Challenges to Coordination: Understanding Intergovernmental Friction During Disasters

All societies hope to minimize the casualties, property damage, and business interruption that come with disasters that, because of human induced-climate change, are becoming more costly (Below and Wallemacq 2018). A number of policies reduce the human and economic costs of crises, including preventing settlement in vulnerable areas through zoning laws, wide scale evacuation before the event, and rapid and organized response from local and national authorities (Metaxa-Kakavouli et al. 2018). Successful post-disaster outcomes are not as common as observers would like. Experts have pointed to a handful of well-coordinated responses following disasters, including the intensive search after the Columbia shuttle disaster that resulted in locals and professional responders rapidly locating the bodies of the astronauts and critical debris from the shuttle. Similarly the largest boat evacuation in history after the 9/11 terror attacks on the Twin Trade Towers in New York City resulted in the successful evacuation of 500,000 people in an afternoon thanks to coordination between the Coast Guard and private boats (Boin and Bynander 2014). Successful disaster responses facilitate strong and rapid population return, generate access to capital for affected businesses and entrepreneurs, create stocks of permanent housing for the displaced, and rebuild the social safety net of schools, hospitals, and social services (BCLC 2010).

But failures in coordination and communication during disaster response have become so common that they are expected (Boin and Richardson 2015). Some USD 95 million of international aid intended for victims of the 1984–1985 famine in Ethiopia was diverted to local warlords and corrupt aid officials, resulting in the deaths of many from preventable starvation. Satellite images from post-Hurricane Katrina New Orleans revealed a fleet of unused school buses that could have evacuated hundreds of residents from oncoming floodwaters had local and regional authorities shared information. In Japan, the Tokyo Electric Power Company did not train with local firefighters on equipment that could cool reactors during a meltdown. As a result of that choice and a lack of preparation at the local and national levels, three reactors at Fukushima Dai-ichi melted down within 50 h of Japan’s 3/11 earthquake and tsunami (Naderpajouh et al. 2018). In all of these cases observers agreed that better coordination—the alignment of goals and methods across multiple levels of response activity—would have created better outcomes, measured in lower mortality rates, more rapid medical service and food delivery, and less waste. I build on past research that has labeled the failure to coordinate effectively as “friction” (Siembieda 2012).

Despite the pressing urgency of disaster response, the conditions under which local, regional, and national authorities and residents synchronize well during disaster remain vague. As a result, proposals to ameliorate discord and a lack of coordination have often focused on technical fixes, such as building shared datasets on disasters (Eyra et al. 2001) and new, phone-based technologies that can gather data from multiple sources (Oden et al. 2012). Moving away from individualized or technical fixes based on uncertainty, this article systematically evaluates the conditions when friction is more likely using a new dataset of 18 disasters from countries with a mix of development levels, governance systems, and friction. Data about each crisis include information on the level of development, governance, the cost of the disaster as a percentage of overall gross domestic product (GDP), the percentage of population killed, nongovernmental organizations (NGOs) per capita, and level of goal sharing. Recognizing warnings from scholars that attempts to judge overall “success” of a disaster response in binary terms “does not adequately describe the problem” (Boin and Richardson 2015), I instead code disaster response “coordination.”

After an initial Tobit regression analysis, I move to a qualitative comparative case (QCA) approach (Ragin 2000) and three mini-case studies to understand when friction is most (or least) likely. Across all three types of analysis (regression, QCA, and case studies), I find that low levels of development, lower economic costs from the crisis, and poor planning and logistical infrastructure correlate with friction. Such mismatches and chaos reduce the efficiency of what ideally would be a joint bottom-up and top-down response process. “This lack of coordination in providing humanitarian assistance can be an enormous hurdle […that] will hinder a proper response to the needs of the affected population” (WHO 2013, p. 1).

