100387.
[36] ASCE (n.d.), Bridges. ASCE’s 2017 Infrastructure Report Card., https://infrastructurereportcard.org/cat-item/bridges/ (accessed on 4 June 2020).
[138] Baize, S. et al. (2014), “Emergence of Zaire Ebola virus disease in Guinea”, The New England journal of medicine, Vol. 371/15, pp. 1418-1425, https://doi.org/10.1056/NEJMOA1404505.
[104] Bakkensen, L. et al. (2017), “Validating Resilience and Vulnerability Indices in the Context of Natural Disasters”, Risk Analysis, Vol. 37/5, pp. 982-1004, https://doi.org/10.1111/RISA.12677.
[13] Berghel, H. (2015), “Cyber Chutzpah: The sony hack and the celebration of hyperbole”, Computer, Vol. 48/2, pp. 77-80, https://doi.org/10.1109/MC.2015.41.
[86] Berkowitz, J. (1999), “A Coherent Framework for Stress-Testing”, SSRN Electronic Journal, https://doi.org/10.2139/SSRN.181931.
[17] Björck, F. et al. (2015), “Cyber resilience – Fundamentals for a definition”, Advances in Intelligent Systems and Computing, Vol. 353, pp. 311-316, https://doi.org/10.1007/978-3-319-16486-1_31/COVER/.
[73] Blunt, K. and R. Gold (2021), “‘A Failure of Texas-Size Proportions’ - State Debates How to Overhaul Its Power Market”, Wall Street Journal, https://www.wsj.com/articles/a-failure-of-texas-size-proportionsstate-struggles-to-overhaul-its-power-market-11618565415 (accessed on 24 June 2022).
[26] Bocchini, P. et al. (2014), “Resilience and Sustainability of Civil Infrastructure: Toward a Unified Approach”, Journal of Infrastructure Systems, Vol. 20/2, p. 04014004, https://doi.org/10.1061/(ASCE)IS.1943-555X.0000177.
[134] Bonwitt, J. et al. (2018), “Unintended consequences of the ’bushmeat ban’ in West Africa during the 2013-2016 Ebola virus disease epidemic”, Social Science & Medicine, Vol. 200, pp. 166-173, https://doi.org/10.1016/J.SOCSCIMED.2017.12.028.
[142] Bostick, T. et al. (2018), “Resilience science, policy and investment for civil infrastructure”, Reliability Engineering & System Safety, Vol. 175, pp. 19-23, https://doi.org/10.1016/j.ress.2018.02.025.
[135] Brown, C. et al. (2017), “What Do We Know About Controlling Ebola Virus Disease Outbreaks?”, Retrovirology: Research and Treatment, Vol. 8, https://doi.org/10.1177/1178123817711115.
[61] Caliendo, C., I. Russo and G. Genovese (2022), “Resilience Assessment of a Twin-Tube Motorway Tunnel in the Event of a Traffic Accident or Fire in a Tube”, Applied Sciences, Vol. 12/1, p. 513, https://doi.org/10.3390/app12010513.
[100] Carías, J. et al. (2020), “Systematic approach to cyber resilience operationalization in SMEs”, IEEE Access, Vol. 8, pp. 174200-174221, https://doi.org/10.1109/ACCESS.2020.3026063.
[137] Centers for Disease Control and Prevention (n.d.), Guidance on Air Medical Transport (AMT) for Patients with Ebola Virus Disease (EVD) | Emergency Services | Clinicians | Ebola (Ebola Virus Disease) | CDC, https://www.cdc.gov/vhf/ebola/clinicians/emergency-services/air-medical-transport.html (accessed on 1 July 2018).
[53] Chang, S. and N. Nojima (2001), “Measuring post-disaster transportation system performance: The 1995 Kobe earthquake in comparative perspective”, Transportation Research Part A: Policy and Practice, Vol. 35/6, pp. 475-494, https://doi.org/10.1016/S0965-8564(00)00003-3.
[41] Chopra, S. and V. Khanna (2015), “Interconnectedness and interdependencies of critical infrastructures in the US economy: Implications for resilience”, Physica A: Statistical Mechanics and its Applications, Vol. 436, pp. 865-877, https://doi.org/10.1016/J.PHYSA.2015.05.091.
