Abstract
The COVID-19 pandemic has drawn worldwide attention to the difficulties inherent in managing disasters. As we learn globally and collectively how the virus spreads, infects, and kills, we contend domestically and separately with levels of development, community structures, and cultural norms that can exacerbate or mitigate COVID-19’s effects, and that introduce new complexities to international relations. As the pandemic continues, scholars of civil unrest, conflict, and cooperation have been considering more generally the impact disasters have on interstate relations, state resilience, patterns of violence and hostility, and the vulnerabilities that condition conflict.
For approximately two decades, studying how disasters condition interstate relations has given rise to a literature on disaster diplomacy (Olson & Drury, 1997; Drury & Olson, 1998; Kelman, 2011, 2016), in which scholars consider disasters, such as the disaster caused by the COVID-19 virus, as ‘shocks’ that can disrupt and potentially terminate civil and interstate conflict. Because disasters, unlike wars or regime changes, are unrelated to the politics of a particular rivalry, they may counteract the inertia inherent in protracted conflicts, revealing rivals’ limitations and levels of commitment, and catalyzing the peace process (Kreutz, 2012; Mandel, 2002; Dreher & Fuchs, 2015; Akcinaroglu, DiCicco & Radziszewski, 2011). At the same time, the shock caused by a critical event may exacerbate existing rivalries and spark or increase levels of violence (Brancati, 2007; Ker-Lindsay, 2000; Nel & Righarts, 2008). Though we know that both conflict and cooperation are possible outcomes of disasters, we know less about the conditions under which we can expect each particular outcome to occur.
In this special issue, ‘Disaster diplomacy: The intricate links between disaster and conflict’, we offer new research on the disaster–conflict nexus. The pieces herein refine our understanding of the conditions under which we can expect disasters to lead to conflict versus cooperation by considering disasters and their influence on the strategies and tactics of people already involved in conflict, as well as on the timing and levels of hostility between rivals. Importantly, these pieces and this special issue also highlight a three-pronged conceptual framework for studying the disaster–conflict relationship. With this framework as the main focus of our introductory article, we argue that adopting it has the potential both to address key gaps in the current literature and to advance disaster diplomacy scholarship even further.
The first prong of the framework is the idea that disasters are endogenous social phenomena. As we explain below, disasters are not only potential drivers of conflict and cooperation – disasters can be driven by civil and interstate tensions as well. Pieces by Lee et al. (2022), Haer & RezaeeDaryakenari (2022), and Koehnlein & Koren (2022) consider this endogeneity carefully in their empirical strategies. By explicitly acknowledging the endogeneity of disasters, these authors are able to expand our understanding of how disasters and conflict unfold together over time, each conditioning the evolution of the other. Their work also demonstrates a gap in the disaster diplomacy literature, which often fails to investigate the reverse causal relationship: the effects of conflict on disaster. Barceló et al. (2022) demonstrate a logical path into examining this causal pathway in their investigation of COVID-19 policies and state repression. In finding that repressive regimes are likely to use pandemic management policies to restrict human rights, their work suggests that the evolution of the pandemic itself will be conditioned by regime type and policies driven by factors outside of standard public health considerations.
Considering the endogeneity of disasters brings us to the second prong of the framework, which is the explicit consideration of disasters and conflicts as intertwined processes or cycles, rather than discrete events. A disaster begins with an event, the effects of which lead to relief, recovery, rebuilding, and eventually preparation and mitigation for the next potential event. This concept of disasters as processes is rarely explicitly recognized in the disaster diplomacy literature, making the articles in this special issue some of the earliest to do so. Lee et al. (2022) expertly demonstrate how fluctuations in levels of violence over time can be modeled as functions of disasters over time, prompting us to think of the two phenomena as concurrent and ongoing processes. Haer & RezaeeDaryakenari (2022) also allude to the disaster process, as they allow each disaster and its effects to be revealed gradually and for months beyond the actual precipitating incident. Chung & Rhee (2022) account for the disaster cycle by examining repeated disasters. In finding that disasters drive negative perceptions of out-groups, their work demonstrates how continued stresses can escalate tensions over time and space. Together, these pieces illuminate how conceptualizing disasters and conflicts as interwoven processes helps us more realistically approximate and more accurately model the relationship.