Next, where previous studies of disaster response have often relied on the analysis of a single case or handful of cases, this study uses data from 18 catastrophes around the world to avoid making inference based on a narrow range of cases from a single country or disaster (Johnson and Olshansky 2017). Instead of basing its results on developed or developing nations, or democratic or authoritarian ones, these findings may have more generalizability. Finally, the counterintuitive findings from this analysis help open up alternative policy responses in future disasters. For example, these data indicate that crises with lower economic costs may not create the sense of urgency that could help coordination between levels of government and civil society. By recognizing that “minor” disasters may in fact be ones where friction is more likely, civil and state authorities can better realize the need to find balance in their interactions.

The article is structured as follows: I first lay out the literature on coordination during crisis and then move to a discussion of potential factors that may minimize (or increase) the likelihood of friction. The middle of the article brings quantitative regression analysis, QCA, and mini-case studies to illustrate patterns of these factors and their relationship to coordination or friction outcomes in these datasets. I conclude the article with a discussion of the shortcomings of this study, future research agendas in this field, and policy recommendations.

Countries around the world struggle to match bottom-up and top-down policies across a variety of public policy areas. Especially for countries with federal or quasi-federal governance systems that face a patchwork of conditions across their jurisdictions, bureaucrats and decision makers often strive to create a top-down, uniform framework for localities that matches bottom-up interests and capacities to facilitate governance in all policy arenas (Girdwood 2013). This is true even in policy fields that do not involve potential loss of life, extensive scrutiny from media, and high levels of regulatory oversight (if occasionally only ex post) as we find in disasters. For example, “[t]he challenge for the Netherlands—as well as for other countries—is to harmonize a national adaptation policy with its spatial planning policy” (De Bruin et al. 2009, p. 25). Government experts and decision makers typically hold access to cutting edge technology and literature on best practices. Nevertheless, local experts, because of their detailed knowledge of hyperlocal conditions, have a better grasp on how best to implement new plans.

So too in the field of energy efficiency, German cities experimenting with energy efficiency, renewable energy, and a smaller carbon footprint have not been able to find a sweet spot in their interactions with the central government. These localities need to balance a bottom-up, entrepreneurial culture of individual designers and architects with standardized top-down planning from regional and federal government experts who set laws and require national compliance (Fastenrath and Braun 2017). The Chinese central government must regularly renegotiate national environmental enforcement protocols and laws with thousands of regional and local governments with varying degrees of economic development and environmental pollutants (Shi et al. 2013).

With these regular problems of friction even in nonemergency policy areas, the challenges of coordination and communication during crisis have more severe and obvious consequences: lives are lost and recovery retarded should responders and decision makers fail to move resources, personnel, and aid in time to disaster struck communities. Nations have moved to create frameworks intended to better handle catastrophe coordination, as “[t]urning policy into practice requires finding the appropriate balance between top down and bottom up engagement” (GNDR 2009, p. vi). Analysts have argued that the ideal disaster response involves simultaneous engagement from community- and locality-based resources in coordination with national authority plans and resources (Muir-Wood 2016). Precisely what that balance is, or the conditions under which it is likely to be reached, are not clear. North American authorities, for example, seeking to coordinate the actions of thousands of agencies, institutions, and NGOs around the country, built the National Disaster Recovery Framework (NDRF) in hopes of a harmonious response structure (DHS 2016). The NDRF mentions coordination more than 100 times in 53 pages. But few would argue that the United States has mastered the art of coordinated disaster response.

Scholars studying disaster response have sought to shed light on this problem in a variety of ways. One researcher created a framework for analyzing disaster coordination based on an economic model of the labor market that involves job seekers and employee seekers. Under such assumptions, governments and relevant actors—labeled external and internal agents in this framework—need information and engage in transactions for stabilization and recovery. In this model, communities hit by disasters serve as the internal agents while national government agencies like the Federal Emergency Management Agency (FEMA) are external ones who provide resources (Siembieda 2012). Other teams of researchers have argued that coordination in disaster response works best in conditions of “emergent coordination,” that is, through indirect, unplanned, and spontaneous collaboration (Faraj and Xiao 2006, p. 1166). Other research has focused on the ways that government agencies and decision makers take credit (or blame) for the response pattern (Boin et al. 2008). Another investigation of extreme weather events and adaptation in the premodern world by Peregrine (2018) looked at greater local participation in decision making and coordination in a balance with the enforcement of norms, and described the core balance as one involving flexibility to adapt to changing circumstances and tight adherence to social norms.