[40] CISA (n.d.), CISA Tabletop Exercises Packages, https://www.cisa.gov/cisa-tabletop-exercises-packages (accessed on 24 June 2022).
[112] Cohen, S. and S. Syme (1985), Social support and health, Academic Press.
[59] Çolak, S., A. Lima and M. González (2016), “Understanding congested travel in urban areas”, Nature Communications, Vol. 7/1, https://doi.org/10.1038/ncomms10793.
[18] Collier, Z. et al. (2014), “Cybersecurity standards: Managing risk and creating resilience”, Computer, Vol. 47/9, pp. 70-76, https://doi.org/10.1109/MC.2013.448.
[9] Collier, Z. and I. Linkov (2014), Decision Making for Resilience within the Context of Network Centric Operations, Army Corps of Engineers Vicksburg MS Engineer Research and Development Center.
[46] Comes, T., V. Bertsch and S. French (2013), “Designing dynamic stress tests for improved critical infrastructure resilience”, ISCRAM 2013 Conference Proceedings - 10th International Conference on Information Systems for Crisis Response and Management, pp. 307-311.
[133] Curson, P. (2015), “The Ebola crisis and the failure of governance”, Geodate, Vol. 28/3, pp. 2-4, https://search.informit.org/doi/10.3316/INFORMIT.370967955195235 (accessed on 24 June 2022).
[76] DesJardine, M., P. Bansal and Y. Yang (2019), “Bouncing Back: Building Resilience Through Social and Environmental Practices in the Context of the 2008 Global Financial Crisis:”, Journal of Management, Vol. 45/4, pp. 1434-1460, https://doi.org/10.1177/0149206317708854.
[54] D’Este, G., R. Zito and M. Taylor (1999), “Using GPS to Measure Traffic System Performance”, Computer-Aided Civil and Infrastructure Engineering, Vol. 14/4, pp. 255-265, https://doi.org/10.1111/0885-9507.00146.
[70] DHS (2017), Statement by Secretary Johnson Concerning Airport Security | Homeland Security, https://www.dhs.gov/news/2017/01/11/statement-secretary-johnson-concerning-airport-security (accessed on 24 June 2022).
[10] DiMase, D. et al. (2015), “Systems engineering framework for cyber physical security and resilience”, Environment Systems and Decisions, Vol. 35/2, pp. 291-300, https://doi.org/10.1007/S10669-015-9540-Y/FIGURES/7.
[127] Dixon, M. and I. Schafer (2014), “Ebola Viral Disease Outbreak - West Africa, 2014”, Morbidity and Mortality Weekly Report (MMWR), Vol. 63/25, pp. 548-551, https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6325a4.htm (accessed on 24 June 2022).
[67] Doss-Gollin, J. et al. (2021), “How unprecedented was the February 2021 Texas cold snap?”, Environmental Research Letters, Vol. 16/6, p. 064056, https://doi.org/10.1088/1748-9326/AC0278.
[72] Douglas, E. (2021), Gov. Greg Abbott wants power companies to “winterize.” Texas’ track record won’t make that easy., The Texas Tribune, https://www.texastribune.org/2021/02/20/texas-power-grid-winterize/ (accessed on 24 June 2022).
[23] Ebi, K. and J. Semenza (2008), “Community-based adaptation to the health impacts of climate change”, American journal of preventive medicine, Vol. 35/5, pp. 501-507, https://doi.org/10.1016/J.AMEPRE.2008.08.018.
[20] Eisenberg, D. et al. (2014), “Resilience Metrics: Lessons from Military Doctrines”, Solutions, Vol. 5/5, pp. 76-87, http://www.thesolutionsjournal.com/node/237200 (accessed on 24 June 2022).
[95] Emmer, R. et al. (2010), Coastal Resilience Index: A Community Self-Assessment, Mississippi-Alabama Sea Grant Consortium, NOAA’s Coastal Storms Program.
