Policymaking through participatory processes

Material Information

Policymaking through participatory processes the driving role of policy beliefs at the Colorado Basin Water Roundtable
Lynn, Jewlya
Publication Date:
Physical Description:
xix, 241 leaves : ; 28 cm.


Subjects / Keywords:
Water-supply -- Government policy -- Citizen participation -- Colorado ( lcsh )
Water rights -- Government policy -- Colorado ( lcsh )
Political planning -- Citizen participation -- Colorado ( lcsh )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Thesis (Ph. D.)--University of Colorado Denver, 2010.
Includes bibliographical references (leaves 229-241).
Statement of Responsibility:
by Jewlya Lynn.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
707489366 ( OCLC )


This item is only available as the following downloads:

Full Text


POLICY.t\L\KING THROUGH PARTICIPATOR'{ PROCESSES: THE DRIVING ROLE OF POLICY BELIEFS .\ T THE COLORADO BASIN WATER ROUNDTABLE by Jewlya Lynn B.A., L'ninrsity of Nebraska, Lincoln, 2001 A thesis submitted to the l'niYersitv of Colorado Demer in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Affairs 2010


Tills thesis for the Doctor of Philosophy degree by Jewlya Lynn has been approved by Peter deLeon Danielle V arda Paul Teske Don Klingner Reagan Waskom ') (J /J '1.11 I Date


Lynn, Jewlya (Ph.D., Public .-\ffairs) Policymaking Through Participatory Processes: The Driving Role of Policy Beliefs at the Colorado Basin Water Roundtable Thesis directed by Full Professor Peter deLcon ABSTRACT The Colorado Basin Roundtable is a participatory policymaking process composed of representatives of water policy and management interests. This thesis argues that the public participation literature has excelled at understanding processes like these, by exploring hO\v individuals should have opportunities to interact and make decisions '"ithin these processes. However, tlus body of theory has stopped short of understanding the role of interests and representation in a participatory process, due to its focus on "public," most broadly defined. \'<'here public participation fails, policy literature can add in its understanding of interests and coalitions; yet, policy theorists are focused on the political context and thus largely ignore the importance of the design of the process. Tlus thesis brings the two bodies of literature together to propose a new model for looking at participatory policymaking processes that includes interests and interest groups. A participatory process can include individuals appointed to the process who represent interests and whose interactions with each other are influenced by their interests, not just the design of the participatory process. To explore these concepts, the study of the Colorado Basin Roundtable collected survey information from 160 respondents, including 46 on the roundtable and outside stakeholders. Questions on beliefs were analyzed using k-means cluster analysis to create the belief coalitions, and the remaining data was used to better understand interactions within the beliefs coalitions and at the roundtable. Membership in belief clusters was found to have a significant relationship to reports of information sharing and trust with other members of the roundtable. However, membership in belief clusters was not found to have a relationship with key demographics often used to select members of a participatory policymaking process, such as age, gender, and organizational affiliation. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. Signed Peter deLeon


ACKNOWLEDGMENT Thank you to the Colorado Interbasin Compact Committee and the I and Outreach Subcommittee for funding the research leading to this dissertation. Thank you to all of the stakeholders who participated, including stakeholders throughout the state \vho helped in the design of the surYey and interpretation of the findings and the 160 stakeholders who took a lengthy surny that asked many personal questions. Without extensive participation from Colorado's \Vater stakeholders during eYery stage of the research, from original de,'elopment of the research questions and sun-ey through interpretation and presentation of the fmdings, this thesis would not haYe been possible. Thank you most of all to the Colorado Basin Roundtable for not only participating in the study, but also using the results to build and irnpro,-e their participatory process. Thank you to Rebecca Kahn and Dr. Lyn Kathlene for their support 111 recruiting participation from the many water stakeholders. Thank you to Dr. Peter deLeon for both patience and guidance and the rest of my dissertation committee for staying engaged and helpful. Thank you to the friends and family 'vho tolerated many weekends of library time instead of joining them as they went hiking, camping, biking, snowboarding, and all of the other things that one should do while living in Colorado.