Given this past research, a number of potential factors stand out for further systematic investigation: level of development, governance, impact of the disaster, civil society capacity, goal sharing between levels of governance, administrative capacity of regional and local governments, and level of logistics and planning. I investigate each of these factors and how they may correlate with disaster response coordination in turn below.

The level of development of a society or nation hit by disaster could impact the likelihood of friction in a number of ways (Manandhar and McEntire 2014). Developed societies, with higher levels of professionalization in their bureaucrats and higher levels of education, may have an easier time coordinating local and national responses. Alternatively developing societies, which more regularly have authoritarian or at least less democratic governance, may find that central authorities can steamroll or overcome local resistance to central plans.

The level of governance in a nation could be correlated with development and therefore have an impact on coordination ability afterwards a shock or disaster (Gall et al. 2014). More democratic and transparent societies might find collaboration easier as they would seek to consult with, rather than overrule, local disaster managers, town mayors, and regional governors. Democratic countries may envision civil society as a desired and active participant in the disaster response stage, helping to organize volunteers into organizations such as voluntary organizations active in disasters (VOADs), community organizations active in disasters (COADs), and other trained community-based aid givers.

The cost of the disaster could potentially reduce or increase the transactions costs associated with coordination (Benson and Clay 2003). I envision costs in two ways: the economic costs of the crisis, captured here in terms of the disaster’s costs as a percentage of overall GDP, and also the human cost, measured as a percentage of the population killed. When a society faces a larger disaster—with either higher human or economic costs—it may be more incumbent on the national government to show leadership by leading local governments and ignoring dissent. This was the case in China’s 2008 Wenchuan Earthquake, the Chilean 2010 earthquake and tsunami, and Japan’s 3/11 triple disasters. However, it may also be possible that, in a massive disaster that creates chaos and reduces the effectiveness of standard service providers in the area, the central government should back away from seeking to standardize and coordinate responses and allow local residents and governments to lead (Blakeman 2017). Or, as others have argued, “because recovery is a process rather than an outcome, it is best accomplished at local levels of government” (Johnson and Olshansky 2017, p. 9).

The ability of the civil society to coordinate and respond, measured here in terms of NGOs per capita, can also alter whether outcomes move smoothly (Shaw and Godao 2004). As with previous variables, the impact of higher levels of NGOs per capita is unclear. It may be that having more NGOs provides little space for top-down decision making and opens the way for conflict and disconnection post crisis. Alternatively societies with more NGOs may be more democratic and therefore more likely to see NGOs as an ally of local, regional, and national governments during crisis.

How well the local authorities and the central government harmonize their visions of a response—a variable I label “goal sharing”—may alter the trajectory of the response (Edgington 2010). If the national government uses the disaster response as a chance to clear residents from vulnerable areas, for example, that may disconnect with local plans of increasing livable and affordable housing in the area. If the central government regularly coordinated before a disaster shock with local authorities on extreme weather, resilience, and disaster management plans—as has been the case in the Netherlands, for example—the arrival of a different major catastrophe, such as a flood, may provide another scenario where cooperation is likely.

Whereas NGO capacity captures the ability of the third sector to respond to needs, local capacity instead seeks to capture how well the local government is equipped to respond in a crisis. Some local governments find themselves underfunded through local tax revenue and heavily dependent on the central government for administrative and financial resources—as in Japan. In other cases local cities and regions may themselves have tremendous resources—such as New York City—and be able to coordinate their own well-oiled responses to crises.