[42] Environmental Protection Agency (EPA) (2015), Peak Stress Testing Protocol Framework, National Service Center for Environmental Publications (NSCEP).
[81] European Central Bank (2021), Stress test shows euro area banking system resilient under challenging macroeconomic scenario, https://www.bankingsupervision.europa.eu/press/pr/date/2021/html/ssm.pr210730~3d4d31f8e8.en.html (accessed on 24 June 2022).
[92] European Commission (2015), Critical Infrastructure Protection topic 3: Critical Infrastructure resilience indicator - analysis and development of methods for assessing resilience, https://cordis.europa.eu/programme/id/H2020_DRS-14-2015 (accessed on 16 January 2023).
[66] Federal Aviation Administration (2020), AC 150/5200-30D - Airport Field Condition Assessments and Winter Operations Safety – Document Information, Office of Airport Safety and Operations, https://www.faa.gov/regulations_policies/advisory_circulars/index.cfm/go/document.information/documentid/1029778 (accessed on 24 June 2022).
[106] Federal Emergency Management Agency (FEMA) (2015), National Preparedness Goal, US Department of Homeland Security.
[105] Federal Emergency Management Agency (FEMA) (2011), National Disaster Recovery Framework, US Department of Homeland Security, https://www.fema.gov/sites/default/files/2020-09/national_disaster_recovery_framework_2nd-edition.pdf (accessed on 24 June 2022).
[80] Federal Reserve Bank of Minneapolis (n.d.), Large bank stress test tool, https://www.minneapolisfed.org/banking/financial-studies-and-community-banking/run-your-own-stress-test-on-large-banks (accessed on 24 June 2022).
[93] Fiksel, J., I. Goodman and A. Hecht (2014), “Resilience: Navigating toward a Sustainable Future”, Solutions, Vol. 5/5, pp. 38-47.
[87] Foglia, A. (2008), “Stress Testing Credit Risk: A Survey of Authorities’ Approaches”, SSRN Electronic Journal, https://doi.org/10.2139/SSRN.1396243.
[21] Fox-Lent, C., M. Bates and I. Linkov (2015), “A matrix approach to community resilience assessment: an illustrative case at Rockaway Peninsula”, Environment Systems and Decisions, Vol. 35/2, pp. 209-218, https://doi.org/10.1007/s10669-015-9555-4.
[32] Francis, R. and B. Bekera (2014), “A metric and frameworks for resilience analysis of engineered and infrastructure systems”, Reliability Engineering & System Safety, Vol. 121, pp. 90-103, https://doi.org/10.1016/J.RESS.2013.07.004.
[62] Galaitsi, S. et al. (2020), “The Need to Reconcile Concepts that Characterize Systems Facing Threats”, Risk Analysis, Vol. 41/1, pp. 3-15, https://doi.org/10.1111/risa.13577.
[47] Galbusera, L. et al. (2018), “A Boolean Networks Approach to Modeling and Resilience Analysis of Interdependent Critical Infrastructures”, Computer-Aided Civil and Infrastructure Engineering, Vol. 33/12, pp. 1041-1055, https://doi.org/10.1111/MICE.12371.
[124] Garmezy, N., A. Masten and A. Tellegen (1984), “The Study of Stress and Competence in Children: A Building Block for Developmental Psychopathology”, Child Development, Vol. 55/1, p. 97, https://doi.org/10.2307/1129837.
[88] Geithner, T. (2014), Stress test: reflections on financial crises, Broadway Books.
[101] Gisladottir, V. et al. (2017), “Resilience of Cyber Systems with Over- and Underregulation”, Risk Analysis, Vol. 37/9, pp. 1644-1651, https://doi.org/10.1111/RISA.12729.
[99] Harris, R. and T. Impelluso (2008), “Virtual stress testing machine and the cyber-infrastructure”, Engineering with Computers, Vol. 24/2, pp. 107-117, https://doi.org/10.1007/S00366-007-0077-7/FIGURES/6.
[6] Hayes, R. (2004), “Network centric operations today: Between the promise and the practice”, RUSI Defence Systems, Vol. 7/1.
