Differentiating between urban flood risk as a unitary problem and as a strand in a braided problem set: Implications for administrative coordination

Flood protection is a leading priority for urban water sustainability. Making cities more resilient to flooding has become urgent as the climate changes and as cities increasingly become the loci of human population and resources. Reducing the risk of future flooding in cities often necessitates different jurisdictions working together. They may do so because they confront a shared problem. This was the case in the City of Lincoln, Nebraska, USA, when partnering agencies shared a single focus on reducing flood risk from Beal Slough to the Nebraska State Penitentiary. In contrast, entities may band together to confront braided problems, intertwined problems that cannot be resolved independently. The Antelope Valley Project, also in Lincoln, Nebraska, USA, combined addressing three problems, for which individual solutions had not been achieved: reducing flood risk from Antelope Creek, improving road transportation safety and capacity, and revitalizing neighborhoods with deteriorating physical structures. Such a scenario is becoming more frequent as cities increasingly face multiple demands on the same location. As this comparative case study demonstrates, there are implications for administrative coordination for whether flood risk reduction can be achieved as a sole focus of a project and when it cannot. The Antelope Valley Project necessitated an innovative management structure and governance process that the Beal Slough Project did not. In the Antelope Valley Project three different, stand-alone entities operating in dissimilar, substantive domains redirected their independent policies to harmonize their problem solving. Collaborative learning among policy actors in the Antelope Valley was more extensive and across policy domains than was required in the Beal Slough Project. Yet, both projects were triggered by policy-oriented learning from the remapping of their respective subwatersheds. The study contributes to understanding interdependency among policy problems and to reducing urban flood risk through administrative coordination.


Introduction
Climate change and urbanization are anticipated to increase catastrophic flooding worldwide [1].Floods are the most frequent natural hazard around the world [1,2] with the widest geographical distribution [2].In the United States, floods occur throughout the country and cause more losses than any other natural disaster.Most of the federally declared disasters list flooding as a leading or consequent source of damage [3].Flooding can be disastrous on its own and can make other calamities worse [4].U.S. federal assistance is not available for over half the floods that cause damage in the country [3].
In addition, many inland cities are experiencing and are anticipated to experience effects of climate change, including more frequent and intense storms in which there is high rainfall amounts over shorter time durations [3].There is the long-recognized threat of fluvial flooding when the capacity of a river channel is exceeded.There is mounting concern about pluvial flooding, when drainage capacity, whether engineered or natural, is exceeded [5].As global climate changes and as cities increasingly become the loci of human population and resources, making cities more resilient to flooding becomes more pressing [5,6] and is key to achieving urban water sustainability [7].
To adapt to increased flooding risk requires innovative strategies to make the built environment more resilient [8].One means to do so is through effective and appropriate administrative coordination for developing urban flood management infrastructure.Administrative coordination occurs through reforms improving how organizations interact and cooperate with each other to develop and execute policies [9].How entities interact and what they achieve jointly varies as a function of the form of their collaborative endeavor [10].Therefore, there is no one form of administrative coordination that is appropriate for all urban flood mitigation projects given how widely they vary in scope, scale, and intent [8,11].An underexplored consideration is whether an urban flood mitigation project can be undertaken by itself or must be executed in tandem with other urban infrastructure projects.In part, this is because of a well-developed trend in urban flood risk management scholarship to focus on reducing flood risk in cities as a self-contained concern [12][13][14][15][16].
One of the core reasons for entities to cooperate is to shape the set of conditions causing, exacerbating, or allowing a problem to continue [17].Sometimes a set of conditions is sufficiently disconnected from the wider context to produce a unitary, shared problem.At other times it is not, producing a set of braided problems.
Explicit [18] and implicit [19] in the collaborative governance literature is that entities join forces to address shared problems.Yet, entities may work together because their problems are sufficiently intertwined with other problems beyond their authorities they can only resolve them in concert with resolving other problems.While the nature of the problem set bears on how entities coalesce and what they accomplish, the distinction between a shared set of problems and a braided set of problems has scarcely been acknowledged, let alone probed in the scholarship related to urban flood risk reduction.