The final prong of the framework is rooted in a phenomenon that is occasionally acknowledged but rarely investigated: the co-determination of disasters and conflict. Acknowledging disasters as endogenous processes, intertwined with conflict, draws our attention to characteristics such as poor governance, ethnic fractionalization, and economic deprivation that can make a group vulnerable to both disasters and conflict. Disasters, then, do not just affect conflict directly – disasters also affect conflict through their effect on the vulnerabilities that condition conflict. Nemeth & Lai (2022), in their investigation into subnational one-sided violence, find that localized disasters can weaken one set of combatants vis-à-vis another. By considering disasters as able to affect other determinants of conflict, they offer a more nuanced understanding of the circumstances under which disasters can lead to conflict versus cooperation. And by acknowledging that the disaster cycle can vary subnationally, their work points future scholars to investigate the possibility that disaster management could be a tool of repression in and of itself, capitalizing on and exacerbating pre-existing vulnerabilities disproportionately for unfavored groups. Barceló et al. (2022) bring this issue to the forefront, demonstrating that regimes with a history of repression are more likely to use pandemic management policies as an excuse for restricting human rights.
This three-pronged framework to conceptualize the disaster–conflict relationship is directly applicable to the COVID-19 pandemic, a disaster of global proportions that we have seen challenge emergency management capacities both domestically and internationally. For example, in terms of co-determination, we have seen the virus spread more rapidly among people who cannot afford to socially distance or stay home from work. In terms of endogeneity, we have seen uneven vaccine roll-out and policy restrictions (such as school and workplace closings and stay-at-home orders) both within and between countries (Hale et al., 2021; Strange & Patkee, 2021). These outcomes have in turn driven civil unrest (Wood et al., forthcoming) and challenged international diplomacy, even between allies (United Nations, 2021). Nearly two years into our awareness of the virus and its contagion, we are all struggling to manage the pandemic as a process that demands relief, recovery, rebuilding, and adaptation to future threats.
There is perhaps no better time to adopt and implement this framework than right now. As we continue to study the COVID-19 pandemic and its effects on conflict and cooperation, we should also keep in mind that the pandemic is unfolding as a process, both conditioning and conditioned by pre-existing vulnerabilities and the conflict environment. In this introduction, we highlight the
First prong: Disasters as endogenous social phenomena
As a group, contributors to this special issue follow the guidance of disaster studies scholarship that characterizes disasters as distinctly social phenomena.
Conceptualizing disasters in this way requires us to acknowledge that societal and governmental factors, such as governance quality, legitimacy, and social cohesion, affect disasters. An event’s effect on society is managed, mitigated, and at times exacerbated or even averted by factors such as preparation, relief, and recovery efforts, levels of citizens’ trust in public officials and the information they give, media narratives, social capital, infrastructure, and more (Aldrich, 2012; Atkeson & Maestas, 2012; Quiroz Flores, 2015; Ross, 2014; MacAskill & Guthrie, 2015; Reinhardt, 2015). The length and severity of disasters are conditioned by the people who manage and live through them, the information they have, and the choices they make before, during, and after the critical events that cause the disruption (Christensen & Lægreid, 2005; Reinhardt & Ross, 2019). Acknowledging this fundamental attribute of disasters has implications for how we study the relationship between disasters and conflict. The pieces in this special issue demonstrate ways to model this endogenous relationship empirically, and point us toward multiple avenues for studying the reverse causal relationship.