[111] Hayslip, B. and G. Smith (2012), “Resilience in Adulthood and Later Life: What Does it Mean and Where Are We Heading?”, Annual Review of Gerontology and Geriatrics, Vol. 32/1, pp. 1-28, https://doi.org/10.1891/0198-8794.32.3.
[83] Heyen, K. (2008), “The Basel II Risk Parameters: Estimation, Validation, and Stress Testing”, https://doi.org/10.1198/jasa.2008.s245, Vol. 103/483, pp. 1318-1319, https://doi.org/10.1198/JASA.2008.S245.
[51] Hoogendoorn, R., B. Van Arem and S. Hoogendoorn (2014), “Automated Driving, Traffic Flow Efficiency, and Human Factors: Literature Review”, https://doi.org/10.3141/2422-13, Vol. 2422, pp. 113-120, https://doi.org/10.3141/2422-13.
[143] Horton, R. et al. (2022), “International airports as agents of resilience”, Journal of Contingencies and Crisis Management, Vol. 30/2, pp. 217-221, https://doi.org/10.1111/1468-5973.12401.
[139] Hynes, W. et al. (2022), “Systemic resilience in economics”, Nature Physics, Vol. 18/4, pp. 381-384, https://doi.org/10.1038/s41567-022-01581-4.
[71] Jin, A. et al. (2021), “Building resilience will require compromise on efficiency”, Nature Energy, https://doi.org/10.1038/s41560-021-00913-7.
[84] Jokivuolle, E., K. Virolainen and O. Vähämaa (2008), “Macro-Model-Based Stress Testing of Basel II Capital Requirements”, SSRN Electronic Journal, https://doi.org/10.2139/SSRN.1267194.
[43] Jovanovic, A. and P. Auerkari (2016), EU project SmartResilience: The concept and its application on critical energy infrastructure in Finland, Baltica X Conference, https://www.researchgate.net/publication/303815036_EU_project_SmartResilience_The_concept_and_its_application_on_critical_energy_infrastructure_in_Finland (accessed on 24 June 2022).
[77] Jüttner, U. and S. Maklan (2011), “Supply chain resilience in the global financial crisis: An empirical study”, Supply Chain Management, Vol. 16/4, pp. 246-259, https://doi.org/10.1108/13598541111139062/FULL/XML.
[15] Kaur, S., S. Sharma and A. Singh (2015), “Cyber Security: Attacks, Implications and Legitimations across the Globe”, International Journal of Computer Applications, Vol. 114/6, pp. 21-23, https://doi.org/10.5120/19983-1932.
[109] Klasa, K. et al. (2020), “Science and Practice of Resilience: Disaster Systems Applications to Aging Resilience”, in Wister, A. and T. Cosco (eds.), Resilience and Aging, Springer, https://doi.org/10.1007/978-3-030-57089-7_4.
[69] Kott, A. et al. (2021), “Cyber Resilience: By Design or by Intervention?”, Computer, Vol. 54/8, pp. 112-117, https://doi.org/10.1109/MC.2021.3082836.
[97] Kott, A. and I. Linkov (2019), Cyber Resilience of Systems and Networks, Springer, https://doi.org/10.1007/978-3-319-77492-3.
[123] Kruk, M. et al. (2017), “Building resilient health systems: a proposal for a resilience index”, BMJ, Vol. 357, https://doi.org/10.1136/BMJ.J2323.
[118] Kruk, M. et al. (2015), “What is a resilient health system? Lessons from Ebola”, The Lancet, Vol. 385/9980, pp. 1910-1912, https://doi.org/10.1016/S0140-6736(15)60755-3.
[50] Kuhn, B. (2010), Efficient Use of Highway Capacity Summary, Federal Highway Administration.
[31] Kumar, N. et al. (2021), “A novel framework for risk assessment and resilience of critical infrastructure towards climate change”, Technological Forecasting and Social Change, Vol. 165, p. 120532, https://doi.org/10.1016/J.TECHFORE.2020.120532.
[27] Labaka, L., J. Hernantes and J. Sarriegi (2016), “A holistic framework for building critical infrastructure resilience”, Technological Forecasting and Social Change, Vol. 103, pp. 21-33, https://doi.org/10.1016/J.TECHFORE.2015.11.005.