Consequently, this paper asks two, paired questions.(1) What are the circumstances in which urban flood risk reduction can be pursued and realized in isolation?(2) What are the circumstances in which policy actors responsible for reducing urban flood risk must engage with related issues that may appear to fall outside their purview?This comparative case study of two urban flood mitigation projects in Lincoln, Nebraska, USA identifies how entities work together to address unitary, shared urban flood mitigation problems and urban flood mitigation problems interwoven with other urban management problems.In doing so, it contributes to understanding a key characteristic of policy problems, interdependency among policy problems [20].The notion of braided problems in which a problem may be defined independently but not resolved independently is the basis for a simple typology that offers guidance in deciding which administrative coordination strategy to pursue.
The next section considers interdependency as a policy problem characteristic, as this is fundamental to understanding the utility of distinguishing between a shared problem and a set of braided problems.A description of the case study approach follows, before the two case studies are presented.The case studies are then used to understand how to characterize collaborative elements of administrative coordination, how learning serves administrative coordination and how mapping provides a means of policy-oriented learning.The concluding discussion highlights the wider implications of distinguishing between a shared problem and a set of braided problems.

Interdependency as a policy problem characteristic
A problem is the discrepancy between an intolerable and a tolerable situation.A policy is a commitment to a course of action by an entity or entities with the authority to impose a course of action within one or more jurisdictions.Interdependency is an important characteristic of policy problems [20] since problems are commonly interlinked within a system [21].Brown [22], Ackoff [23] and Peters [24] observe that policy problems do not exist independently since they act with others as components of a structure of problems.
Ackoff's depiction of messes provides an important historical touchstone in thinking about interconnected problems [20,24,25].Ackoff [23] (21) coined the term a mess to describe a system of problems, ". ..a system of external conditions that produces dissatisfaction."Ackoff [26] (99) went on to explain that messes are "dynamic situations that consist of complex systems of changing problems that interact with each other".
Interdependence increases the complexity of resolving problems by expanding how many organizations need to be involved [27].The more interdependent problems are the more difficult they are for politically created coalitions to resolve.Managing interdependent problems creates a need, at a minimum, for coordination and potentially for integrating policies and organizations [28].The greater the interdependency among problems, the more likely organizations will have to develop new routines to address them [29].
The extent of interdependence among policy problems can be understood by considering if a problem can be defined independently or must be defined in conjunction with other problems and whether or not it can be resolved on its own or must be resolved in conjunction with other problems.See Table 1.A stand-alone policy problem can be defined and resolved independently.The term, braided set of problems, captures how under certain, specific circumstances, there are interdependent policy problems that, on one hand, can be defined independently and, on the other hand, cannot be resolved independently.While it is possible to articulate separate policy problems in a braided problem set, each of them can only be resolved as part of a package where all of them are resolved.This contrasts with Ackoff's [23] notion of messes in which interdependent policy problems can neither be defined nor resolved separately [24].In policy making, coordination has been regarded as different policy actors working together to achieve a common end [30,31].Consequently, the distinction between working on a shared, stand-alone problem and working on a set of braided problems, arguably, has not been highlighted to the extent it warrants.The assumption of a shared, single problem may have resulted from the lack of theoretical grounding for policy problem classification schemes [24] or theory that placed an emphasis on other dimensions of explanation (for example Stone [32]).Regardless, the assumption of a shared, single problem has been widespread [24,[32][33][34][35][36][37][38][39][40][41].For example, Gray [18] (5) notes "parties who see different aspects of a problem can constructively explore their differences and search for solutions that go beyond their own limited vision of what is possible".

Case study approach
A pilot case study approach was employed to enable an in-depth multi-faceted investigation of phenomena within real world contexts in which they occur, thus providing a rich appreciation of complex situations, historical specificity [42,43] and policy subtleties [44].A case study approach enabled producing tangible context-dependent knowledge [45].The case studies were purposely selected to acquire salient information that could not have been attained in such richness otherwise [46].