The ability of the society to get personnel, material, and information where it needs to go during a crisis can be categorized in terms of the strength of logistical infrastructure. Even before the Haiti Earthquake, for example, that nation struggled to provide potable drinking water, education, and health care to its citizens. After the earthquake killed many members of the armed forces, police and fire teams, and government officials, it became even more of a struggle for the government to assist its citizenry. In contrast Japan’s air, rail, road, and shipping infrastructure made it possible to deliver needed relief supplies within days even to the most damaged areas of Tōhoku following its 3/11 triple disasters.

The degree to which all authorities engaged in realistic disaster planning may serve as a final factor that can influence coordination or discord after disaster. While colleagues have referred to such plans as “fantasy documents” (Clarke 1999) because of their tremendous disconnect from reality, some communities have invested time and effort in building pragmatic frameworks for response and recovery.

Using a variety of public sources, including EM-DAT (CRED 2019), secondary and tertiary sources from newspapers, and academic articles, I assembled a dataset on 18 disasters¹ that took place over the past five decades across developing and developed countries alike. I selected these 18 disasters for several reasons. First, they represent a strong cross section of nations at various levels of development, from developing, to mid-range developed, to advanced development. Next, these disasters bring with them a wide range of government capacity and democracy. These disasters draw from a number of regions around the world, including Africa, the Americas, and Asia. Finally, these disasters have a broad range of economic and human costs, from very few deaths to 0.05% of the overall nation’s population. Most of the cases come after 2001 because of the availability of consistent data on these events. Using a team of coders I sought to capture the variables outlined above from secondary and tertiary reports in relatively simple terms, often using a binary outcome but at other times using a continuous scale if it was feasible across all observations. The unit of analysis is therefore the disaster response itself. Table 1 lays out the variables, their means, standard deviation, minimum, and maximum.

Moving beyond a descriptive approach, I use quantitative regression analysis, qualitative comparative analysis, and case studies (in that order) to illustrate how factors may influence coordination or friction in crisis.

One suggested solution to the friction found in many of the cases here and in the broader dataset involves advancing mitigation and response simultaneously in parallel at different levels. Local residents, experts urge, need to live in cultures of disaster while decision makers and national leaders must be part of strong institutions (Muir-Wood 2016). Local disaster culture translates into homeowners, families, and businesses being mindful of threats and risks and creating and drilling crisis responses. Institutional engagement means that decision makers and bureaucrats in national governments create and maintain strong regulations and governance practices that minimize harm. One example would be optimizing the bottom-up and top-down responses in vulnerable, coastal regions of nations with a high likelihood of tsunami occurrence, whether Talcahuano, Chile, Portland, Oregon, or Sendai, Japan. In an ideal world, citizens living in such areas would evacuate each time they felt or were warned of an earthquake and school children and adults alike would regularly practice walking, moving via wheelchairs, or running to higher ground. At the same time central government officials would put in physical infrastructure that effectively mitigated threats while creating escape zones that minimize dangers of inundation. Government agencies would also have pre-placed materials—including rescue, housing, food, and water—out of inundation zones but along resilient transportation routes into these areas.

But it is not immediately obvious that these two factors—bottom-up culture and top-down institutions—are in fact always compatible. This is because national frameworks can ignore local practice and indigenous knowledge may violate standardized laws. First, national, top-down responses may ignore indigenous practices. Hundreds of years ago, Japanese communities erected stones warning their neighbors against building homes at lower elevations near the ocean. For hundreds of years religious institutions in Tōhoku such as shrines and temples were all built above the inundation heights of the 3/11 tsunami. Despite this institutional memory, regional and local zoning laws opened up low lying coastal property for residential development (Takase et al. 2012). Over time, these laws pushed locals to ignore embodied memory and move homes and businesses into vulnerable coastal zones.