TABLE OF CONTENTS Figures ............................................................................................................................... x Tables ............................................................................................................................... xi CHAPTER 1. INTRODL1CTION ......................................................................................................... ] Research Purpose and Research Questions ................................................................. 5 Research Setting ............................... : ...................................................................... 9 Organization of the Dissertation ................................................................................. 15 2. WATER POLICY IN COLORAD0 ......................................................................... 17 The Lay of the \'\later .................................................................................................... 18 Consumptin Water Uses in Colorado ................................................................ 20 Non-Consumpti,e Water Usc in Colorado ........................................................ 22 The Colorado Basin ................................................................................................ 25 Water Quantity and Future Projections .............................................................. 26 The l,a\vs of the \'Vater ................................................................................................. 28 Water Law and Non-ConsutnptiYe l'ses ............................................................ 30 1"he I.ords of the \'Vater ................................................................................................ 33 How do the Stakeholders Interact? ...................................................................... 37 \YJhat can Policymakers do? ................................................................................... 38 The Future of the Water ............................................................................................... 40


The Colorado Basin Roundtable ................................................................................. 43 Conclusion ...................................................................................................................... 4..J. 3. Lll'ER1\"fl1RE RE\'IEW ............................................................................................ ..J-6 The Roundtables as a Public Participation Process .................................................. 48 Participation as a Necessary Component of Democracy .................................. 48 r-.todels of Public Participation .............................................................................. 52 Defining Success in Participation Processes ....................................................... 56 Representation in Participatory Processes .......................................................... 61 The Roundtables as a Public Policy Process ............................................................. 69 The Contribution of Public Policy Theory ......................................................... 71 Shared Beliefs and Public Participation ............................................................... 78 The Research Propositions .......................................................................................... 79 Proposition 1 ........................................................................................................... 81 Proposition 2 ........................................................................................................... 82 Proposition 3 ........................................................................................................... 83 Conclusion ...................................................................................................................... 8..J. 4. I\fE1'HODS ..................................................................................................................... 86 Research Approach: Participatory Action Research ................................................. 86 Identifying Data Needed .............................................................................................. 89 Data to L'nderstand Beliefs ................................................................................... 91 Data to L'nderstand Demographics ..................................................................... 93


Network Data to l'nderstand the Relationships at the Roundtable ............... 94 Data to l'nderstand the Perceptions of the Roundtable Process .................. 101 Piloting the SurYey ................................................................................................ 1 02 Collecting the Data ...................................................................................................... 1 08 Description of Subject Population ..................................................................... 109 l)ata Collection ..................................................................................................... 11 0 .\nalyzing the Data ...................................................................................................... 111 Analysis Strategy for the Belief Clusters ............................................................ 114 .-\nalysis of Demographics ................................................................................... 115 i\nalyses lising the Net\vork Data ..................................................................... 115 .\nalyzing the Net\vork Variables ....................................................................... 120 Analysis of Perceptions of Success .................................................................... 121 Interpreting the Rcsults ........................................................................................ 121 \' aliditY and Lirnitations .............................................................................................. 122 Litnitations ............................................................................................................. 125 Conclusion .................................................................................................................... 129 5. AN,\J .)'SIS .................................................................................................................... 130 Developing Belief Coalitions ..................................................................................... 130 Results of the Cluster Analysis ............................................................................ 131 Comparing the Clusters ....................................................................................... 13 7 Summary of Agreement on Belief Statemems .................................................. 14 7 Vll


Summary of Disagreement on Belief Statements ............................................. 14H Demographics of Each Belief Coalition ................................................................... 149 Comparing .\ffiliation to Clusters ...................................................................... 152 Relations Among and Between Belief Coalitions ................................................... 156 Belief that Other Roundtable l\Iembers Share Goals ...................................... I 57 Frequency of Information Exchange between Roundtable l\lembers .......... 159 Importance of Information Receind from Roundtable l\Iembers ............... 161 Trust Roundtable Members to Keep Interests in l\Iind .................................. l 6.) Confident Roundtable Members will Follow through on a Commitment ... 165 Influence of Each Roundtable l\Iember in \Vater Policy and 1\Ianagement 167 Belief Coalitions and Perceptions of Success .......................................................... 170 Perceptions of Colorado Basin Roundtable Success ....................................... 170 Perceptions of HB 11 77 Process Success .......................................................... 17 3 ,Exploring the Clusters ................................................................................................ 176 Protecting ConsumptiYe Needs .......................................................................... 1 77 Protecting .\gricultural Needs ............................................................................. 180 Balancing N ceds .................................................................................................... 183 Protecting Non-Consumptin Needs ................................................................ 185 Current SYstem Broken ........................................................................................ 188 Conclusion .................................................................................................................... 190 6. CONCLliSION ........................................................................................................... 191 \'ill