The case studies investigated constitute best case scenarios for addressing problems through interlocal cooperation.They address what Rittel and Webber [47] would describe as tame rather than wicked problems and what Dunn [20] would describe as well-structured rather than ill-structured problems.Uncertainty is less of a consideration for tame or well-structured problems while it is a defining characteristic of wicked or ill-structured problems [17].The well-structured problems of the case studies had adequate information to alleviate them, clear goals as to what needed to be addressed about them, and had reasonably well scoped arrays of possible solutions, which could be framed in the same political terms as the problems being addressed [48].The capacities of the entities involved and the projects' resources were sufficient to solve the problems [17].The problems were amenable to resolution through negotiated agreement because they were straight forward.The issues involved were broadly recognized with sufficient clarity to act upon them [21].The projects were adequately resourced and operated within a facilitating legal framework and lacked a history of conflict between key actors and organizations [17].Neither project was impeded by lack of commitment by participating entities, dominant actors manipulating interactions or distrust among participants [49].There was consensus on how to define and address the problems in the two case studies.
The instrumental case studies were intended to contribute to a broad appreciation of the phenomena [50] of addressing problems through interlocal cooperation.Of interest was learning about and explaining causal links and pathways [42] developed as a consequence of highly focused, interlocal initiatives.
A longitudinal case study approach employing archival documentation was used to advance understanding of collaboration dynamics and process performance [17].For both the Beal Slough Project and Antelope Valley Project case studies, information from the City of Lincoln and the Lower Platte South LPSNRD Archives was utilized.In addition, for the Antelope Valley Project documents from the University of Nebraska-Lincoln Archive were reviewed.

The case studies of the Beal Slough Project (BSP) and the Antelope Valley Project (AVP)
Flood mitigation provided the common, topical content for the Beal Slough Project (BSP) and the Antelope Valley Project (AVP).Intensification of urbanization upstream directly contributed to increasing the flood hazard threat in both case studies.Human actions in flood prone areas, such as concentrating economic activity or escalating settlement density, increase exposure and vulnerability to flooding [1,51].Urban flood risk is exacerbated when the rate of new development exceeds that of drainage infrastructure upgrades [5] and when urban land cover is altered by expanded built areas and increased surface imperviousness [52].
The most common cause of flooding is when natural water courses don't have capacity to convey excess water produced during heavy rainfall [2].Both case studies involved interventions to increase the conveyance capacity of natural water courses to reduce exposure and vulnerability to flood events [1,53,54].Flood defense measures were undertaken to reduce flooding by keeping water away from people [1,[54][55][56].In both case studies, intergovernmental cooperation was required as different governmental levels were responsible for financing, constructing, and maintaining the infrastructure for the respective flood defense projects [54].
Local entities needed to cooperate to address improvements to Beal Slough to increase its carrying capacity: the Nebraska Department of Correctional Services (NDCS) with its responsibility for the penitentiary, the City of Lincoln with its responsibility for floodplain regulations and transportation, and the Lower Platte South Natural Resources District (LPSNRD) with the lead responsibility for the Beal Slough channel under the terms of a 1963 agreement with the City.The NDCS didn't have the statutory authority to undertake flood reduction for this project on its own.It was, however, able to collaborate on the project through the authority to enter into interlocal agreements granted by the Nebraska Interlocal Cooperation Act.A simple, interlocal agreement was negotiated.Costs were estimated and allocated to each partner.The LPSNRD was assigned to be the lead agency.There was no joint activity or sharing of authorities.Through the Nebraska Emergency Management Agency (NEMA), Federal Emergency Management Agency (FEMA) funds from a previous disaster were applied to this project [57,58].
The Beal Slough Project (BSP), Lincoln, Nebraska, USA had a primary focus, reducing the flood risk from Beal Slough to the Nebraska State Penitentiary.Mitigating the flood risk from Beal Slough was an urgent priority because the penitentiary housed an incarcerated population.It was considered preferable, if possible, to avoid the need to evacuate the penitentiary due to a flood event [58].BSP resulted in incidental flood protection to upstream private properties and improved transportation infrastructure.The project involved enlarging and stabilizing the channel of Beal Slough, removing and/or replacing three bridges and constructing minor levees in a one kilometer reach of the stream.It took six years from planning to completion.The total cost of the project was $7.3 million, with 75% of the funding from a FEMA Hazard Mitigation Grant administered by NEMA, 8.3% from the Nebraska Department of Correctional Services (NDCS), and 8.35% respectively from each of the City of Lincoln and the Lower Platte South Natural Resources District (LPSNRD) [57].Table 2 lists expenditures and funding sources.The four partnering agencies operated through an interlocal agreement made possible by the Nebraska Interlocal Cooperation Act with the LPSNRD as the lead partner [57].