Further, national institutions may press localities to take on standardized plans that go against local visions and goals and constrain local culture and innovation. Following the 1995 earthquake, the Kobe city government proposed innovative ways to raise money to create more independence in the recovery process but were not allowed to do so by the central government (Aldrich 1997). Despite a recognition that physical infrastructure may harm tourism and local fisheries, Japan’s central and regional governments have pushed local communities to accept standardized seawall heights at 14.5 meters (Takezawa 2016).

Finally, indigenous knowledge and traditional practice may circumvent national laws or push beyond normative behavior (Mearns and Norton 2010). Local cultures of disaster may embrace events like flooding seen by decision makers as major threats to be managed. The Jamuna char dwellers in Bangladesh, for example, see flooding as a natural part of the life cycle that can bring benefits as well as threats, where Western educated engineers define them solely as risks (Schmuck-Widmann 1998). Similarly, the aboriginal cultures in Australia have long embraced active burning of local landscapes to reduce the likelihood of far more destructive fires (Spillman and Cottrell 2004). Following the 2004 Boxing Day tsunami, the Indian government insisted that all coastal dwellers move back 5 km from the sea in order to receive a new home, but local residents merely kept their old homes along with their new, more distant dwelling (Aldrich 2012).

In short, it may be that friction during disaster response is the norm, not the exception, and any attempts to move forward to reduce it must balance the importance of recognizing indigenous culture and practices whether in Japan, Sri Lanka, Nepal, or the Philippines (Shaw et al. 2009).

This study has used a medium sized dataset of disasters around the world and as a result the dataset cannot be thought of as a representative sample. Because of data availability, most of these events occurred in the early twenty-first century. Further, while we used coders when categorizing our nine variables of interest, future teams should consider machine coding to provide stronger reliability in coding decisions. This is especially true for the outcome of interest, namely the level of coordination (or friction) in the response, which scholars have argued involves subjective judgment (Boin and Richardson 2015). Future studies should simultaneously build up a broader dataset of disasters—perhaps along the structure of EM-DAT (CRED 2019)—while employing teams of simultaneous coders who can seek to accurately capture more ambiguous variables of interest.

This article has sought to use qualitative comparative analysis, Tobit regression analysis, and mini-case studies to illuminate the factors that lead to friction—and not coordination—during disaster response. Across these methods the factors of the economic costs of the event, the level of development in the country, and the strength of logistical infrastructure having proven critical. In short, developed nations facing larger disasters with strong logistical infrastructure seem less likely to encounter friction during crisis. Unfortunately, less developed nations facing smaller scale disasters and using weaker infrastructure are more the norm, and more likely, based on these results, to have a lack of coordination.

The problem of balancing top-down and bottom-up approaches and ensuring smooth coordination will require creative political and administrative frameworks, perhaps ones not influenced strongly by political regime changes (Wood and Waterman 1991). While this article has focused partly on extreme weather-related events (including typhoons and floods), the pressing challenge of climate change will require simultaneous bottom-up adaptation and top-down legislative and institutional efforts that need to be coordinated well to avoid waste and achieve efficiency. One experimental approach to climate change adaptation has involved social learning, knowledge exchange, and social networks across levels of agency (Butler et al. 2015). Similarly large scale, cross-national private sector projects require horizontal and vertical integration of teams across firms and countries, and research has underscored the importance of identification and empowerment of leaders (Muller et al. 2018). Many industries struggle with the attempt to balance inflexible rule frameworks with local needs. In attempts to preserve critical aspects of heritage and culture, for example, conservators in museums and cultural institutions must navigate top-down initiatives from executives with bottom-up mandates from public stakeholders (Stoner 2009).

One cautionary tale about attempts to create coordination through top-down responses, as has often been the norm in developing and developed nations alike, revolves around the importance of bottom-up responses. The Fa’a Samoa, or Samoan culture, provided local residents with a bottom-up, accountable structure for responding to the 2009 tsunami without need for top-down institutions (Rumbach and Foley 2014). Societies that have unbalanced the interaction between top-down and bottom-up policies have paid the price. As societies face extreme weather events and climate change, they must recognize the need to integrate local residents and governments into response and recovery plans.