[63] Lambert, J. et al. (2012), “Prioritizing Infrastructure Investments in Afghanistan with Multiagency Stakeholders and Deep Uncertainty of Emergent Conditions”, Journal of Infrastructure Systems, Vol. 18/2, pp. 155-166, https://doi.org/10.1061/(ASCE)IS.1943-555X.0000078.
[107] Larkin, S. et al. (2015), “Benchmarking agency and organizational practices in resilience decision making”, Environment Systems and Decisions, Vol. 35/2, pp. 185-195, https://doi.org/10.1007/S10669-015-9554-5/FIGURES/1.
[131] Lee-Kwan, S. et al. (2014), “Support Services for Survivors of Ebola Virus Disease — Sierra Leone, 2014”, Morbidity and Mortality Weekly Report (MMWR), Vol. 63/50, https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6350a6.htm (accessed on 24 June 2022).
[132] Levin-Sparenberg, E. et al. (2015), “Ebola: The Natural and Human History of a Deadly Virus By David Quammen”, American Journal of Epidemiology, Vol. 181/2, pp. 151-151, https://doi.org/10.1093/AJE/KWU354.
[82] Levy-Carciente, S. et al. (2015), “Dynamical macroprudential stress testing using network theory”, Journal of Banking and Finance, Vol. 59, pp. 164-181, https://doi.org/10.1016/J.JBANKFIN.2015.05.008.
[28] Linkov, I. et al. (2014), “Changing the resilience paradigm”, Nature Climate Change, Vol. 4/6, pp. 407-409, https://doi.org/10.1038/nclimate2227.
[19] Linkov, I. et al. (2013), “Resilience metrics for cyber systems”, Environment Systems and Decisions, Vol. 33/4, pp. 471-476, https://doi.org/10.1007/S10669-013-9485-Y/TABLES/1.
[35] Linkov, I. et al. (2018), “Tiered Approach to Resilience Assessment”, Risk Analysis, Vol. 38/9, pp. 1772-1780, https://doi.org/10.1111/RISA.12991.
[102] Linkov, I. et al. (2006), “From comparative risk assessment to multi-criteria decision analysis and adaptive management: Recent developments and applications”, Environment International, Vol. 32/8, pp. 1072-1093, https://doi.org/10.1016/J.ENVINT.2006.06.013.
[94] Linkov, I. and B. Trump (2019), The Science and Practice of Resilience, Springer International Publishing, Cham, https://doi.org/10.1007/978-3-030-04565-4.
[4] Linkov, I., B. Trump and C. Fox-Lent (2016), “Resilience: Approaches to Risk Analysis and Governance”, in IRGC Resource Guide on Resilience, International Risk Governance Council.
[68] Linkov, I. et al. (2021), “Enhancing Resilience in Post-COVID Societies: By Design or by Intervention?”, Environmental Science and Technology, Vol. 55/8, pp. 4202-4204, https://doi.org/10.1021/acs.est.1c00444.
[74] Linkov, I. et al. (2022), “Resilience stress testing for critical infrastructure”, International Journal of Disaster Risk Reduction, Vol. 82, p. 103323, https://doi.org/10.1016/j.ijdrr.2022.103323.
[140] Linkov, I. et al. (2022), “Stress-test the resilience of critical infrastructure”, Nature, Vol. 603/7902, pp. 578-578, https://doi.org/10.1038/d41586-022-00784-2.
[16] Lino, C. (2014), “Cybersecurity in the federal government: Failing to maintain a secure cyber infrastructure”, Bulletin of the Association for Information Science and Technology, Vol. 41/1, pp. 24-28, https://doi.org/10.1002/BULT.2014.1720410111.
[24] Longstaff, P. (2005), Security, Resilience, and Communication in Unpredictable Environments Such as Terrorism, Natural Disasters, and Complex Technology, Program on Information Resources Policy, Harvard University and the Center for Information Policy Research, http://www.pirp.harvard.edu/pubs_pdf/longsta/longsta-p05-3.pdf (accessed on 24 June 2022).