Examining the Propositions ....................................................................................... 192 Exploring Proposition One ................................................................................. 195 Examining Proposition Two ............................................................................... 197 Examining Proposition Three ............................................................................. 198 Implications of the Findings ...................................................................................... 200 Policy Implications for the Colorado Basin Roundtable and other Participatory Processes ....................................................................................................................... 202 Grounds for Further Research .................................................................................. 204 APPENDIX .\. FINJ\L COLOR.\DO B.\SIN ROCNDT.-\BLE SURVEY ............................... 208 B. NORMAL DISTRIBL"TIONS ................................................................................. 228 ..................................................................................................................... 229 lX


2.1. HB05-1177 Water Basins ............................................................................................... 19 4.1. Example roster for answering the information exchange network question ........ 101 5.1. !\leans on Question 1 .................................................................................................... 139 5.2. l\Ieans on Question 2 .................................................................................................... 140 5.3. !\leans on Question 3 .................................................................................................... 141 5.4. l\1eans on Question 4 .................................................................................................... 142 5.5. l\leans on Question 5 .................................................................................................... 143 5.6. !\leans on Question 6 .................................................................................................... 144 5.7. l\Ieans on Question 7 .................................................................................................... 145 5.9. l\Ieans on Question 9 .................................................................................................... 147


LIST OF T.-\.BLES 3.1. and Roh's (2008) hypothesized effects of resources and trust on the strength of cooperatin relationships benveen policy actors ............................................... 77 4.1 .. -\daptation of nenvork measures ................................................................................. 1 06 4.2. Example Nenvork l\Iatrix: Influence Question for Cluster 1 ................................. 118 4.3. Example calculation of a\erages for the Influence Question from Cluster 1 members to all other members of Cluster 1 ......................................................... 118 4.4. Example calculation of averages for the Influence Question from Cluster 1 members to Cluster 2 members ............................................................................. 119 5.1. !\leans and standard deviations for Cluster 1, Protecting Consumptive Needs ... 133 5.2. Means and standard deviations for Cluster 2, Protecting Agricultural Needs ...... 134 5.3. Means and standard deviations for Cluster 3, Balancing Needs ............................. 135 5.4. l\leans and standard de,iations for Cluster 4, Protecting Non-Consumptive Needs ..................................................................................................................................... 136 5.5. l\Ieans and standard deviations for Cluster 5, Current System Broken ................. 137 5.6. Significant Differences Benveen Clusters on Question 1: HB 1177 Should Support Water Transfers to High Growth Areas/Sectors ................................................ 139 5. 7. Significant Differences Benveen Clusters on Question 2: HB 1177 Should Support Protecting the Agricultural Economy and Way of Life ...................................... 140 5.8. Significant Differences Benveen Clusters on Question 3: HB 1177 Should Support Protecting the Recreational Economy and its Water Needs .............................. 141 5.9. Significant Differences Benveen Clusters on Question 4: HB1177 Should Support Protecting Ecosystems and Non-Human Species (or just the Environment) 142 Xl


5.1 0. Significant Differences Between Clusters on Question 5: HB 1177 Should Support Increasing Cooperation :\mong \'\'ater Basins ..................................... 143 5.12. Significant Differences Between Clusters on Question 7: HB 1177 Should Support Balancing Water Supplies and Demands ............................................... 145 5.13. Significant Differences Between Clusters on Question 8: H B 1177 Should Support Protecting Existing IndiYidual Water Rights ........................................ 14(> 5.14. Significant Differences Between Clusters on Question 9: HB 1177 Should Support Allocation and Management of \Vater Resources through the l\Iarket ..................................................................................................................................... 147 5.15. Age by cluster ...................................................................................... 150 5.16. Gender by cluster membership* ................................................................................ 151 5.17. Owns indiYidual water rights by cluster membership*' .......................................... 151 5.18. Represents an organization who owns \Vater rights by cluster membership* .... 152 5.20. i\Yerage Response on "extent to which you belie,'e the following roundtable members agree with or disagree with your goals for the roundtable process" (Scale: -2 =Strongly Disagree to +2 =Strongly Agree) .................................... 158 5.21. ANOV A results from responses on "extent to which you belien the following roundtable members agree with or disagree with your goals for the roundtable process" ..................................................................................................................... 159 5.22. AYerage Response on "how often in the last year you haYe exchanged information related to water policy and management issues" (Scale: 0 = Ne,er to 4 = Daily) .............................................................................................................. 160 5.23. ANOVA results from responses on "hO\v often in the last year you haYe exchanged information related to water policy and management issues" ....... 161 5.24. AYerage Response on "extent to which the information you receive from each roundtable member is important, that is, helps you achieYe your goals for water policy and management issues" (Scale: 0 = Not at all important to 4 = \'cry important) .................................................................................................................. 1 62 5.25. ANOV,\ results from responses on "extent to which the information you recciYe from each roundtable member is important, that is, helps you achieve your goals for water policy and management issues" .................................................. 1 6.) xu


