THE DEVELOPMENT OF COLLABORATION OF RESPONSE TO HURRICANE KATRINA IN THE DALLAS AREA SCOTT E. ROBINSON University of Texas at Dallas BRITT BERRETT Medical City Hospital, Dallas, Texas KELLEY STONE Homeland Security - Collin County, Texas Hurricane Katrina placed tremendous stress on America’s emergency response systems. No single agency possessed the capacity required to address the needs of those affected by the hurricane. One element seen lacking in the immediate response to the hurricane was a well-coordinated effort taking advantage of the outpouring of support from distant jurisdictions and organizations. This article looks at some of the theories of how collaborative relationships like these sought in the aftermath of Hurricane Katrina emerge in disaster situations and what factors influence the development of stable emergency response collaborative partnerships. Two case studies in the disaster response in the Dallas/Fort Worth area serve to illustrate the importance of communication and interaction in the development of disaster response networks. The article concludes with thoughts about what questions remain from the case studies and potentially fruitful avenues for research on collaboration and disaster response.

Keywords:

organizational collaboration; emergency response; policy networks

H

urricane Katrina seized the attention of the American public in a way not seen since the terrorist attacks of September 11, 2001. Americans across the country were glued to their television sets watching the horrific images of destruction, deprivation, and desperation from the Gulf Coast. This was a disaster, in this case natural rather than man-made as were the terrorist attacks, that hit close to home. The American people saw their own as victims of this disaster and wondered why little (certainly not enough) had been done to prevent the disaster and what was being done to assist those now in dire need. Americans knew that they were watching history being made as one of America’s greatest cities was decimated. Hurricane Katrina also potentially serves as a signal event in the development of American public administration. For two related reasons, when we look back at the first decade of the

PUBLIC WORKS MANAGEMENT & POLICY, Vol. 10 No. 4, April 2006 315-327 DOI: 10.1177/1087724X06289053 © 2006 Sage Publications

Scott E. Robinson is an assistant professor of political science and public affairs at the University of Texas at Dallas where he researches the management of public organizations and the dynamics of public policy. His public management research focuses on the methods by which organizational decision makers deal with demands for change and conflicting values. Currently, he is developing a large-scale research project investigating patterns of collaboration between schools and other public organizations in the context of emergency preparedness. Britt Berrett has served as president and chief executive officer of Medical City, a 598-bed hospital in Dallas, since April 2000. Prior to joining Medical City, Berrett served as chief executive officer of Sharp Chula Vista Medical Center, a 306-bed facility located in San Diego, and Daniel Freeman Memorial Hospital in Los Angeles, California. He received his bachelor’s degree in finance from Brigham Young University and his master’s degree in hospital administration from Washington University School of Medicine in St. Louis. He is currently a PhD student at The University of Texas at Dallas. 315

316

PUBLIC WORKS MANAGEMENT & POLICY / April 2006

Kelley Stone has more than 30 years of experience in law enforcement and currently serves as the director of Homeland Security for Collin County, Texas. As director of Homeland Security, he is responsible for county facility security, the Fire Marshal’s office, bioterrorism planning and response, and Collin County’s Emergency Management. He received his bachelor of business management from LeTourneau University and his master of business administration from The University of Texas at Dallas (UTD). Currently, he is enrolled in a PhD program at UTD in public affairs.

How government can coordinate the activities of a mobilized community may well become the central question of public administration in the coming years.

21st century, public administration may see Hurricane Katrina as a turning point in public administration. First, after a decade or more of the steady privatization of domestic policy tasks1 to private and nonprofit organizations, the public called for governmental action to address the needs of victims and wondered where the governmental action was that could have mitigated the damage in the first place. When they looked, the public found a public agency in disarray led by a political appointee with little practical experience in the area of disaster preparedness and response.2 Calls for a more professionalized agency were loud and, by and large, univocal. A second reason that the image of the government response to Hurricane Katrina may serve as a seminal moment in the history of American public administration is more subtle. Although the public looked for government responses to the plight of the victims of Katrina, what they found were a variety of nongovernment actors attempting to step in to serve what has traditionally been seen as a governmental role—disaster relief. Whether it was Wal-Mart sending trucks of bottled water or local churches organizing massive efforts to serve victims of Katrina, the public saw a community of private sector, nonprofit, religious, and local government actors (from all around the country) stepping up to respond to the need of Katrina victims (including the needs of those refuges of Katrina who flooded neighboring areas—and some nonneighboring areas). The public saw the potential in this effort but soon realized the complexities of coordinating such an effort. Although the public wondered where government was, they saw the need for a stronger government capable of coordinating the remarkable resources turned out from the nontraditional providers of aid: private businesses offering goods, nonprofit and religious organizations offering volunteers, and local emergency response offices from distant jurisdictions offering trained personnel. The public saw the need for collaboration between a capable government and an interested community of nongovernmental personnel. How government can coordinate the activities of a mobilized community may well become the central question of public administration in the coming years. This is not an entirely new concern to scholars of public administration. Laurence O’Toole (1997) argued that networks were the central method of policy implementation. Later work by Hall and O’Toole (2000) found that networks have long been the implementation instrument implied by national legislation. Work by Provan and Milward (1991, 1995) and Agranoff and McGuire (2003) served to emphasize the importance of networks in public administration and, thereby, the essential need for an understanding of collaboration on the part of public managers. It being clear that collaboration promises to be an integral activity for public management in the coming years, particularly in areas like disaster preparedness and response, this article assesses the origin of disaster response networks. If we are to understand collaborative networks sufficiently to utilize their power to address public problems, we need to understand how these networks are born and develop over time. The next section reviews some of the major theories of the birth of collaborative networks—two of these theories are drawn specifically from the literature of disaster preparedness and response. Following the review and integration of existing theory on the origins of collaborative networks, the article presents two brief case studies of the birth of disaster response networks in the Dallas, Texas, area to respond to the needs of victims of Hurricane Katrina. These two case studies will trace the early history of the Katrina response networks and the role of preexisting relationships on the nature of the Katrina response networks. The article concludes with a discussion of the patterns that emerge from these cases and the potential areas for further research on network origins.