The Antelope Valley Project (AVP), also in Lincoln, Nebraska, USA was necessitated by three problems unsolvable independently that brought the City of Lincoln, the Lower Platte South Natural Resources District (LPSNRD) and the University of Nebraska-Lincoln (UNL) together [60][61][62].The LPSNRD's efforts to reduce the flood risk from Antelope Creek, the City's efforts to improve the transportation network to increase capacity and safety, and the University's efforts to shift arterial traffic out of campus and to continue campus expansion to the east had all been unsuccessful as individual projects.At the same time, the built urban community east of the University was experiencing blighted conditions.The flood problem meant the impacted neighborhoods were economically depressed, in part because the flood risk made property owners reluctant to invest in the housing stock [61][62][63].These multiple problems and circumstances were all concentrated geographically in the lower Antelope Creek Watershed [60,63].
In the AVP, three different, stand-alone entities, the City of Lincoln, the Lower Platte South Natural Resources District and the University of Nebraska-Lincoln, operating in dissimilar, substantive domains redirected their independent policies to harmonize their problem solving.This was accomplished through creating, implementing and sunsetting a new joint separate entity, the Joint Antelope Valley Authority (JAVA), endowed with salient, necessary powers from each of the partnering entities and created by the three collaborating partners solely to implement the AVP.JAVA was delegated or authorized ". ..any power or powers, privileges, or authority exercised or capable of being exercised by any individual Partner. .." and ". ..any governmental service, activity or undertaking, which at least one of The Partners is authorized by law to perform. .." [64] (6).
In an effort to be as inclusive as was achievable within time and financial constraints, the Antelope Valley Project partners appointed a nearly-seventy-member, diverse citizen advisory group that was actively engaged through the five years of planning, meeting at least monthly.The group included representatives from many different constituencies, including residential and business neighborhoods, UNL students, community organizations, federal and state agencies, and staff and administration of the partners [63,66].In addition, the partners' consultant conducted a broad, public involvement campaign.
An extensive study and a public involvement process were integral to planning and designing the project on several square kilometers east of the City's downtown and of the (UNL) City Campus.Major components of the project included three kilometers of new and reconstructed open waterway for flood control, twelve traffic and pedestrian bridges, ten kilometers of new roadways, five kilometers of new bicycle trails, new and expanded parks and recreation facilities, new business and residential construction and neighborhood infrastructure upgrades [65,66].The project took twenty-three years.The total project cost was $241 million [59].Table 2 lists the expenditures and sources of funding.The LPSNRD, the City, and UNL entered into an interlocal agreement through the Nebraska Interlocal Cooperation Act and created the Joint Antelope Valley Authority (JAVA) as a joint separate entity, delegating the implementation of the AVP to JAVA with authorities granted to it by the partnering entities.JAVA can be understood as a policy venue created because there was not an appropriate one in place.It employed rules promoting consensus and mitigating conflict to ensure each of the problems in the braided problem set was solved through to completion.Also involved in planning, designing, and funding of the project were three federal agencies: US Army Corps of Engineers, Federal Highway Administration, and Housing and Urban Development.Once the AVP was completed, as per the interlocal agreement, JAVA was terminated with ongoing responsibilities assigned back to the local partners [64].