5.26. Average Response on "Extent to which you would trust the following roundtable members to keep your interests in mind on water policy and management issues, regardless of whether he/ she shares your interests" (Scale: 0 = NeYer to 4 = Completely) ............................................................................... 164 5.27. ANO\'A results from responses on "Extent to which you \Vould trust the following roundtable members to keep your interests in mind on water policy and management issues, regardless of whether he/ she shares your interests" ..................................................................................................................................... 165 5.28. Average Response on "confident each of the following roundtable members will follow through on a commitment" 1-'-'rom each Cluster to Each Cluster (Scale: 0 =Not at all confident to 4 =Very confident) ..................................................... 166 5.29. ANOV .A results from responses on "confident each of the following roundtable members 'vill follow through on a commitment" ............................................... 166 5.30. ;\Yerage Response on "extent to which you consider each person listed below to be influential in water policy and management issues in the STATE OF COLORADO; that is, the people who seem to han pull, weight, or clout with others on these issues" (Scale: 0 = Not at all influential to 4 = Very influential) ..................................................................................................................................... 168 5.31. ANO\'A results from responses on "extent to which you consider each person listed below to be influential in water policy and management issues in the STATE OF COLORADO; that is, the people who seem to have pull, weight, or clout with others on these issues" .................................................................... 168 5.32. i\lean Responses to the Success of the Colorado Basin Roundtable by Belief Clusters of Colorado Roundtable Basin i\lembers (5 = very successful and 1 = not at all successful) ................................................................................................. 172 5.33. Analysis of \' ariance for Responses to the Success of the Colorado Basin Roundtable Process as a Function of Belief Clusters ......................................... 173 5.34. 1\lean Responses to the Success of the HB 1177 Process by Belief Clusters of External Stakeholders (5 = very successful and 1 = not at all successful) ...... 175 5.35. Analysis of\' ariance for Responses to the Success of the HB 1177 Process as a Function of Belief Clusters ..................................................................................... 176 X ill


6.1. Summary of the FiYe Clusters Affiliations (Proposition 1 ), Lenl of Cooperation within the Cluster (Proposition 2), and Engagement in the Participatory Process (Proposition 3)* ......................................................................................... 194 B.1. Significance }eye] for normal distributions by cluster on each network nriable using a Shapiro-Wilk test. Significance of < .05 indicates assumption of normalitY is Yiolated ................................................................................................. 22R XlY


PREFACE In 2005, in partnership with the Colorado Institute of Public Policy, 1 began to work with the lnterbasin Compact Committee (IBCC), newly formed as a result of legislation. The IBCC was mandated to OYersee the statewide implementation of nine \Vater basin roundtables, whose charge was "to facilitate continued discussions within and bet\veen basins on water management issues, and to encourage locally driYen collaborative solutions to water supply challenges" (37 -75-104, C.R.S.). I had the opportunity to support one of the ftrst legislatiYely mandated tasks of the IBCC, de,eloping a negotiation charter, by completing a white paper on strategies for negotiating and consensus building in the context of the work they were charged with undertaking. From these ftrst steps, I could see the challenge that lay ahead for this new, statewide, and quite oYerwhelmingly large participatory policymaking process. As the nine roundtables began their work, many different interests sat together monthlY, if not for the ftrst time, at least for the ftrst time within a statutorily authorized context. The roundtables began to study and explore water policy and management issues, \Vith mer 400 representatiYes of municipal, agricultural, industrial, recreational, and emironmental interests meeting monthly across the state to try to ftnd solutions to water policy problems. XY


As time passed, the roundtables achieYed small wtns and continued to be supported at the statewide level through the passage of two more bills intended to fund and guide the process (HB06-1400 and SB06-179). HoweYer, the actiYities of the roundtables were largely focused on process and outreach. The lack of full engagement by all of the interests seemed to be an issue just under the surface of many meetings I attended. How could any roundtable achien policy recommendations that could be implemented if they did not have strong representation from all of the inYested interests, and trust between the interests? Hoping to explore this issue further, I submitted a proposal in partnership \Vith the Colorado Institute of Public Policy to the Public Education and Outreach committee of the IBCC. The purpose of the research, as described in the proposal, was: With the establishment of nine water basin roundtables as a result of HBOS-1177, a new mechanism exists to address long-term conflicts in water policy and plan for the future. Gi,en the dinrsity of participants some of whom have long been actively imohed in water issues and others who ha,c been more peripheral players the roundtables sene as an important Yenue for addressing both intra and interbasin issues. Collaborati,e decision-making structures .like the roundtables han tremendous potential, but they also suffer from the complexity of interpersonal and interest-based stakeholder interactions. One way to address these complexities is to understand them through an analysis of the networks of actors directly and indirectly involnd in the roundtable process. The Public Education and Outreach Committee recommended funding the study and the Colorado Basin Roundtable, under the leadership of Dave .Merritt, agreed to be the site of the study. The survey tool was designed in partnership with water stakeholders representing many interests and parts of the state. The roundtable members ga,c an xn