The Theory of Network Emergence The study of collaborative networks is by no means limited to the scholarship in public administration. To illustrate the growing on the emergence of collaborative networks, this section will first review Rosabeth Moss Kanter’s (1994) developmental view of business

Robinson et al. / COLLABORATIVE RELATIONSHIPS

alliances with particular attention to the commonalities between the business alliances that Kanter discusses and the public administration collaborations discussed in this article. Following this review, we review the literature on the development of collaborative networks in disaster response by surveying the works of Louise Comfort (1994) and Saundra Schneider (1995). The conclusion of the theory section integrates these theories to suggest patterns we expect to find in the cases of network emergence in the Dallas area response to Hurricane Katrina. ROSABETH MOSS KANTER AND COLLABORATION AS HUMAN RELATIONSHIPS

Kanter studied collaborative relationships between private businesses—with particular attention to partnerships across national boundaries. What she found is that collaborations are, in many ways, like one of the most well known human relationships—marriage. The most interesting aspect of this analogy is that Kanter identified a time path for collaborative relationships she saw across cases. Kanter (1994) found that collaborations begin with a courtship period. In this period “two companies meet, are attracted, and discover their compatibility” (p. 98). The key for Kanter was that the early period of collaboration depended on the testing of “compatibility on broad historical, philosophical, and strategic grounds: common experiences, values and principles, and hopes for the future” (p. 99). Many relationships are broken off in this reflective stage as the partners seriously consider the relationship and find that they are not sufficiently compatible for a long-term collaboration. The courtship period is followed by a period Kanter (1994) calls “engagement.” Here the collaborative relationship gets serious. Kanter describes parts of this stage as “meeting the family” wherein the leaders of enterprises have to seek the approval from “other people in the companies and of other stakeholders” (p. 100). This period closes with a process of taking “vows” wherein the partners embark on specific, well-defined projects together. Usually these “vows” include explicit recognition of plans to expand the relationship in the future and incorporate “clear signs of continuing independence for all partners” (p. 101). It is only after this initial period that collaborative relationships become formalized and stabilized. Kanter (1994) went on to describe later stages of collaboration (including “housekeeping” and “old marrieds”); however, our focus is on these early stages. Not everything that Kanter saw in the private sector can translate directly into collaborative relationships in disaster preparedness and response. For example, potential partners are not usually drawn from a large marketplace. Most collaborations, in Kanter’s analogy, are like arranged marriages. The only eligible partners in an area often must partner with each other and are sometimes required to do so by hierarchically superior organizations like regional and/or national offices or legislation. Furthermore, disaster situations do not allow for a leisurely courtship as one might encounter during a series of golfing excursions between chief executives in a private enterprise. Instead, partners in disaster response may have to find each other quickly in an uncertain environment. The slow development of a relationship between freely participating partners is unrealistic in this environment. However, much of what Kanter saw in the collaborations between international enterprises has a direct bearing on the dynamics one would expect to see in collaborative partnerships involved in disaster preparedness and response. The image of a hesitant, self-reflective search on the part of potential collaborators seems to be a likely pattern one would see in voluntary or arranged collaborative partnerships. The pace of the movement from courtship to engagement will likely differ—but the nature of the human relationship suggests that we will likely see these stages, at least in truncated form, in emergency response collaborations. What is most important from Kanter is the recognition that collaborative relationships likely follow a time path through a series of stages—proposed to be much like the stages of a marriage relationship. In this respect, one should focus on the age of a relationship to understand key aspects of the relationship—rather than other potential factors such as the nature of the

317

318

PUBLIC WORKS MANAGEMENT & POLICY / April 2006

participants (private enterprises, public agencies, nonrofit and/or volunteer organizations), though this may be important as well. When one recognizes that the developmental stage of a collaborative relationship is a key aspect of that relationship, researchers should turn their attention to relationships at specific stages such as courtship. LOUISE COMFORT AND THE EMERGENCE OF COLLABORATIVE PARTNERSHIPS IN DISASTER RESPONSE