The timelines for Beal Slough and Antelope Valley Projects are laid out in Table 3.The project activities occurred in four discernable, chronological partnering phases identified by the LPSNRD general manager as pre-partnering, informal, formal and post project.The pre-partnering phase involved necessary, precursor activities to conceptualizing the projects.The informal phase involved activities that set the stage for establishing formal commitments among project participants.The formal phase occurred when interlocal cooperation agreements were in place.Post project phase activities were activities that resulted from the project having taken place and were the responsibility of individual entities operating independently.Both projects arose in an ad hoc, evolving manner.In the BSP, the City and the LPSNRD in 2006 cosponsored a feasibility study and analysis of potential alternatives to alleviate flooding from Beal Slough, including the State Penitentiary area [68].Following the study, the City, LPSNRD, and the NDCS began meeting with the common goal of reconstructing the waterway along Beal Slough adjacent to the Nebraska State Penitentiary.The objective was to avoid devastating impacts a major flood would have on the Penitentiary.Collateral benefits were flood protection for adjacent properties along Beal Slough and improvements to transportation on 14 th street [57,58,68,69].Participants worked together to solicit funding from the State, FEMA, and their own entities.In April 2008 they entered into an interlocal agreement under the Nebraska Interlocal Cooperation Act [57].
In the AVP, the City, the UNL, and the LPSNRD administrators recognized collaboration was needed because the University, the LPSNRD, and the City had overlapping jurisdictions, responsibilities, programs, and administrative structures.The City, the University and the LPSNRD while pursuing their individual projects, had been informally communicating and interacting with each other [61,85,86].Several formal agreements were approved between the City and the LPSNRD to cooperate on the repairs to the underground conduit, a key element of the flood mitigation infrastructure for Antelope Creek, and between the City and the University for the Malone neighborhood redevelopment [77,79].In 1992, the executives of the three entities agreed to form an Antelope Creek Partnering Group (ACPG) to continue informal cooperation on studies [60].
The process the ACPG went through to generate a problem statement and recommend a conceptual framework for developing a transportation/flood control/revitalization plan included each local partnering agency independently listing its individual needs and then merging them into a set of shared goals and objectives and mission statement.See Table 4. Another element of discovery was the entities learning about potential federal partners, their programs and funding, and the roles they would play in the process.The anticipated involvement of funding and design programs of several federal and state agencies further spurred local cooperation [60,87,88].The process reached a critical milestone in a local partnering charrette on March 11-12, 1993.The charrette resulted in a report, The Antelope Creek Basin Development Plan Problem Statement, which presented consensus goals and objectives for land use, circulation, image, operations and management, and planning processes.It also articulated a vision for the project and conceptualized five project alternatives [60].
In 1995 the City of Lincoln, the University of Nebraska-Lincoln and the LPSNRD entered into a formal agreement to cost share the Flood Control Feasibility Study [78].The informal partnering group continued through the planning process from 1995 to 2000, when the formal interlocal agreement creating JAVA was approved.AVP functioned during the five years of structured planning through a Management Committee, including staff from various City and University departments, from the LPSNRD, and from the federal agency partners, facilitated by a consultant team under City contract.They met monthly to make critical planning decisions, receiving valuable input from the Citizen's Advisory Committee, referred to earlier.When there was a change in the top leadership of one of the participating entities a meeting was held to evaluate the status and direction and reach interim consensus.The governing bodies of the three partners met jointly three times during the planning phase to take interim go/no go decisions on the full AVP [63,74,75,89,98].
Procedural decisions are critical in addressing a set of braided problems to keep all partners at the table until everyone's problems are resolved to their satisfaction.Without such an arrangement, an entity could leave the partnership once its problem(s) gets resolved, even if others have not had their problems resolved.The AVP partners understood the potential challenges the project faced due to the lengthy implementation phase, to the changes in elected officials and corresponding shifts in priorities that could occur, and from budget or funding issues.So, they provided certain conditions in the interlocal agreement to ensure continuity through completing the project, including institutional restrictions on the structure and function of the JAVA Board of Directors.Specific elements included a JAVA Board made up of administrative staff of the partners and not appointed or elected officials, a 100% quorum requirement along with unanimous decisions, formal bylaws, annual budgets and audits, and withdrawal disincentives [64,91,93].JAVA was structured to take into account the differences between the partners and their procedures and institutions, to create a common, centralized approach, thus creating a business-as-usual approach for JAVA.For example, the partnering entities centralized a number of procedures and functions in JAVA rather than each entity performing them under their own rules and policies.These included right-of-way and relocations, administration and accounting, award of contracts, and professional services [64,93].JAVA was brought into existence with a sunset clause, so from the outset the partnering entities decided which entity would take on which responsibility post JAVA [93].JAVA was terminated in 2016 [104].