hour of their time to taking the suney and aided in securing contact information to oYer 100 additional stakeholders who participated in a shorter version of the same suney. The generous information prmided in the sun'ey allO\ved for an in depth exploration of interactions at the water roundtable. L1tilizing additional data not discussed in tlus dissertation, a report was drafted that explored participation at the Colorado Basin Roundtable and statewide. The results were carefully reYiewed by key stakeholders, who helped interpret the findings, and after two reYiew processes, the report was finalized. The report, issued as a publication of the Colorado Institute of Public Policy and aYailable at, utilizes some of the same data and some of the same analysis as this dissertation. However, the report and this thesis are fundamentally different documents. The purpose is different; rather than a careful exploration of theory and analysis together, the report focuses on key findings that can help the Colorado Basin Roundtable and the roundtable process explore participation issues. The report makes no attempt to relate the inf01mation gathered from the Colorado Basin Roundtable to public participation or public policy theory, which is the main thrust of this dissertation. The report presents similar, but less in depth analysis, of the belief coalitions, demographics, and data on success. A different analysis than the one seen here was conducted of the network data and presented in the report. The report also includes analysis of network data from another 400 stakeholders that is not part of tlus dissertation. Finally, the report is designed for a lay audience and buries most of the research detail in the appendices. XYll


The report and this dissertation share an important component in common. The interpretations gathered from stakeholders to understand the belief coalitions \Vere used for both the report and dissertation, as the participatory nature of the research necessitates stakeholder engagement in both design and interpretation. In 2008, the report was presented to the Colorado Basin Roundtable. The meeting participants reported that the findings resonated and a dialogue began on how to address the issues that the report raised .. \fter four facilitated sessions, the final result of the report was a series of decision-making tools for the roundtable to utilize in their attempts to form consensus and take action on issues of importance. These tools represented an opportunity for the roundtable to mme from discussion to action, an actiYitY that had been an ongomg challenge. The, also helped focus conYersations toward outcomes, with a goal of increasing the need for the less engaged stakeholders to increase their participation, as their roundtable mo,ed from safe acti,ities like outreach and studies to less safe actions that could influence policy in their basin and statewide. In 2010, 1 learned that the decision-making tools that \Vere de,eloped m response to the report haYe accomplished what they were intended to do. The Colorado Basin Roundtable is making decisions that not only han ramifications for their O\Vn basin, but are influencing policy statewide. i\Iembers of the roundtable shared examples of how they are using the decision-making tools to strongly advocate as a roundtable at the state leYel, and that they are advocating for policy changes that combine the interests at their table. They are developing consensus on topics as controYersial as interbasin xnu


transfers and state legislation intended to change water policy. The goal of any participatory research process is to benefit the stakeholders. It is with pleasure that I conclude this preface by reporting that the roundtable that participated in this study, resulting in both a report and a dissertation, has benefited from their effort in concrete ways that are making a difference in water policy and management in Colorado. XlX


CH"\PTER 1 INTRODL'CTION Participatory policymaking and collaboratiYe decision-making structures ha,e been a defining feature of the practice and study of policymaking for more than two decades (Krimskey, 1979; Fischer, 1981; Dahl, 1994; Barber, 2004). In practice, participatory processes arc becoming the norm in many ,enues and for many issues, sometimes because goyemmcnt imites the public to participate and other times because the public forces itself into the process when the goYernment fails to engage them (e.g. Kathlene & l\lartin, 1991; Hajer & Wagenaar, 2003; Welsh, 2004). In emironmental policymaking, participation processes haYe ranged from nationwide, engaging thousands (Depoe, Delicath, & Elsenbeer, 2004) to statewide, engaging the broader public and leaders of enYironmental advocacy moyements (Ambruster, 2008) to local lc,cl, engaging key stakeholders with a nsted interest in their local \Vatersheds (Griffin, 1999). With the increasing use of participatory processes comes an increasing focus in academic literature on creating high quality participation processes, and in recent years, theorists haYe developed more sophisticated ways of assessing the purpose, quality, and outcomes of participation (e.g. Kathlene & l\Iartin, 1991; Webler, 1995; Beierle, 1999; Rowe & Frewer, 2000; Webler & Tuler, 2000; Petts & Leach, 2000; Leach, 2006). As a value, participatory democracy theorists argue that it is necessary and even demonstrably better when government engages citizens in its processes (Dryzek, 1990;