Empirically modelling disaster endogeneity
If the disruptions we refer to as disasters are not an exogenous shock, identifying their relationship with conflict is a challenge. The articles in this special issue tackle the endogeneity of disasters in a variety of ways that both expand and deepen our understanding of the disaster–conflict nexus. Lee et al. (2022) conduct their estimations by using standard deviations from the mean and time to next dispute, which helps isolate changes in levels of violence over time and link a disaster event to its long-term effects. Haer & RezaeeDaryakenari (2022) distinguish short-term from long-term effects of disaster events on one-sided violence by lagging all explanatory variables to decrease the likelihood of problems caused by a reciprocal association with their dependent variable. Nemeth & Lai (2022), by focusing solely on natural disasters in their investigation into disaster effects on negotiations, reduce ‘the possibility that a disaster endogenous to the conflict – like the destruction of infrastructure due to fighting – affects the likelihood of negotiations’ (p. 34). And Koehnlein & Koren (2022) conduct multiple tests to account for possible endogeneity as they find that the pandemic should be explicitly considered as a draw on state capacity when studying the determinants of violence by nonstate actors.
Conflict as a cause of disaster
Beyond this work and the vast majority of the disaster–conflict literature, which has examined the effect of disasters on conflict, the pieces in this special issue suggest great potential in examining the underexplored reverse causal relationship: the effect of conflict on disasters. For example, Barceló et al. (2022) help us understand how cross-national variation in response to international disaster events is determined in part by a country’s history of political violence. Lee et al. (2022) find that violence in longstanding rivalries after disasters increases when leaders have multiple rivalries, suggesting the use of violence as a diversion from internal challenges of disaster management. When ignoring or failing to meet these challenges, leaders are prolonging the disaster itself, which then increases pressure for humanitarian and relief aid (Bluhm et al., 2016). Future scholarship could further investigate the extent to which diversionary shows of force exacerbate the human and material costs of disasters and recovery from them.
One means through which conflict is sure to affect disaster outcomes will be via states’ resilience in the face of disasters, and the ways they cope domestically and internationally with disasters’ consequences. Learning processes within governments and organizations, as well as in complex intergovernmental systems such as those called upon during disaster recovery, take time. A vast literature on resilience (Demiroz & Haase, 2018) examines the ability of a society to bounce back from an unplanned shock and resume or surpass its pre-shock development trajectory, though little of this work focuses on resilience of the state apparatus itself. If, as Nemeth & Lai (2022) demonstrate, a state’s willingness to negotiate with rebel groups depends on the relative effect of a disaster event on one group versus another within intrastate conflict, it is likely that the existence and/or outcome of negotiations will condition the resilience of the state to the event itself.
The endogeneity of disasters also has direct implications for the study of between-group tensions and equality. We know from previous scholarship that living through disasters creates a shared experience (Ross, Rouse & Mobley, 2019) that may fuel grievances and civil unrest (Wood et al., forthcoming). Chung & Rhee (2022) demonstrate an increase in negative perceptions of outgroups due to disasters. In the reverse relationship, as Nemeth & Lai (2022) suggest, these tensions are likely to lead to disproportionate disruptions from disaster events, particularly when rivalries cause groups to live geographically separated from each other. For example, when an earthquake or cyclone strikes a religious or ethnic enclave, those religious and ethnic groups will experience the disruption disproportionately compared to those groups living farther away. Even in the case of a disaster event that strikes across regions populated by different groups, those suffering marginalization or hostility due to intrastate strife will also suffer greater disruption as the state channels resources into relief and recovery for more favored groups.
Second prong: Disasters as a process
Following from the concept of disasters as endogenous, disaster scholarship has a history of conceptualizing disasters as part of a hazards cycle, involving preparedness, mitigation, critical events, relief, recovery, rebuilding, and adaptation (see discussions in Khan, Vasilescu & Khan, 2008). At any given moment, every society or country is in some stage of this cycle. For some, the time span between disaster-precipitating events is long enough to forget how to adapt to long-term environmental changes; for others, the time span is too short to be able to prepare for the next event. Countries experiencing multiple hazards may experience overlapping phases of the cycle at once, such as a country that is simultaneously recovering from a deadly cyclone, preparing for an oncoming hurricane, and adapting to climate change.