Understanding Beal Slough and Antelope Valley Projects as collaborative, administrative arrangements
Both case studies can be understood, using Gray's [18] early typology of collaborative forms, as emerging out of optimism, employing collective strategies, beginning with a clear, if mundane vision, and recognizing the necessity and value in working together rather than being constituted as a result of conflict.Collaboration occurs when entities engage informally and formally over time through repeatedly negotiating, developing, and executing obligations [105].Successful collaboration is predicated on each entity being able to further its own goals [106] and achieving better performance or lower costs than if it had acted alone [107].
Beal Slough and Antelope Valley Projects were site-specific undertakings [108].Each focused on planning and implementing around a specific set of on the ground circumstances [17].Each project was formed through voluntary, self-initiation by their respective partnering entities, rather than being independently convened or externally directed [17].Each faced an acute policy (physical problem) challenge, with a clear, bounded geographic focus and authority was concentrated in the coordinating authorities responsible for the decision making [17].
Cross-boundary collaboration takes place among parties from dissimilar authorities, sectors, or bodies [107,109].Guided by coming together around arrays of values, customs, roles, and ways of interacting, governance is how participants in a specific, public policy domain deliberate, decide, exert authority and influence, delegate command, act and oversee implementation [110].In a regime, participants' expectations about a concern coalesce through a governing arrangement instilled with decision-make processes, rules, customs and ethics [111].
Building on the above terms, Emerson, Nabatchi and Balogh [112] (6) define a collaborative governance regime (CGR) as a "particular mode of, or system for, public decision making in which cross-boundary collaboration represents the prevailing pattern of behavior and activity".Emerson and Nabatchi [17] (210) explain CGRs are "fluid, dynamic systems of processes, relationships, capacities, and functions" composed of procedural and institutional arrangements.To identify and organize the concepts for understanding CGRs, Emerson and Nabatchi [17] developed an integrative framework for collaborative governance applicable across a range of settings necessitating cooperation among heterogeneous entities.
Beal Slough and Antelope Valley Projects exhibited four distinguishing features of CGRs.They were oriented to broad public policy or service, involved a range of autonomous entities with differing interests, created institutional and procedural norms and rules and had recurring interactions over an extended time frame [17].See Table 5.

Learning in the service of administrative coordination
Collaborating across entities necessitates creating and using new, pooled information [17].Leach et al. [113] (592) define learning in the context of public administration "as the process by which people develop a more comprehensive and accurate understanding of the science, technology, law, economics, and politics that underlie the decisions they make or the recommendations they advance."The need for collaboration grows as a function of increasing complexity and interdependence of institutional infrastructures and as knowledge becomes more specialized and dispersed [17,49].
In addressing a shared problem, the requisite knowledge may come predominantly from a single lead entity with substantial expertise in the shared problem.The LPSNRD led the BSP because of its flood control expertise.Not all parties have to contribute technical knowledge.For example, in the BSP neither the NDCS nor NEMA contributed flood control expertise while the former gave financially and the latter provided oversight for funds originating from FEMA [57].The deliberation process for the BSP focused primarily on design alternatives, costs, and funding for the shared flood control problem.All parties were familiar with floodplain regulations and flood control works, while recognizing the lead expertise of the LPSNRD [57].In the BSP the partners selected a preferred solution, allocated costs and arranged responsibilities for project implementation to achieve their shared goal of reducing the flood threat to a specified area.In doing so, they removed the Nebraska State Penitentiary from the 100-year floodplain [57,58].The interlocal agreement for the BSP [57] outlined the roles and responsibilities of the partners, their financial shares in the project and designated the LPSNRD to administer implementing the project.Unlike for the AVP, no novel administrative structure was created, and no subsequent, substantive decisions were needed.