[38] Ludin, S. and P. Arbon (2017), “Improving community disaster resilience through scorecard self-testing”, Disaster Prevention and Management: An International Journal, Vol. 26/1, pp. 13-27, https://doi.org/10.1108/dpm-08-2016-0177.
[128] Magill, A. et al. (2013), Hunter’s tropical medicine and emerging infectious disease, Saunders, London, https://doi.org/10.1016/C2009-0-51934-4.
[136] Massaro, E. et al. (2018), “Resilience management during large-scale epidemic outbreaks”, Scientific Reports, Vol. 8/1, pp. 1-9, https://doi.org/10.1038/s41598-018-19706-2.
[120] Masten, A. (2014), “Global Perspectives on Resilience in Children and Youth”, Child Development, Vol. 85/1, pp. 6-20, https://doi.org/10.1111/cdev.12205.
[116] Mayo Clinic (n.d.), Resilience: Build skills to endure hardship, https://www.mayoclinic.org/tests-procedures/resilience-training/in-depth/resilience/art-20046311 (accessed on 24 June 2022).
[110] Mcleroy, K. et al. (1988), “An Ecological Perspective on Health Promotion Programs”, Health Education & Behavior, Vol. 15/4, pp. 351-377, https://doi.org/10.1177/109019818801500401.
[85] Miu, P. and B. Ozdemir (2008), “Stress-Testing Probability of Default and Migration Rate with Respect to Basel II Requirements”, SSRN Electronic Journal, https://doi.org/10.2139/SSRN.1365842.
[91] Mullan, M. and N. Ranger (2022), “Climate-resilient finance and investment: Framing paper”, OECD Environment Working Papers, No. 196, OECD Publishing, Paris, https://doi.org/10.1787/223ad3b9-en.
[96] Murphy, B. et al. (2014), “Planning for disaster resilience in rural, remote, and coastal communities: moving from thought to action”, Journal of emergency management, Vol. 12/2, pp. 105-120, https://doi.org/10.5055/JEM.2014.0165.
[2] National Research Council (2012), Disaster resilience: A national imperative, National Academies Press, https://doi.org/10.17226/13457.
[3] Nienaber, M. (2017), “G20 adopt list of principles to strengthen economic resilience | Reuters”, Reuters, https://www.reuters.com/article/us-g20-germany-resilience/g20-adopt-list-of-principles-to-strengthen-economic-resilience-idUSKBN16P0QY (accessed on 24 June 2022).
[90] OECD (2021), Strengthening Climate Resilience: Guidance for Governments and Development Co-operation, OECD Publishing, Paris, https://doi.org/10.1787/4b08b7be-en.
[1] OECD (2019), “Resilience-based Strategies and Policies to Address Systemic Risks”, OECD, Paris, https://www.oecd.org/naec/averting-systemic-collapse/SG-NAEC(2019)5_Resilience_strategies.pdf.
[12] Osawa, J. (2011), “As Sony Counts Hacking Costs, Analysts See Billion-Dollar Repair Bill - WSJ”, Wall Street Journal, https://www.wsj.com/articles/SB10001424052748703859304576307664174667924 (accessed on 24 June 2022).
[29] Panteli, M. et al. (2017), “Metrics and Quantification of Operational and Infrastructure Resilience in Power Systems”, IEEE Transactions on Power Systems, Vol. 32/6, pp. 4732-4742, https://doi.org/10.1109/TPWRS.2017.2664141.
[119] Panter-Brick, C. and J. Leckman (2013), “Editorial Commentary: Resilience in child development – interconnected pathways to wellbeing”, Journal of Child Psychology and Psychiatry, Vol. 54/4, pp. 333-336, https://doi.org/10.1111/JCPP.12057.
[14] Petrie, C. and V. Roth (2015), “How Badly Do You Want Privacy?”, IEEE Internet Computing, Vol. 19/2, pp. 92-94, https://doi.org/10.1109/MIC.2015.37.
[65] Pickard, S. and I. Gençsü (2021), The Airport Tracker: Uncovering aviation emissions, ODI, London, http://www.odi.org/en/publications/the-airport-tracker-uncovering-aviation-emissions (accessed on 24 June 2022).