Thinking of hazards as cycles or processes enables us to see how the evolution of disasters can be interlinked with fluctuations in civil unrest, patterns of violence, and protracted armed conflict. Acknowledging the hazards cycle also illuminates how the pieces in this special issue point us toward untapped areas of investigation with implications for policy regarding peacekeeping, disaster relief, and ethnic strife. Further, despite the implementation of clever empirical strategies, we are forced to recognize the limitations of current disaster data, and to point out further lines of inquiry that may not be adequately addressed with discrete event data.
Policy implications of disasters as a process
Haer & RezaeeDaryakenari (2022) model violence against civilians and find that it decreases immediately after a disaster, then increases in the long term. This trend is not only conditioned by the evolution of the conflict, but also by the evolution of the hazards cycle. Extensions of their discovery can inform important public policy debates, such as the degree to which UN peacekeeping forces should be deployed to protect civilians in civil wars (Hultman, 2013) or immediately after disaster events.
Also in this issue, Nemeth & Lai (2022) use geocoding to examine the effects of disasters on one-sided violence at the subnational level. Future work could explore the extent to which disaster management becomes a tool of repression in and of itself, allowing governments to target and advantage pro-regime areas by selectively providing relief and recovery programs. These areas are then likely to become less vulnerable to future disasters as well as to attacks from combatants. If so, external observers and policymakers could understand how to more precisely target their own disaster relief and, at the same time, their peacekeeping efforts.
Viewing disasters as a process also highlights the importance of examining recurring and expected critical events, such as landslides due to expected precipitation, annual hurricane and typhoon seasons, and droughts caused by larger global patterns such as El Niño. Chung & Rhee (2022) find that disasters increase negative perceptions of other groups, suggesting that over time, these perceptions change as a community moves through a disastrous event to relief and recovery. This work is critical to understanding more about the foundations of conflict. They illuminate the possibility that the hazards cycle is linked to a cycle of hostility between groups which, though ebbing and waning around each critical event, still shows an overall increase over time. As the authors astutely suggest, knowing these trends exist can help design policy tools to counteract xenophobic and racist sentiments when we expect hostilities to surge.
Discrete event disaster data and unanswered questions
Investigating the disaster–conflict nexus by incorporating the concept of disasters as a process has at least one serious drawback, which is the availability of data. Largely, disaster–conflict literature has relied on the Emergency Event Database (EM-DAT 1 ), developed by the Centre for Research on the Epidemiology of Disasters (CRED), to supply data on disasters. The availability of these data has enabled disaster diplomacy scholarship to expand theoretically and empirically by offering a standardized dataset of disasters across time and space. With it, we have learned that natural disasters serve as shocks that can induce conflict within states (Brancati, 2007; Ker-Lindsay, 2000; Nel & Righarts, 2008). For example, Nel & Righarts (2008) find that natural disasters significantly increase the risk of violent civil conflict in the short and medium term. Brancati (2007) shows that natural disasters promote civil conflict because they both exacerbate resource scarcities and attract international aid, which can be used to strengthen rebel groups. And Schleussner et al. (2016) offer evidence that risk of outbreak of armed conflict is enhanced by climate-related disaster occurrence in ethnically fractionalized countries.
Now that we have extensive insights based on these event data, and in light of the view of hazards as a cycle, it might be tempting to eschew event data that do not capture disasters as a process. After all, viewing disasters as one moment in time inhibits our ability to investigate the role of disaster preparedness in fueling conflict and unrest. States’ disaster preparedness affects leaders’ tenure in office (Quiroz Flores & Smith, 2013), concerns about which may prompt diversionary violence or repression from regimes under pressure. Do disaster-prone regions pose opportunities for external actors to further their own security goals by supporting resilience and preparedness in strategic ways, or even opportunities to step in and influence the outcome of previously existing conflict? If so, policymakers seeking to alleviate the use of state repression around the world could help governments fortify preparedness as a means to achieve their own strategic and humanitarian goals (Balla & Reinhardt, 2008). Yet measuring disasters only as an isolated event makes such questions much more difficult to explore.