Louise Comfort (1994) analyzed the emergence of collaborative partnerships in the specific context of disaster response, adopting a developmental model similar to that found in Kanter (1994). Comfort sought to address the nature of the emergence of collaborative relationships in more detail and drew on self-organization theory from the natural sciences for assistance. Natural scientists have long wondered why, at times, small changes in the environment can lead to large changes in a physical system (reviewed in Comfort, 1994). One oft-used example is the “sandpile” analogy. If individual grains of sand are added to a sand dune, it is not clear when an individual grain of sand will be enough to cause the sand dune to collapse and reform a new dune. At some point, a single grain will cause the collapse and reformation of large parts of the sand dune. This is difficult to accommodate within the traditional natural science paradigm that assumes that small changes in the environment should create proportionately small changes in the physical system. In this case, a small change (the addition of a single grain of sand) causes a large effect (the restructuring of the physical system of the sand dune). To explain these processes, Prigogine (1984) and other natural scientists developed the notion of self-organizing systems. Systems seek low energy states. However, to move from one state to another, they have to overcome energy barriers. It is easiest to imagine that systems live in energy valleys. The only way to move from one valley to another—even a preferred, lower valley—would be to climb the hills separating the various valleys. For this reason, most systems are stable. They do not possess the energy to climb over the hills to explore other valleys for preferred energy states. However, when some external source adds energy to the system, it can overcome these hills and sometimes find a preferable new state. The single grain of sand does not change the system, it simply adds the final amount of potential energy necessary to overcome the barriers to change and allow the system to enter a new energy state by collapsing. Social theorists applied the same logic to social system through the process of “autopoesis” (Luhmann, 1995). Social systems are organizations that seek ends much like natural systems seek low energy states. The specific equilibrium of a social system is not determined by the will of any one social agent, but by the will of the various members combined in complex ways. In this way, social systems can adapt to changes in the environment by the simultaneous reaction of social agents. These are massively parallel-processing systems (Kauffman, 1993). The result is that societies can be viewed as massive learning systems. Comfort (1994) applied this thinking to the analysis of the Pittsburgh Oil Spill of 1988. Comfort found that many different organizations responded to the oil spill. The respondent organizations varied from private to public to nonprofit organizations. These various organizations spontaneously organized into a collaborative system even without centralized knowledge of who exactly was responsible for which activities. This type of spontaneous organization, autopoetic self-organization, required several key systemic elements. Comfort found that the system relied heavily on five qualities: the structures necessary to hold and exchange information, sufficient flexibility to adjust behavior to changes in the environment, shared goals among the participants, reoccurring opportunities for interaction, and the capacity to integrate evolving information. It was these qualities that allowed the social system to respond dynamically to the sudden change in the environment represented by the oil spill. It is the ability of the various components of the social system to communicate—even if not through a centralized leader or organizer—that allowed the participants to seek a new equilibrium following the environmental disaster. The decentralized communication of participants eventually led to a new equilibrium through the interaction of the diverse participants

Robinson et al. / COLLABORATIVE RELATIONSHIPS

in the emergency response system. Comfort noted that the repeated interaction of these parties was a key to the development of an autopoetic, self-organizing system when she stated: Organizations that worked alone showed less willingness to adjust their behavior. Recognition of long-term association and continuing interaction by participating organizations and ready access to communication facilities are critical elements in this process. Opportunities for reoccurring interaction would have been facilitated by a stronger communication capacity. (p. 405)

This account closely resembles the skeletal theory of the stage of collaboration presented in Kanter (1994). The spontaneous organization of the emergency response system in Pittsburgh depended on the ability of the organizations to enter a courtship—a process of investigating the possibilities for continued relationships. This was only possible when the organizations were allowed frequent interaction in the shared goals of disaster response. The Comfort account of self-organization provided a theoretical structure for this stage of the collaborative process. The courtship stage involves the communication of various components of a social system in their attempt to organize in response to environmental changes—even in the absence of a centralized leader. The Comfort account also focused attention on the importance of multiple parties in a collaborative enterprise—rather than the two-party system imagined in a marriage metaphor. Collaborative partnerships, particularly those addressing public issues (including disaster preparedness and response), may involve dozens of participants— an image poorly accommodated by a marriage metaphor. The combination of the Kanter notion of a developmental process of collaborative partnership with the Comfort appropriation of self-organization theory provides a strong foundation for our understanding of the emergence of collaborative partnerships in disaster response. SCHNEIDER AND THE NORM GAP IN DISASTER RESPONSE

Saundra Schneider (1995) focused on similar themes of spontaneous emergence of social organization in her comparative case study of governmental responses to natural disaster. In her analysis of governmental response to natural disasters, she found that processes of spontaneous organization led to the emergence of norms among the affected populations. Borrowing from sociological theory (Merton, 1957), Schneider argued that mutual expectations about the actions of others lead members of society to form accepted rules of conduct. When natural disasters occur, the previously accepted set of norms can be disrupted. At this point, processes of collective action take over. Groups of individuals seek collective action to establish new norms in the face of the new conditions generated by the norm-disrupting disaster. This collective action process follows through a series of phases of collective action Schneider collectively calls the “emergent norm perspective of collective phenomenon” (from McPhail, 1991). In this model, the affected population passes through four stages of norm development. In the first phase, the affected population engages in “milling” (Schneider, 1995, p. 51). In this stage, people face situations in which their preexisting norms of behavior do not work. The environment does not correspond to the expectations around which the original norms formed. People are then left without a set of norms to guide their actions. The result is that people “mill” about seeking new norms based on their postdisaster sets of expectations. The second phase begins as the milling people begin to interact with each other and spread rumors. It is at this point that people interact with each other and new norms can begin to develop. As rumors circulate and people look for a new set of norms to guide their behavior, some rumors circulate more broadly and influence people more deeply. The processes of selections wherein some rumors become dominate in the population produces “keynoting” (Schneider, 1995, p. 52). Keynoting involves the social identification of the symbols and ideas that will eventually coalesce in a new series of norms. The keynoting separates the dominant themes and symbols from the lesser known (or less accepted) potential themes. In the final stage, the new themes and symbols create a new set of norms that operate within the newly established