Renn, Wehler, & Weidemann, 1995; Schneider & Ingram, 1997; deLeon, 1997; Petts, 2001; Narayan, 2002; Barber, 2004). In the current political context, the complex cmironment that has resulted in policy arenas with many Je,els of government and types of stakeholders inYoked has made it impossible not to make policy hand in hand with non-gonrnmental stakeholders (Dryzek, 1996; Hajer & Wagenaar, 2003; Bryson & Crosby, 1992; Susskind & Cruickshank, 1987). Proponents of deliberative processes extend this argument to suggest that not only does the public need to be imohed, but that decisions should come from the deliberations of "dinrse groups of citizens -not just experts and professional politicians" (Gastil & Keith, 2005, p. 3). Howe,er, participatory processes that allow for ongoing policymaking can be dominated by agencies and representatives of business, industry, recreation, the em-ironment, and other interest groups, and may or may not include participation from the general public or indiYiduals paid to advocate for the public's interests (Griffin, 1999). Perhaps in recognition of this pattern, it is not uncommon for a deliberative, participatory process to purposefully engage interest groups, rather than the general public. Bishop and Da,-is (2002) refer to these as COIJI/Jllmitio, where there is a "long-term dialogue between government agencies and a series of interest groups imoh-ed in particular policy arenas" and often a formal mechanism for deliberation, such as an advisory board (p. 23). Though the general public is certainly still involved in participatory processes throughout the country, it is also true that participation for many policy arenas has been defined as also being about expertise and experience, not just


desire to participate. For this reason, some of the researchers exploring representation outside of electoral politics arc discussing participatory processes as a place for repreJenlalil'eJ o{publk inll're.riJ, not just members of the public, to shape policy (e.g. sec Curry, 2001; Weber, 2000; Kenny, 2000; Petts & Leach, 2000; Rowe & Frewer, 2000; It...-in & Stansburg, 2004; Rockloff & i\loore, 2006). If scholars and practitioners belieYe representation of interests is a necessary component for participatory policymaking processes, then it is important to understand what it means to represent an interest. .\lthough most participatory process theorists whether coming from the public participation, the participatory democracy, or the public interest literature hme emphasized the 'ralue of di,erse representation, few ha,e gone beyond the concepts of demographics and affiliations as indicators of representation (Petts & Leach, 2000). Other bodies of literature, including the AdYocacy Coalition Framework, a theory from the public policy literature, haYe emphasized that beliefs and shared values, rather than demographics and affiliations, connect indiYiduals and lead to collectiYe action (Saba tier & Jeluns-Smith, 1993). Yet, participatory process theorists and practitioners are slmv to assimilate these concepts. In defining representation through such a limited lens, participation theorists may be limiting their ability to help achie,e the desired outcomes of participatory processes, including shared understandings, high quality agreements that represent a compromise between interests, inno\'ation, learning, and change (Connick & Innes, 2001; Innes & Booher, 1999). 3


l11e .-\dYocacy Coalition Frame\vork ("\CF), a policy process theory that explores how people come together to influence policy, pays little to no attention to demographics and affiliations. Rather, it describes a policymaking world where core beliefs bring people together into coalitions and lead them to joint action in pursuit of shared policy outcomes (Sabatier, 1999; Sabatier & Jenkins-Smith, 1993; Sabatier, 2007). The theory argues that policymaking is the result of adYocacy coalitions, groups of actors who are di,erse in their roles, but who haYe a shared belief system including "a set of basic nlues, causal assumptions, and problem perceptions" and whose ongoing actiYities show a leYel of coordination (Sabatier, 1999, p. 25). The .-\CF concept that core beliefs are \Vhat driYe collectiYe action, a concept supported in the literature much more widely than just through ACF (e.g. see Smith-Lmin, & Cook, 2001), falls short of being complete in its failure to consider alternati,e scenarios, ones where shared action may occur without shared beliefs, or vice ,ersa. L'nderstanding that coalitions are more than just a function of their shared beliefs, but also a function of their resources and of the individual members of the coalition (Kim & Roh, 2008), expands the concept of a belief coalition in a way that is Yital if this line of policy theory is to help inform public participation theory. Just as belief coalitions are composed of individuals, so too are participatory processes, whose individuals represent interests, and thus may be members of belief coalitions, whether engaged in collectin action or not. These theoretical perspectins when applied to a participatory policymaking setting suggest that understanding the core beliefs that 4