Acknowledging the limitations of data that do not capture the hazards cycle also forces us to acknowledge the merit of case study work that probes the evolution of one country or hazards cycle at a time. Such in-depth work uncovers new phenomena to examine, such as intergovernmental blame shifting (Atkeson & Maestas, 2012) and the role of adaptation in prompting recognition of marginalized groups (Plein, 2019). Suppose we want to investigate how small states, for example, may choose to enhance their status in the international system by providing development and/or humanitarian aid to states hit hard by natural disasters, as Israel does. As the processes of relief and recovery cycle back into preparedness and adaptation, does this behavior engender other foreign policy allegiances or cooperation for the small donor state?
All the same, we would not argue that the EM-DAT dataset is no longer useful for exploring the disaster–conflict relationship, nor that its potential for informing research and policy debates has been completely tapped. Rather, the pieces in this special issue demonstrate just how useful such resources can be, and that with careful attention and appropriate modelling, the hazards cycle can be considered, as Haer & RezaeeDaryakenari (2022) and Lee et al. (2022) expertly demonstrate. We do suggest, however, that we remain mindful as a discipline of the need to account for time trends and cycles in our analyses, and that event databases are not at the end of their own evolution. If such data were able to distinguish between various phases of the hazards cycle in addition to disaster events and their ultimate outcomes, they would more closely approximate the realities of the disaster experience and its relationships to processes of conflict, peace, and cooperation.
Third prong: Co-determination of disasters, conflict, and cooperation
Now that we can think of disasters as endogenous social processes that are intertwined with conflict, we can also more easily see how disasters and conflict might be co-determined by a common set of vulnerabilities. Countries experiencing unrest or conflict are experiencing repeated iterations of the disaster cycle. The scale, location, intensity, frequency, and cost of disasters depend on factors that also condition the scale, location, intensity, frequency, and cost of conflict. Thinking about these vulnerabilities as co-determining conflict and disasters can help us in terms of both measurement and theoretical advancement.
Measuring vulnerabilities
The pieces in this special issue account for a variety of pre-existing vulnerabilities in their empirical investigations, including levels of economic development, pre-disaster state repression, civil and interstate dispute histories, power dyads and ratios, negotiation histories, terrain, and ethnic fractionalization (Lee et al., 2022; Nemeth & Lai, 2022). These vulnerabilities are largely considered to be control variables both in this issue and in traditional disaster–conflict scholarship. Yet the concept of the endogenous disaster encourages us to view disasters and conflict as conditioned on overlapping sets of vulnerabilities. Vulnerabilities such as economic deprivation, poor health, and inadequate infrastructure can both lead to conflict directly by creating grievances, and affect conflict by deepening and lengthening the effects of a disaster and complicating its management (Wood & Wright, 2016).
Theoretical advancement
One of the most important contributions of this collection, therefore, lies in helping us identify new avenues for research that investigates the vulnerabilities that determine both disasters and conflict, thus refining the traditional question of ‘How do disasters affect conflict?’ to ask instead ‘How do disasters influence the vulnerabilities that condition conflict?’ (see Hollis, 2018 for further elaboration on this point). Nemeth & Lai (2022) present a natural jumping-off point for further research in this vein by showing how disasters that affect only one actor in a rivalry can help prompt negotiations that ultimately end violence. Further investigations could probe this relationship more deeply, investigating which vulnerabilities are most likely to influence the effects of disasters and conflict on each other, thereby to inform the policy and relief community about the most pertinent areas to target with development efforts.