319

320

PUBLIC WORKS MANAGEMENT & POLICY / April 2006

postdisaster environment. The disaster shattered the world in which the old norms operated. It is the four-stage process of emerging norms that creates a new set of norms to replace the disaster-displaced old norms. Note how similar this collective-action developmental process resembles the selforganization processes discussed in Comfort and the courtship processes in Kanter. Following some impetus to seek new relationships, potential collaborators hesitantly seek partners. Repeated interaction between potential partners leads to the development of common norms that did not hold in the predisaster environment. It is the interaction between independent agents that allows the new norms to emerge. For Schneider, a key problem in this process is the separation between the bureaucratic operating procedures of governmental agencies (particularly federal agencies) and the norms that emerge locally among victims of disasters. The bureaucrats within disaster response organizations are not generally those people most affected by a disaster. Instead, the bureaucrats handle problems according to historically developed standard operating procedures (p. 46). In that the disruption presented by the disaster is not a disruption to the bureaucratic agency itself, there is little reason for the agency to develop new norms. The result is a situation where the target population for disaster response is developing new norms while the government agencies are operating according to old norms. The result is a norm gap. Where this gap is small, the government agency is able to overcome any gap and provide adequate relief to victims of natural disasters. This was the case Schneider (1995) saw in “normal disasters” like the flooding in South Carolina in 1990 (pp. 134-140). In these situations, the normality of the disaster (how much the disaster is expected, prepared for, and mitigated) prevents the need for radically new norms to emerge in the victim population. The result is a small gap between the norms of the victims and the norms of the agency—and a relatively easy disaster to which agencies must respond. Where the gap is large, governmental response is difficult and often ineffective. Schneider (1995) found a large gap in the traumatic effects of Hurricane Hugo in the Caribbean islands in 1989 (pp, 80-86). Although the Caribbean Islands experience hurricanes with sufficient frequency for hurricanes to be considered “normal” disasters—Hurricane Hugo was so strong that it knocked out the communication systems on which the disaster responses to “normal” hurricanes relied. Without these communication systems, the disaster response systems collapsed and the government was not able to respond in its traditional ways. The gap between the norms that emerged in the population in response to the dramatic hurricane and the topdown expectations of the government was large and the relief efforts were largely considered a failure. The analogy between the governmental response to Hurricane Hugo and the response to Hurricane Katrina is eerily close. Schneider (1995) added an important consideration to Comfort and Kanter’s account of the emergence of collaborative relationships. Schneider pointed out that the emergence of collaborative partnerships and new norms among the victim population can create tension between the victim population and the often-insulated governmental agencies. Without attention to this potential gap, disaster response can be undermined by new norms emerging in the victim population or the inappropriate implementation of disaster response based on norms ill-suited for the postdisaster environment. PATTERNS IN COLLABORATIVE RESPONSES TO DISASTERS

These different theories of the emergence of collaborative relationships agree at a number of key points. The most obvious point of agreement is the hesitant, searching behavior in which potential collaborative partners are likely to engage. While Kanter focuses on the voluntary initiation of a search for a collaborative partner, in all three accounts the participants are left to search with little assurance of what they will find. The uncertainty is amplified in the disaster situations wherein old structures and norms have been disrupted. Without familiar guideposts for organizations in seeking partners, collaborators have to make up new norms as they go along.

Robinson et al. / COLLABORATIVE RELATIONSHIPS

The emergence of new norms is similarly a common theme across all three accounts. Whether the new norms emerge from collective behavior (as in Schneider’s account), from the spontaneous search for a new stable equilibrium (as in Comfort’s case), or from the drafting of new vows following a courtship and engagement period (as in Kanter’s account), collaborative relationships create demands for new ways of carrying out organizational tasks and new ways of interacting with organizational environments. Such a process closely resembles what scholars of policy change would classify as a punctuated equilibrium (Baumgartner & Jones, 1993). A long period of stability—characterized by a well-known and constant set of expectations about social behavior—is interrupted by some dramatic, usually external, event. This interruption destabilizes expectations and initiates a search for new expectations. Only when there is consensus on the new expectations is a new equilibrium established and a new era of stability begun. The three theories of collaboration anticipate a punctuated equilibrium of collaborative relationships. Dramatic disruptions (such as natural disasters) can disrupt old relationships. The partners in these relationships then engage in searches for a new set of relationships to establish a new equilibrium. The next section reviews two cases of disaster response in Dallas, Texas, with particular attention to the affect of Hurricane Katrina on the disestablishment of prior relationships and the search processes of a variety of organizations looking for new collaborative partners and a new stable equilibrium of service delivery.