In addressing a set of cross-sector braided problems, it is unlikely the expertise for addressing all the problems will be housed within a single entity.A wide range of pooled knowledge is required, since the problems can be substantively different from each other.Consequently, to address their problem, experts in one field must negotiate accommodating knowledge very different from their own.This happened in the AVP.LPSNRD and University representatives had limited understanding of traffic engineering and yet they needed to understand the City transportation engineers sufficiently to be able to address the problems for which they were responsible.LPSNRD brought flood management expertise to the partnership, which the University did not have and the City had, as noted above, delegated to the LPSNRD in a 1963 agreement.The University team brought institutional campus knowledge, including about campus facilities, such as playing fields, and experiential knowledge of student pedestrian behavior, which the City and LPSNRD did not have.The City's authority and expertise were in transportation, floodplain regulation, urban land use, including zoning and community revitalization.Clear and complementary domains of expertise, led to partners being willing to leave issues to subject matter experts.
Addressing a braided problem set benefits from a rigorous, up-front process in which all entities clarify the problems each need addressed and maps these problems relative to the other problems as a prerequisite to developing joint or coordinated solutions.Ideally, participants prioritize championing the complete package of solutions.While problems may vary greatly in terms of scale, resources to address them, costs of ameliorating them, technical complexity and the vociferousness of their advocates, to achieve success all problems require equal standing in the collaboration.The consensus requirement in the AVP meant community redevelopment had veto power, notwithstanding 50% of the project cost was transportation infrastructure alone [65].Inevitably, in addressing a set of braided problems there will be unequal understanding of each problem among participants even while outsiders hold all participants accountable for the entire undertaking.Participants in collaboration may confront reduced control over what happens and murky accountability [114].
In both cases, the value of leveraging the relevant, substantive expertise existing in collaborating entities became apparent.It is important to acknowledge the technical contribution among participants inevitably will be uneven, driven as it is by specific, situational requirements.At the same time, each entity has a vital role in ensuring the collaborative process works for all engaged in it.

Remapping as policy-oriented learning
While a disaster as a major perturbation external to a policy system is a common pathway to policy change [115], the remapping of the Beal Slough and Antelope Valley floodplains exemplify policy-oriented learning as a pathway to policy change.In both cases, the remapping documented a greater area would be exposed to the flood threat than had previously been thought.FEMA developed Flood Insurance Rate Maps (FIRMs) throughout the country as flood prone communities were identified.Each FIRM represents what the risk was when the study on which it was based was conducted.Not incorporated are factors likely to increase flooding in the future, including the consequences of future urbanization, development, and climate change [3].
The BSP was necessitated by the findings of a 2005 FEMA floodplain remapping project under the Digital Flood Insurance Rate Map Program.The entire watershed had become urbanized, and with urbanization, the total volume and the timing of the runoff increased significantly.In the previous mapping of the Beal Slough Floodplain, the State Penitentiary had been only partially in the 100-year floodplain.In the updated floodplain map the entire facility was included, with both expanded extent and flood elevations .6 to 1.2 meters higher, putting the penitentiary in harm's way.Improvements to Beal Slough to increase its carrying capacity would take the State Penitentiary back out of the floodplain.This was important to do because the State of Nebraska realized it would be challenging to evacuate the penitentiary quickly because of the incarcerated population inside [58,71].
The AVP was necessitated by the findings of a routine, second generation FEMA floodplain remapping project in 1986.The remapped floodplain was greater in extent than the floodplain that had been mapped previously, with increased base 100-year flood elevations, and without a designated floodway.This resulted in a significant, heightened risk to the area [60,84].
The remapping of the Beal Slough and Antelope Valley floodplains resulted in up-to date and accurate assessments of the physical exposure to the flood hazard in each setting.The legal, economic, and political ramifications of the new technical information for each floodplain made the status quo no longer acceptable.The new information from the remapping resulted in altered intentions about pursuing policy objectives and as such constitutes policyoriented learning [116].That there was the necessary coordination between the authority generating the flood hazards maps and the authorities able to act on the information acquired cannot be taken for granted [117].Policy learning here is instrumental, as the information generated from the remapping exercises was used to inform new decisions and actions required to mitigate the flood hazards as now understood [118,119].The instrumental use of information refers to when a direct connection can be discerned between information and its specific use in problem solving or decision making [119].