The vulnerabilities that co-determine disasters and conflict are perhaps at no time more apparent than during a pandemic. Pandemics stem from events that are rapid in onset but long in duration, which makes their effects differ from those of both slow onset disasters like famines and relatively brief events like hurricanes and earthquakes. Critically, the length, depth, and breadth of a pandemic’s effects are conditioned not just by the ability of a state to manage the event, but also on its ability to manage individual event-prolonging behavior. In contrast to most disaster events, in pandemics individuals fuel the event as they react to it, because their very behavior determines where and how quickly the virus spreads. Vulnerabilities such as high population density, poor sanitation, inadequate housing, and weak infrastructure then have the potential to both prolong the pandemic and amplify its effects on unrest and hostility (see full discussion in Wood et al., forthcoming).
The COVID-19 pandemic has also highlighted international and cross-national vulnerabilities that condition public health and human rights outcomes. Domestically, as Barceló et al. (2022) demonstrate, regimes with a history of repression may be more likely to use the pandemic as an excuse to implement policies that restrict human rights. We also see cross-national cultural, regime, and leadership effects on death and infection rates of COVID-19 (Windsor, Dowell & Graesser, 2014; Windsor et al., 2020). Internationally, we see the COVID-19 vaccine becoming a currency in and of itself, distributed swiftly and firstly in the Global North, and then channeled to the Global South with greater delay (Strange & Patkee, 2021). While anti-vax groups in countries such as Belgium, Italy, New Zealand, and the UK take to the streets to oppose mandatory vaccines as rights violations (AFP, 2021; Al Jazeera, 2021; Menon & Awasthi, 2021; Paterlini, 2021; Toner, 2021), grievances are also rising over the inequitable cross-national vaccine roll-out (AFP, 2021; BBC News, 2021; McClure, 2021). Can international organizations help developing countries navigate the diplomatic aspects of vaccine acquisition while they attempt to balance domestic tensions and unrest?
There are multiple relevant extensions of research along these lines. For example, what is the extent of disasters’ influence on the effects of issue areas and issue characteristics on conflict or cooperation? Do disasters have varying influence on cooperation in areas with high costs of cooperating, such as national security, compared to areas like trade and the environment where cooperation costs are lower? Do disasters lead to increased incentives for negotiation or mediation between belligerents in interstate disputes? And how do national leaders use disasters to take actions that will increase their tenure?
Conclusion
In this special issue, we offer a three-pronged theoretical framework for understanding and studying the intricate links between disasters and diplomacy. Contributors provide fresh theoretical and empirical insights into the relationship and point us toward multiple potential avenues for future research. The articles add to the existing international relations literature on disaster diplomacy by identifying the conditions under which we should expect disasters to lead to conflict versus cooperation, analyzing temporal patterns that affect regional and international actors in the aftermath of disasters, and highlighting the role of domestic institutions in shaping states’ behavior in the wake of disasters. These substantive advancements are achieved by a unified understanding of three central principles of the disaster–conflict nexus that comprise our proposed theoretical framework: disasters are endogenous social phenomena; disasters and conflict are intertwined processes or cycles; and disasters and conflict are co-determined.
The six articles in this special issue offer multiple suggestions for empirically navigating the complexities this framework generates. Using proportional hazards and duration models, as well as geocoding and temporal lags, enables contributors to reduce the effects of endogeneity and refine our understanding of one causal relationship – the effect of disaster on conflict. We also suggest a variety of benefits that could come from investigating the effect of conflict on disasters, as well as multiple policy implications and avenues for further exploration. In particular, we note the potential of future work to help plan relief and peacekeeping efforts, as well as to inform management strategies for ongoing and future disasters, such as the COVID-19 pandemic.
Recent events have served as an impetus for a growing interest in the topic of disasters and conflict. This special issue and our proffered framework demonstrate the opportunity for further systematic research exploring the dimensions of the links between disasters and diplomacy in general, and the idiosyncrasies of pandemics as disasters in particular. To gain a more comprehensive understanding of the intricate links between disasters, conflict, and peace, we therefore encourage the reader to consider disasters and conflict as endogenous, co-determined processes, the investigation of which can reveal critical insights into human and societal behavior.