Dallas Responds to Hurricane Katrina—Two Cases When Hurricane Katrina hit the Gulf area, Texas was inundated with people fleeing the destruction in the hurricane’s wake. This forced systems in the Dallas, Texas, area to respond to the needs of the people displaced by the hurricane—a task of unanticipated scale and urgency. The section reviews the efforts of Dallas area organizations to respond to the needs of those displaced by Hurricane Katrina in two areas. The first case reviews the work of a Dallas area county Homeland Security office to coordinate efforts by local religious and relief organizations. This case provides a “government’s eye view” of the relief effort from the perspective of a public agency. The second case reviews the works of hospitals in the Dallas/Fort Worth (DFW) metro area to address the needs of some of the most vulnerable people displaced by the hurricane. In each of these cases, collaboration was a key element in the response to Hurricane Katrina. The case studies are organized to emphasize the state of collaborative partnerships before the hurricane and an account of the evolution of collaborations in the immediate aftermath of the hurricane as, for various reasons, organizations search for new collaborative partners to address the needs of those displaced by Hurricane Katrina. COLLIN COUNTY’S HOMELAND SECURITY OFFICE

Collin County is a rapidly growing and urbanizing county of approximately 660,000 residents situated immediately north of Dallas County. In 2002, Collin County recognized the growing importance of emergency management and established the Collin County Homeland Security Office (CCHLS). The new department was tasked with creating coordinated efforts between similar agencies at the local, regional, and state level. In addition to emergency management departments, CCHLS initiated collaboration between nongovernmental organizations (NGOs) and other government agencies, such as the local school districts. Although Collin County is not prone to catastrophic events, emergency management plays a key role in daily activities. One of the organizations within the newly designed CCHLS was the Collin County Bioterrorism Response Team (CCBRT). The CCBRT observes and evaluates public health threats ranging from rabies to anthrax on a daily basis. An established relationship between medical providers, including hospitals, the Dallas/Fort Worth Hospital Council (DFWHC), Collin County Hospital Committee, the medical examiner, and private medical practices, transmit data to the CCBRT epidemiologist for statistical analysis of infectious

321

322

PUBLIC WORKS MANAGEMENT & POLICY / April 2006

diseases. Analysis of the data allows for a compressive summary of any potential threats to be disseminated to the public. In the event of an infectious disease outbreak, the CCBRT, under the guidance of CCHLS, would open and operate dispensing sites to provide prophylaxis and/or vaccinations to the general public. An extensive amount of collaboration between multiple agencies at all levels is necessary. At the federal level, the Centers for Disease Control would coordinate with the Texas Department of State Health Services to dispense prophylaxis and/or vaccinations through the Strategic National Stockpile Program (SNS) to state, regional, and local governments. The State Operations Center (SOC), operated by the Texas Department of Public Safety (DPS), would coordinate dispersal throughout affected areas. The DPS houses the Governor’s Division of Emergency Management and the Texas Homeland Security Director, both of which are involved during emergency operations. As established during previous events, the SOC would coordinate daily conference calls for all regions affected. Such collaboration allows for consistent data dissemination to regional and local governments. From the SOC, information and resource operations are handled regionally by Department of State Health Services (DSHS) officials and the Disaster District Committee (DDC). The DDC handles state emergency management, whereas DSHS handles public health events within this framework. Although state and federal collaboration is required, local coordination is the key to successful response efforts. Inter-local agreements between Independent School Districts and Collin County allow the use of school buildings for SNS dispersal, along with sheltering and mass care operations. Predetermined agreements reduce the amount of time required to allocate resources. In addition to the public sector, NGOs provide needed resources for emergency management. Currently, the Collin County Community Organizations Active in Disasters (COAD) centralizes NGO activity in the county. The COAD is in the early stages and has only recently established by-laws and elected officers; however, CCHLS has been involved during the entire process allowing for a strong tie between the two. Within the COAD are representatives from agencies that provide volunteer coordination to amateur radio operators that assist in communication needs. The American Red Cross (ARC) and the Salvation Army are other NGO agencies that provide services during an emergency. Under the provisions of the Collin County Emergency Operations Plan (EOP), the ARC is responsible for assisting the county with shelter and mass care needs.

Four days after Katrina’s landfall in the Central Gulf Coast, Collin County Judge Ron Harris received notification that Collin County should anticipate approximately 5,000 persons needing shelter.

Hurricane Katrina coordination and collaboration. In the days following the second landfall of Hurricane Katrina in the central Gulf Coast, Texas initiated shelter operations. The SOC contacted the DDC who contacted local governments to begin preparation for incoming persons evacuated from the central Gulf Coast, primarily the New Orleans area. The DFW metroplex was designated as a metropolitan area to receive evacuated individuals from the New Orleans area. Prior to the decision to open shelter operations in the area, thousands of persons who had evacuated prior to Katrina’s landfall had entered the area to stay with friends and family or at local hotels. With such an influx of persons, the Governor’s Division of Emergency Management (GDEM) coordinated statewide efforts at the SOC in Austin. Dallas, Tarrant, Denton, Collin, and neighboring counties were designated to receive busloads of New Orleans and surrounding community residents. Four days after Katrina’s landfall in the central Gulf Coast, Collin County Judge Ron Harris received the notification that Collin County should anticipate 5,000 approximate persons needing shelter. On September 2nd at 0930 CDT, Collin County opened its Emergency Operations Center (EOC) to begin planning and operations for incoming buses. Prior to arrival, the evacuees would arrive in the Dallas suburb of Mesquite at the Big Town Center Mall to be processed and routed to appropriate shelters. Because of the overwhelming nature of the situation nationwide, the ARC, which was designated in the EOP to operate county shelters, was absent in the process. Therefore, the immediate concern for Collin County was the location and facilities to be used. Several prospects were suggested; however, not all were sufficient and feasible for the situation. The primary location became a privately owned facility in the small rural community of Princeton, the Lake Lavon Baptist Encampment. A verbal contract between the county and the owner of the