[44] Pitilakis, K. et al. (2016), “Systemic Vulnerability and Risk Assessment of Transportation Systems Under Natural Hazards Towards More Resilient and Robust Infrastructures”, Transportation Research Procedia, Vol. 14, pp. 1335-1344, https://doi.org/10.1016/J.TRPRO.2016.05.206.
[78] Plummer, R. and D. Armitage (2007), “A resilience-based framework for evaluating adaptive co-management: Linking ecology, economics and society in a complex world”, Ecological Economics, Vol. 61/1, pp. 62-74, https://doi.org/10.1016/J.ECOLECON.2006.09.025.
[126] Pourrut, X. et al. (2005), “The natural history of Ebola virus in Africa”, Microbes and infection, Vol. 7/7-8, pp. 1005-1014, https://doi.org/10.1016/J.MICINF.2005.04.006.
[34] Reeves, M. and R. Varadarajan (2020), When Resilience Is More Important Than Efficiency, BCG Henderson Institute, https://doi.org/10.1515/9783110745511-002 (accessed on 24 June 2022).
[8] Roege, P. et al. (2014), “Metrics for energy resilience”, Energy Policy, Vol. 72, pp. 249-256, https://doi.org/10.1016/J.ENPOL.2014.04.012.
[39] Sachinthana, P., S. Chandana and P. Shehara (2022), “Development of critical infrastructure resilience index for cities in Sri Lanka”, International Journal of Disaster Resilience in the Built Environment.
[52] Sami, J., F. Pascal and B. Younes (2013), “Public Road Transport Efficiency: A Stochastic Frontier Analysis”, Journal of Transportation Systems Engineering and Information Technology, Vol. 13/5, pp. 64-71, https://doi.org/10.1016/S1570-6672(13)60123-3.
[45] Samoylenko, A., A. Panychev and S. Panychev (2017), “Evaluation of telecommunication system reliability via stress testing”, 2017 International Siberian Conference on Control and Communications, SIBCON 2017 - Proceedings, https://doi.org/10.1109/SIBCON.2017.7998430.
[49] Samuelsson, A. and C. Tilanus (1997), “A framework efficiency model for goods transportation, with an application to regional less-than-truckload distribution — Eindhoven University of Technology research portal”, Transport Logistics, Vol. 1/2, pp. 139-151, https://research.tue.nl/en/publications/a-framework-efficiency-model-for-goods-transportation-with-an-app (accessed on 24 June 2022).
[57] Schrank, D. et al. (2015), 2015 Urban Mobility Scorecard, Texas A&M University System, https://trid.trb.org/view/1367337 (accessed on 24 June 2022).
[122] Southwick, S. et al. (2014), “Resilience definitions, theory, and challenges: interdisciplinary perspectives”, European Journal of Psychotraumatology, Vol. 5, https://doi.org/10.3402/EJPT.V5.25338.
[103] Stankey, G., R. Clark and B. Bormann (2005), Adaptive Management of Natural Resources: Theory, Concepts, and Management Institutions, United States Department of Agriculture, https://doi.org/10.2737/PNW-GTR-654 (accessed on 24 June 2022).
[64] Stevens, N., D. Baker and R. Freestone (2010), “Airports in their urban settings: towards a conceptual model of interfaces in the Australian context”, Journal of Transport Geography, Vol. 18/2, pp. 276-284, https://doi.org/10.1016/J.JTRANGEO.2009.05.007.
[33] Sun, X., S. Wandelt and A. Zhang (2022), “STARTUPS: Founding airlines during COVID-19 - A hopeless endeavor or an ample opportunity for a better aviation system?”, Transport Policy, Vol. 118, pp. 10-19, https://doi.org/10.1016/J.TRANPOL.2022.01.013.
[129] Tosh, P. and P. Sampathkumar (2014), “What clinicians should know about the 2014 Ebola outbreak”, Mayo Clinic proceedings, Vol. 89/12, pp. 1710-1717, https://doi.org/10.1016/J.MAYOCP.2014.10.010.