indiYiduals bring to a participatory, deliberatiYe process may help explain how or why they engage with other stakeholders in the process and those peripheral to the process. For deliberati,e processes that assume a representatiYc will connect back to the interest they represent, the ACF perspecti,e on core beliefs is eYen more important, as it suggests that indiYiduals \vill seck to build collective action with those who share their beliefs, rather than those who merely share their demographics or affiliations. The combination of these theories may well create the opportunity for public participation theory to embrace the role of interests, as an interest defined by beliefs is less inherently elite and non-participatory than an interest defined in the more classic political theory of interest groups. Research Purpose and Research Questions When the literature on public participation is brought together with the .\CF literature on the role of beliefs in policymaking processes, there is an opportunity for increased understanding of the dynamics of participatory processes. The exploration of public participation in the chapters to follow reYeals that the definition of representation has been dominated by a focus on demographics, leading to a failure to account for interests. The addition of public policy literature, particularly ACF and its focus on coalitions, adds to the policy process literature the concept of interests and expanded the understanding of beliefs. Howe,er, ACF theorists han largely ignored the role of the indiYidual, something that public participation processes often emphasize. To bridge the two theories at the individual lenl, Kim and Roh's (ZOOS) model of individual 5


interaction within coalitions introduces the concept of deliberatiYe dialogues between members of interest-based coalitions. With the three theories, this dissertation will argue that a model of a public participation process can include (1) individuals appointed to a participation process whose (2) beliefs and actions align to create (3) coalitions with others both at the table and outside the table, and whose (4) interactions at the table may differ in part based on these beliefs and coalitions. The chapters to follmv argue that the public participation literature has excelled at understanding processes for participation and how indi,iduals should have opportunities to interact and make decisions within these processes, but has stopped short of understanding the role of interests and representation in a participatory process, due to its focus on "public," most broadly defined. Where public participation fails, policy literature can add in its understanding of interests and coalitions; yet, policy theorists are focused on the larger political context and thus largely ignore the importance of a participatory process, including such things as deliberation, trust, and interactions. \X'ith this focus on the connections between the individual, the participatory process, and the interests tn mind, the following research questions are posed: 1. What is the relationship between demographics and beliefs among participants in a participatory process? This exploratory question makes the first connection between the two bodies of literature: the indiYiduals imoh-ed in the public participation process han policy beliefs that they bring to the process. This is an often neglected first step in the study of 6


participation processes, as scholars do not typically seek to understand interests through beliefs. Perhaps it is ignored in part because the challenge of understanding an "interest" lies in the need for that understanding to be placed in a broader context than just the participation process itself. What point would there be in merely learning of the beliefs of participants, for without knowledge of the beliefs outside the table, one cannot claim participants npreJml anything beyond their own beliefs. 2. What is the relationship between beliefs and interactions within and between coalitions in a participatory policymaking process? This exploratory question focuses on the policy literature, looking for how the indi,iduals with beliefs function within their coalitions, but in its focus on the indi,idual, it brings in the emphasis of the public participation literature on the role of the indiYidual in the decision-making process. The Advocacy Coalition Framework suggests that beliefs lead to coalitions (Sabatier & Jenkins-Smith, 1993). The research question focuses on participatory policymaking processes and key issues with them. Policy research in the age of complex policy problems cannot seek the solutions to problems without considering the social context of individuals and groups (Hajer &Wagenaar, 2003). Authors such as Fischer (1995, 2003), YanO\v (2003), Gottweis (2003), Durning (1999), and Lynn ( 1999) all argue for policy analysis to respect the interpersonal dynamics of the interactin policy environment, whether through the study of discourse, ,-alues, beliefs, or symbolism. In posing the second research question, we arc taking precisely this approach, examining the interpersonal dynamics of the representatives at the water roundtables and explicitly using their beliefs to understand those interactions. 7