Robinson et al. / COLLABORATIVE RELATIONSHIPS

Encampment allowed the partial use of the facility to shelter persons. Food and medical services were provided and reimbursable by the county through FEMA assistance. The facility provided a capacity of 150; therefore, additional facilities were necessary. The City of Wylie, operating under the Collin County Emergency Plan, designated its National Guard Armory as a second shelter facility. Because of the nature of the facility, the SOC had to approve the location along with the National Guard. Acceptance of the Armory was declined until industrial air conditioning units were allocated. Fortunately, county and city coordinated efforts resulted in the completion of the task, and the National Guard Armory became an acceptable solution. With regards to the large cities of Collin County, Allen, Frisco, McKinney, and Plano, only one additional shelter was activated by local government. The City of McKinney converted a recently vacant Wal-Mart facility into a shelter site. Conversion of the vehicle maintenance bay to bathrooms and showers and additional structural changes allowed for temporary shelter. As stated before, numerous persons were residing in hotels, and with the determination by FEMA and the ARC that the influx of persons in hotels would overwhelm shelter operations nationwide, a voucher system was initiated that allowed for reimbursement of room costs. Such actions at the federal level provided relief for local operations. In addition to the influx of persons via bus, the faith-based organizations began transferring persons to government shelters. Even with the end of faith-based shelters, the organizations provided volunteer support for the government shelters. Volunteers from all agencies and individual volunteers provided irreplaceable support to county operations. The entire shelter team at the Lake Lavon Baptist Encampment comprised volunteers and was assisted by county personnel. County personnel from several departments, including, but not limited to Auditing, Purchasing, Public Health, Homeland Security, Fire Marshal, Information Services, and Public Works, all handled EOC operations. Liaisons from other jurisdictions and agencies were provided for EOC operations. Staffing and overtime costs for the operation were reimbursable for 100% of cost due to a federal Disaster Declaration for the 254 counties of the State of Texas (Disaster Declaration FEMA 3216 EM TX). Sufficient staffing and volunteer response allowed for an adequate transition for evacuated persons upon arrival to Collin County shelters. At shelters, persons were provided with sleeping quarters and communal space for living areas and meal times. Additional medical needs that were not addressed at the processing center were handled by small clinic operations at each shelter. The City of McKinney’s Wal-Mart facility opened a clinic for daytime use by persons that were evacuated but did not stay at one of the shelters. Local physicians and emergency care facilities, such as E-Care in McKinney, provided additional medical needs. Other needs that were addressed included medication provisions and special needs patients. Dependent on the special need, all patients were transported to proper facilities to remain. For persons at the shelter that had alternative locations to stay, such as family members in other locations, transportation was provided for one-way travel at no charge. National Travel System, a private travel agency, was contracted by the state to coordinate travel arrangements requested by the EOC. Travel programs allowed persons to reunite with family elsewhere and alleviate the number of persons requiring shelter. Nearly 3 weeks after the opening of shelter operations in Collin County and many other counties, long-term temporary housing was facilitated. Nearly a dozen apartment complexes and private rental homes were allocated for an initial term of 6 months for all remaining sheltered persons. More than 45 contracts were entered into by Collin County. Special addendums and waivers of deposits on rented units and utilities were necessary. Transition to the new rental units began the day before Hurricane Rita prompted evacuation in the Texas Gulf Coast region and brought more persons to the Collin County area. The number of persons was reduced at the Lake Lavon Baptist Encampment and the Wylie Armory had closed operations; however, the City of Plano provided a short-term shelter at its Community Center for a capacity of nearly 400 persons. After the shelter operations ceased and the EOC was closed, CCHLS and other departments remained active in reimbursement paperwork and procedures, along with activities

323

324

PUBLIC WORKS MANAGEMENT & POLICY / April 2006

regarding persons in rental units. Monthly check-ins and constant case management for remaining evacuated persons allowed for easy transition for persons as conditions at their former residence improved. Several persons that arrived because of the destruction of the Gulf Coast have chosen to remain and become residents of the DFW metroplex. DFW METRO HOSPITALS

Relationships that were developed prior to the 2005 Katrina hurricane were critical to respond to the issues that emerged as the impact of the disaster unfolded.