Concluding discussion
With the appreciation of the value of integrating flood control measures into urban planning [120] comes the need to consider what form of administrative coordination will be required.The Beal Slough Project employed an interlocal agreement, an established procedure in the State of Nebraska, to create and employ a straightforward administrative mechanism to achieve the particular goal of flood mitigation in a specific subwatershed.The AVP required a more innovative approach to administrative coordination than BSP because it necessitated forging a new policy subsystem to address the braided problem set.That problems in a set of braided problems are distinct means the range of considerations is broader than in addressing a shared problem.Coordination in addressing braided problems is paramount.Keeping everyone engaged in the process until all problems are resolved is critical because partial solutions may result in some constituencies getting nothing.Competition among problems, each with its own constituency is a real prospect.As a result, consensus arrangements are important in assuring project completion.
Flood protection is a leading priority for urban water sustainability [7].Consequently, there is potential value in conducting a cross-case investigation with more urban flood management case studies or using a case-oriented approach to probe more cases with wider variation in administrative coordination than were considered in this proof-of-concept exercise.For example, BSP and AVP have the properties of unitary subsystems, with strong cohesiveness and coordination, a great deal of independence from other subsystems and structured, amicable decision-making venues.Conditions would be different in collaborative subsystems and even more different in adversarial subsystems [121].
As floodplain remapping revealed the need for BSP and AVP, future floodplain remapping around the world may reveal changes, including anthropogenic climate change, requiring new adaptation and mitigation.Ideally, remapping will benefit from increasing accuracy, stemming partly from adding years to the rainfall record into calculating flooding probabilities, from advances in probabilistic flood forecasting [122] and analytical strategies, such as projections and probabilities for planning water infrastructure [123].
When policy actors are not animated by solving the same problem, the deliberation process benefits from including structured exchanges to facilitate solving each of the separate problems for which partners may have non-overlapping constituencies and interests.This makes how problems get prioritized and how to frame issues important.Rather than pitting one issue against another, the intent is for those affected to see merging points and interdependences.How the exchange of information get structured becomes key in enabling policy actors to recognize and appreciate the links between their entities' problems and those of other entities.
A key implication of this work is the value of a simple delineation between a largely, standalone public policy problem and one that is part of a braided problem set.Recognizing at the outset whether a flood mitigation problem can be addressed independently or not is to acknowledge from the beginning how administratively complex its resolution likely is to be.Distinguishing between a standalone problem and a braided problem set is to recognize whether the parties involved are focused on a shared problem or are giving time to a problem that is not their primary concern as a means to solve another problem that is.This recognition points to when reducing flood risk in cities can and cannot be treated as a largely autonomous matter.
Regardless of where it is generated, new knowledge about evolving flood threats generated through routine mapping can be a sufficient spur to undertake extensive and long-term flood mitigation.While there continue to be valuable and innovative non-structural approaches to flood mitigation there are circumstances, such as in urban corridors where space is limited, for effective and appropriate use of structural solutions.Establishing the administrative coordination necessary to undertake such solutions takes considerable time and resources.It works best when there is a good fit between the capacities of entities and the responsibilities assigned to them.Creating administrative coordination mechanisms intended to exist in perpetuity is not always advisable.Therefore, it is constructive to recognize why and when it makes sense to create a mission executing entity that only exists for the duration required to fulfill its mandate.Completing the BSP and AVP highlight the necessity of negotiating the requisite aspects of the specific contexts in which flood mitigation projects are made necessary.
Even in best case scenarios, problems are not solved indefinitely, situations change, and problems get redefined or morph [23,34,124].Flood risk changes all the time for many reasons, including changes in land use and patterns of rainfall [3].Concurrently, collaborative processes evolve [107].So much of flood risk management in the context of sustainable, resilient cities requires jurisdictions to cooperate.Consequently, it is worthwhile to refine our understanding of how this occurs with the intent of achieving the most appropriate, effective practices and intended outcomes.

Table 1 . Distinguishing among policy problems based on whether or not their definition and resolution are independent or interdependent of other policy problems.
https://doi.org/10.1371/journal.pwat.0000090.t001