Abstract
In an 1878 report to Congress, John Wesley Powell unsuccessfully advocated that political jurisdictions in the American west should be organized to conform to watersheds. Instead, governing North America’s water resources remains fraught with cross-jurisdictional challenges, through which water flows with no concern for borders on maps. The Columbia River in the Pacific Northwest is the embodiment of boundary-spanning stories, as its arms cross and re-cross national and international borders on its journey to the Pacific Ocean. The Columbia River Basin spans seven U.S. states, Canada’s British Columbia, and 30 recognized Indigenous tribes on both sides of the international border, with complex social and ecological interactions throughout.
Water governance styles in large river reaches and reservoir basins across the western U.S. and Canada vary greatly, related in part to the magnitude of federal investment in infrastructure and projects for irrigation, hydropower, and navigation. Yet even the more top-down, command-and-control situations have evolved towards collaborative arrangements of government and non-government organizations ranging from federal agencies to regional and local government and advocacy groups. The devolution of powers from a central government to place-based collaborative governance allows for local experimentation, one of the central tenets of adaptive co-management. Nobel Prize winner, Elinor Ostrom, champions this idea of independent, multi-level actors governing collectively in polycentric arrangements. It is still unclear, however, to what extent collaborative governance arrangements lead to positive outcomes in managing the resources.
Interactions among organizations that together govern the resources can be seen as a network, like a spiders web, where actors are the nodes, and connecting threads are avenues of communications between them. Social network analyses are rigorous scientific methods that are often applied to evaluate the structure and interrelations among actors in governance networks. But the relative paucity of comparable datasets leads to gaps in our understanding of networks for governing water resources, such as whether the networks have a consistent or dominant shape or pattern.
This study examines water resource governance networks, with a focus on fisheries and water quality. I seek to answer three central research questions:
1. Is there a dominant shape for water resource governance networks?
2. What qualities determine the position of an actor within a water resource governance network?
3. To what extent can internal dynamics of the water resource governance network be correlated with action outcomes in terms of changes in fisheries or water quality parameters?
I use a mixed-methods approach of personal outreach and an online survey to collect data for lake health metrics, actors in the governance networks, and internal dynamics of those networks in five reservoir basins within a geographic transect of the Columbia River Basin. Lakes Chelan, Roosevelt are in the U.S. state of Washington, Lake Pend Oreille is Idaho, and člq̓etkʷ (Flathead) is in Montana. The fifth reservoir, Lake Koocanusa, is just over half in Montana, with the rest across the international border in British Columbia, Canada. Similarities, such as federal water quality standards, fisheries regulations, and federally licensed hydroelectric dams that impound each basin, provide a consistent regulatory framework for each governance network. Differences in local geography and social contexts contribute to individuality in the networks.
This dissertation is organized into five major chapters, of which the first three constitute the background needed before the analyses. Chapter one describes the methodology for survey development and the questions that are in the survey instrument. I describe the process of scoping to determine the network boundaries and actor lists for each basin. I explain the process of preparing data for analysis, and the analyses that are performed in course of this dissertation. The chapter finishes with an in-depth discussion of complexities in the process of in-person outreach to solicit survey responses, and subsequent richness in qualitative information that adds context to the networks in individual study basins.
Chapter two is a sketch of the geography and biography of each basin in this study. I present similarities among the settings, but also key differences. Qualitative data from both text entry survey responses and personal communications illustrate issues in governance scenarios in individual basins. Some overarching themes, such as calls for better funding, are common to all of the basins governance groups.
In chapter three I conduct a literature review to examine the prevalence of a core-periphery network pattern in water resource and natural resource governance case studies. The 30 pieces reviewed span nearly 60 years of research in both developed and developing nations. I approach the observed pattern from three disciplinary lenses: network structures, internal dynamics of social capital, and polycentric relations of multi-level collaborative governance. While the three lenses involve different scales and theoretical approaches to evaluating governance networks, results show considerable overlap in the concepts. Furthermore, the core-periphery pattern does appear to be commonly observed, which underscores the need for further examination of this phenomenon in governance networks.
Chapter four examines the core-periphery network pattern with a stepwise process of analytical methods. The central premise is that actors’ positions in a network are based on what they do (the number of roles they fulfill in a suite of roles), rather than who they are (the actor type). I construct a suite of roles on three characteristics: resource roles (data, expertise, funding, political support), formal roles (land and resource tenure, regulatory, juridical), and informal roles (collaboration, information sharing by hosting forums and via websites, policy entrepreneurship or proactive networking). With the network data organized by actor types and management focus (all actors, fisheries focus, water quality focus) I first visualize the networks. The patterns are confirmed analytically using a core-periphery test in the social network analysis software. Comparing the network structure by quadratic assignment procedure reveals very low correlation of patterns by actor-type. Poisson regression of actor position (centralization on indegree, or how many nominations an actor receives) shows strong and significant correlations to the three role categories, with individual variations by basin and focus.
Chapter five explores internal dynamics of the governance networks in the five basins relative to changes in lake health indices. Literature for social-ecological systems shows that relating social factors with physical management outcomes remains challenging. Measurements are operationalized using Likert-scale responses from the surveys. Aggregated data across all five basins reveal few discernible trends in lake health outcomes. Strong correlation patterns emerge, however, when network dynamics are related with physical outcomes in individual basins. Qualitative information from the surveys, archival documents, and scoping meetings, supplemented adds basin-specific context and interpretive explanations to these differences.
The five main chapters are followed by discussions of overarching themes, limitations to the study, and future directions. This research makes contributions to both theoretical and practical applications of network analysis. The literature review and subsequent network analyses build understanding of core-periphery patterns in resource governance networks. This cross-basin study adds much needed data to a field that suffers from paucity in comparative, multi-community studies. Using the networks themselves as boundary objects to delimit the basins renders a network analysis approach more accessible to practitioners who wish to improve communications in their basin. Results suggest multiple directions for future studies, such as whether governance networks are actually approaching a polycentric government arrangement, or if they are achieving the goals adaptive management. A bold next step would be to expand this study to other large dam catchments of the Columbia River Basin, and on both sides of the U.S.-Canadian border.