The delivery of hospital health care in the United States is unique. More than 5,000 hospitals are in operation throughout the country, and they collaborate and coordinate their efforts in a chaotic and seemingly disorderly manner. Absent from the American landscape is centralized public health care planning for the delivery of services and the coordination of resources. Although other countries have pursued single-payer economic models to respond to escalating health care expenses, the uniquely American approach has been to allow the free-market to reign. Several statewide efforts have created Certificate of Need (CON) bureaucracies that have attempted to purposefully regulate and manage the proliferation of new health care assets and equipments. Notably, CON efforts are exclusive to capital planning and do not include health care services delivery or coordination of services. This lack of public health care planning at the national level is also absent at the regional level and is restricted to sporadic state efforts. Because of the lack of national health care planning and coordination, some of the void has been filled by the creation of integrated health care systems (IHS) by previously unaffiliated hospitals and health care organizations. A reaction to imperfect federal funding, the IHS is typically the consolidation of multiple hospitals within a geographical area. Efforts are expended to consolidate duplicative services and drive economic improvements throughout the organization. An unintended benefit of the consolidation of hospitals is the development of relationships among the entities. Although only a few regional integrated health care services organizations have achieved national scope, the majority are relegated to regional and state coverage. Notably, proprietary hospital companies have achieved national coverage as part of a business-planning model focused on responding to acquisition opportunities nationally. HCA, based in Nashville, Tennessee, is the largest of these proprietary entities and operates more than 180 hospitals from Richmond, Virginia; to Miami, Florida; to San Jose, California; to Anchorage, Alaska. Centralized planning is conducted through the home office with regional leadership directing multistate efforts. Because of the lack of national public health care planning and chaotic and fragmented health care planning efforts, regional health care collaboration has developed through a number of informal and unstructured relationships. Professional organizations and not-forprofit associations within the health care field fuel much of the collaboration and relationship development between hospitals. The American Medical Association and the American Nursing Association are examples of two trade associations that develop relational bonds between individuals within the health care delivery system. State and regional efforts such as the Texas Hospital Association (THA) or the DFWHC also assume the responsibility to respond to the demands of their member hospitals and have demonstrated their ability to fuel collaboration and coordination among member hospitals. The relationships developed in these organizations create the foundation for collaboration that is called on during moments of crisis. Collaboration and coordination in the moment of crisis. Relationships that were developed prior to the 2005 Katrina hurricane were critical to respond to the issues that emerged as the impact of the disaster unfolded. Within hours from the collapse of the levee system, hospitals throughout the New Orleans basin realized that all disaster planning and preparedness that was conducted on a local basis was completely and utterly futile. It became clear that key administrative leaders would need to call on regional and national relationships previously developed to respond to the cascading series of events. Because of the scope of the disaster,

Robinson et al. / COLLABORATIVE RELATIONSHIPS

regional assistance became critical. Tulane University Medical Center made the call to HCA headquarters in Nashville, and within hours emergency and evacuation plans were launched. The Medical City transport team in Dallas was immediately dispatched to the area and retrieved nine neonatal intensive care unit babies from hospital rooftops. By Thursday of that week state and federal agencies began to respond to the crisis but were nationally criticized for their lack of preparedness and responsiveness. Moving parallel with local evacuation efforts were state organizations such as the THA and the DFWHC. The executive director of the Hospital Council recognized the magnitude of the disaster and immediately implemented the local Hospital Emergency Incident Command System (HEICS). A command post was established in the Hospital Council chambers in Dallas, and resources were identified by member hospitals that would be available for immediate transportation to the site of the disaster. In addition area hospitals coordinated efforts in preparation to receive victims of the hurricane. As evacuees inundated the DFW metroplex, local agencies and organizations immediately responded. Health care clinics were established at the Reunion Arena and the Convention Center in downtown Dallas by highly motivated health care workers. Requests were made for supplies such as bandages, pharmaceuticals, and communication systems that were without delay approved. Within the week the appropriate health care–planning organizations such as medical schools and public hospitals had marshaled the resources to prepare for the deluge of evacuees; however, in those wee early hours the coordination and collaboration was shouldered by individuals Clearly the regional and national response to the health care needs during the disaster was chaotic. Health care issues were addressed by local New Orleans health care staff with no regional or national direction or support. Personal relationships were called on to respond to immediate needs and limited regional and national responses were driven by ownership such as HCA. Following close behind the Katrina hurricane was the much-anticipated destruction of hurricane Rita. Once again, state agencies such as the THA assumed responsibility to manage and direct the chaos. Daily conference calls were established to communicate critical information among hospitals. Startling was the lack of familiarity that existed from one region to the next. Determining availability of resources and assessing needs became a key function of the THA. As information was difficult to assess and requests for assistance became frenzied, the THA functioned as the intermediary. At one point prior to the actual landfall of Rita, calls were made by hospitals from the Houston, Texas, area requesting assistance for transfers of critical care patients. These calls were redirected to the THA and the command center of the D/FW Hospital Council to facilitate and manage resources. It was clear that the lack of familiarity with available resources and an evident crisis mentality drove hospitals on the Gulf Coast to attempt to establish untried relationships.

Lessons From the Dallas Response to Hurricane Katrina The preceding cases serve as illustrations of the development of collaborative relationships in the Dallas area as it responded to the needs of those displaced from the Gulf Coast. These cases are not a comprehensive review of all responses or the development of all partnerships intended to address the needs of the people displaced by Hurricane Katrina. However, these cases do illustrate common patterns in the development of partnerships as organizations quickly realized that no one organization could adequately handle the needs of those arriving in the Dallas area. The hesitant search for partners following a disruptive event is clearly evident in both of these cases. In the case of the CCHLS, the primary coordinator of disaster response was taken out of the Dallas area response process by the dire need in the Gulf Coast itself. The result was a local disaster response system that had to spontaneously reorganize and seek partners to provide services to those arriving from the Gulf Coast. The days following the influx of

325

326

PUBLIC WORKS MANAGEMENT & POLICY / April 2006

Planned local responses were incapable of handling the disaster on their own.