[48] Tsionis, G. et al. (2016), The STREST Project: Harmonized Approach to Stress Tests for Critical Infrastructures against Low-Probability High-Impact Natural Hazards, https://www.researchgate.net/publication/306372408_The_STREST_Project_Harmonized_Approach_to_Stress_Tests_for_Critical_Infrastructures_against_Low-Probability_High-Impact_Natural_Hazards (accessed on 24 June 2022).
[60] Turnbull, K. (2016), Transportation Resilience: Adaptation to Climate Change, Transportation Research Board, https://doi.org/10.17226/24648.
[37] UNDRR (2017), Disaster Resilience Scorecard for Cities - Toolkit, https://www.unisdr.org/campaign/resilientcities/toolkit/article/disaster-resilience-scorecard-for-cities.html (accessed on 24 June 2022).
[25] UNDRR (2015), UNDRR Sendai Framework, https://www.undrr.org/implementing-sendai-framework/what-sendai-framework (accessed on 24 June 2022).
[79] US Congress (2010), Dodd-Frank Wall Street Reform and Consumer Protection Act, https://www.congress.gov/bill/111th-congress/house-bill/4173/text (accessed on 25 July 2022).
[117] USAID (2013), The resilience agenda: Measuring Resilience in USAID.
[30] Vugrin, E. et al. (2010), “A framework for assessing the resilience of infrastructure and economic systems”, in Gopalakrishnan, K. and S. Peeta (eds.), Sustainable and Resilient Critical Infrastructure Systems: Simulation, Modeling, and Intelligent Engineering, Springer Berlin Heidelberg, https://doi.org/10.1007/978-3-642-11405-2_3/COVER/.
[125] Wagnild, G. and J. Collins (2009), “Assessing resilience”, Journal of Psychosocial Nursing and Mental Health Services, Vol. 47/12, pp. 28-33, https://doi.org/10.3928/02793695-20091103-01.
[114] WHO (2017), Building resilience: a key pillar of Health 2020 and the Sustainable Development Goals: examples from the WHO Small Countries Initiative, World Health Organization, Regional Office for Europe, https://apps.who.int/iris/handle/10665/338752.
[130] WHO (2015), An emergency within an emergency: caring for Ebola survivors, World Health Organization, https://www.afro.who.int/news/emergency-within-emergency-caring-ebola-survivors (accessed on 24 June 2022).
[113] Wister, A., K. Klasa and I. Linkov (2022), “A Unified Model of Resilience and Aging: Applications to COVID-19”, Frontiers in Public Health, Vol. 10, https://doi.org/10.3389/fpubh.2022.865459.
[22] Wood, M. et al. (2012), “Flood risk management: U.S. Army Corps of Engineers and layperson perceptions”, Risk analysis : an official publication of the Society for Risk Analysis, Vol. 32/8, pp. 1349-1368, https://doi.org/10.1111/J.1539-6924.2012.01832.X.
[56] Yamashita, T., K. Izumi and K. Kurumatani (2004), “Car navigation with route information sharing for improvement of traffic efficiency”, IEEE Conference on Intelligent Transportation Systems, Proceedings, ITSC, pp. 465-470, https://doi.org/10.1109/ITSC.2004.1398944.
[55] Yan, G. et al. (2006), “Efficient routing on complex networks”, Physical Review E - Statistical, Nonlinear, and Soft Matter Physics, Vol. 73/4, p. 046108, https://doi.org/10.1103/PHYSREVE.73.046108/FIGURES/4/MEDIUM.
[11] Zhao, J. and S. Zhao (2010), “Opportunities and threats: A security assessment of state e-government websites”, Government Information Quarterly, Vol. 1/27, pp. 49-56, https://doi.org/10.1016/J.GIQ.2009.07.004.
Notes
← 1. In some countries, for example, the United States, critical infrastructure has a precise definition. In the United States, it is defined as “[s]ystems and assets, whether physical or virtual, so vital to the United States that the incapacity or destruction of such systems and assets would have a debilitating impact on security, national economic security, national public health or safety, or any combination of those matters.”
← 2. Based on Horton et al. (2022[143]).