people from the Gulf Coast were then filled with attempts to build new relationships, for example, with the Wal-Mart and the Baptist Encampment. These relationships were informal and limited at first (note the verbal nature of the agreement with the Baptist Encampment) as the partners were hesitant in defining their role in the partnership and cautious about committing to new collaborative relationships. Similarly in the case of health care organizations, the early reactions to the needs of those arriving from the Gulf Coast were chaotic. Although the regional hospital organizations provided some structure to the initial response of health care organizations, the collapse of systems in the Gulf Coast left many health care providers without an organization on which to rely. The result was the reported reliance on nontraditional health care networks dominated by personal connections rather than organizational plans. However, the shadow of the past loomed large in the development of the collaborative responses to the needs of people arriving from the Gulf Coast. It seemed at times, that the theories reviewed above suggested a clean slate on which organizations drew up new collaborative partnerships. Although the absence of the past may have been an analytical exaggeration to focus attention on the development of relationships following a disrupting event, the case studies suggest that past relationships serve as a starting point for many of the searches for new collaborative partners and provide support for organizations that have to seek new local partners. In the CCHLS and health care organization cases, state and regional organizations provided support in coordinating efforts while new local collaborative relationships were forged to accommodate immediate local needs. The role of regional and state organizations is particularly interesting in the context of localized disasters. Local organizations in the Gulf Coast area were quickly overwhelmed as transportation and communication systems were brought down by the hurricane and its aftermath. Planned local responses were incapable of handling the disaster on their own. Even in Dallas, where the hurricane itself did not hit, the influx of people from affected areas quickly overwhelmed the sheltering and medical care capacities of the area. The development of new local relationships occurred within the context of established regional organizations—state emergency organizations and regional health care organizations in these cases. This introduces an element of consistency to what might otherwise be a purely chaotic search for a new equilibrium following an event that fundamentally disrupts the social system. Although the collective searches for new norms described in the literature occurred in both cases, then, it was complemented by consistency in an overarching set of organizations and norms. Although these norms may potentially create conflict (in the norm gap of which Schneider, 1995, warns), they can also provide a skeletal structure to hold up local efforts as they seek stabilized local relationships. FUTURE RESEARCH IN THE DEVELOPMENT OF DISASTER RESPONSE COLLABORATIONS

The case studies bring out areas where further theoretical development is needed. The cases largely confirm the hypothesis that collaborations begin with a hesitant search for potential partners, especially following a disruption in the social system. At the local level, the cases confirm the punctuated equilibrium dynamics implied by the theories of the emergence of collaborative relationships. Future work should trace the persistence of these relationships over time. What is the probability that relationships forged in a crisis such as this persist and become formalized? How likely is it that the verbal agreement with the Baptist Encampment becomes a more formalized agreement that can persist for years, and crises, to come? What qualities predispose a relationship to persist where other relationships do not? The cases studies also raise questions about the role of overarching regional and state organizations in the development of new local relationships. Both cases reported continuity in regional and state relationships in the midst of the efforts to build new local relationships. The mixture of local discontinuity (punctuated equilibrium) in the presence of regional continuity suggests that regional organizations can be influential in the process of developing new local

Robinson et al. / COLLABORATIVE RELATIONSHIPS

partnerships. The stable regional relationships can serve as the bedrock for organizations seeking new partners. This may introduce further dangers of Schneider’s norm gap or it may facilitate the sort of communication and interactions that Comfort (1994) emphasizes are so important to the development of spontaneously organized systems. These questions suggest that there is a lot to learn about collaborative public management in disaster response. Although there is a lot of work ahead, the potential contribution of scholarly understanding of collaboration to the practice of disaster response is tremendous. The events following Hurricane Katrina make it clear that our disaster response systems have a lot to learn about development of collaborative partnerships to respond to disruptive events. Better collaboration seems a key step in making sure that services can be delivered in a timely and comprehensive fashion to the victims of future disasters.

Notes 1. There has been a notable empowerment of governmental authority in the area of foreign affairs, military intelligence, and counterterrorism. However, this stands in contrast to the move away from government in domestic affairs represented by the popularity of such movement as welfare privatization, charitable choice, and school choice. 2. The resignation of the director of the Federal Emergency Management Association (FEMA) soon after Hurricane Katrina, and the widespread criticism that led to this resignation, point to the general social dissatisfaction with the management of the preparation for and response to Hurricane Katrina.

References Agranoff, R., & McGuire, M. (2003). Collaborative public management: New strategies for local government. Washington, DC: Georgetown University Press. Baumgartner, F. R., & Jones, B. D. (1993). Agendas and instability in American politics. Chicago: University of Chicago Press. Comfort, L. K. (1994). Self-organization in complex systems. Journal of Public Administration Research and Theory, 4(3), 393-410. Hall, T. E., & O’Toole, L. J. (2000). Structures for policy implementation: An analysis of national legislation, 1965-66 and 1993-4. Administration & Society, 6, 667-686. Kanter, R. M. (1994). Collaborative advantage: The art of alliances. Harvard Business Review, 72(4), 96-109. Kauffmann, S. A. (1993). The origins of order: Self-organization and selection in organization. New York: Oxford University Press. Luhmann, N. (1995). Social systems (J. Bednarz, Jr., Trans.) Stanford, CA: Stanford University Press. (Original work published 1984) McPhail, C. (1991). The myth of the maddening crowd. New York: Aldine du Gruyter. Merton, R. K. (1957). Social theory and social structure. New York: Free Press. O’Toole, L. J. (1997). Taking networks seriously. Public Administration Review, 57, 45-52. Prigogine, I. (1984). Order out of chaos. New York: Bantam. Provan, K. G., & Milward, H. B. (1991). Institutional level norms and organizational involvement in service implementation networks. Public Administration Review, 64, 681-693. Provan, K. G., & Milward, H. B. (1995). A preliminary theory of network effectiveness: A comparative study of four health systems. Administrative Science Quarterly, 40, 1-33. Schneider, S. K. (1995). Flirting with disaster: Public management in crisis situations. Armonk, NY: M.E. Sharpe.

327