The argument made by this research question is that interpersonal interaction within a public policy process focused on water policy are a function of beliefs held about that policy process and what issues it should prioritize. By engaging in a case study design, with a wide range of data about a specific participatory process, this thesis seeks to establish as much of a casual relationship as is possible in a complex social and political cmironment such as water policy. 3. What is the relationship bet\veen coalitions formed by beliefs and how the coalitions are engaging in a participatory policymaking process? Here again, this exploratory question connects the t\vo theories: public participation theon focuses on how individuals participate in a process, while policy theory is more interested in the beliefs that bring people into a policy process. The question asks not how indiYiduals are participating in a process, but how coalitions based on shared beliefs arc participating in a process. It is the final question that builds the connection to the goals of a participatory policymaking process. If the process intends to achieYe better decisions through such things as representation and deliberation (the criteria established by thejaimeJJ approach (Wehler, 1995), then the coalitions must be acti,e participants in the process. Gi,en their exploratory nature, these research questions lend themselves to a case study approach, rather than attempting to look across many participatory processes for patterns. A case study creates the opportunity for a depth of understanding not anilable in a research approach that focuses on large random samples. Additionally, in a study that uses network data, large random samples are not appropriate (\X'asserman & 8


Faust, 1994). In studying an issue in depth, case study research often uncmers the examples and eYen exemplars that can inform later comparatiYe studies that utilize larger samples (Flyvbjerg, 2001 ). In fact, to understand these research questions, randomness is the least appropriate approach, as the interest is in understanding in detail the small group of actors imoh-ed in a participatory process and their interactions with stakeholders who are members of their advocacy coalitions. For this reason, the research questions will lead to research propositions, posed in chapter three, which will drive the analysis and findings. Research Setting To understand the concept of representation as an tssuc of core beliefs, not merely demographics and affiliations, this study explores the role of beliefs in the participation and interaction of stakeholders in a participatory policymaking process. Colorado's water policy community provides an ideal setting to explore these questions. Water is both a policy issue rife with conflicts and interest groups, and vitally important to society. Water is fundamental to the growth of communities. CiYilizations develop around water sources, with social and economic forces tied to the equitable, efficient distribution of water to meet needs and demands (Wittfogel, 1957; Caponera, 1 992). The management of water throughout history and throughout the world has been complex, rarely without conflict, but ,ery necessary as "adequate management of \Vater resources is a prerequisite for ci,ilized progress and human survival" (Caponera, 1992, p. 25). 9


In Colorado, an arid, western state, water has become a source of conflict and contention among the many users, from municipal goYcrnments to agricultural producers to emironmentalist and recreational users. The necessity of water for the surviYal of different interests has dri,en the denlopment of a complex legal system entrenched in years of statutes, tradition, and legal precedents. The consequence is a water management system that is sometimes referred to as "archaic" and "inflexible" by water stakeholders (Colorado Institute of Public Policy, 2006). The extensive legal system that underpins Colorado's water policy was designed for different times and different problems than currently face Colorado. As demands ha,e grown and issues have evohed, ne\v stakeholders and new challenges han emerged. To make significant changes would require broad support, but resistance is unavoidable, as some interests benefit from the design of the current water policy system and others fear they would be harmed by any significant changes. In response to these problems and specifically to address predicted \Vater shortages, policymakers in Colorado have initiated a complex participatory process: the water roundtables. The water roundtables take the participatory policymaking trend that has been occurring in watershed level management (Chess & Purcell, 1999; Leach & Pelkey, 2002; Koontz & Johnson 2004; Lubell 2004; Weible, Sabatier, & Lubell, 2004; Lachapelle & McCool, 2005), and expand it to a statewide approach that engages eight water basins in a collaborative, state\vide process composed of nine participatory roundtables. The ninth \Vater roundtable represents the Demer-metro area, rather than 10


a water basin. The ne\vly undertaken process is heaYily dependent on water stakeholders being willing to engage and compromise. It is also dependent on the ability of the actors who han been brought together through the water roundtables in each water basin to reach consensus not only among themsehes, but between the nine different roundtables. These dependencies arc a consequence of the roundtables haYing a statutory mandate to make decisions, but not to implement them in any manner that conflicts \Vith an existing law gmcrning water usage and rights (37-75-102, C.R.S.). As will be explored more in the following chapter, the lack of statutory authority means that the implementation of any decisions made by the roundtables is dependent upon the indiYidual water rights holders. It ts their choice whether to implement a decision made by the water roundtable, as the market-based SYStem continues to be the mechanism for implementation, not the water roundtables. The lack of authority creates an unusually high need for the participatory process to be seen as a credible decision-making process by water policy stakeholders who are not at the table. The need is paramount for the process to result in decisions that are widely supported by many stakeholders. Understanding that need, the roundtables were designed tn statute to reqUlre representation from both the "usual" and "unusual" suspects. As of May 2007, the nine water roundtables had largely unchanged membership since their ftrst few months, resulting in nine roundtables, each with benvecn 25 and 50 people, \vorking together monthly on this complex policy issue. The roundtables were only 18 months into their work together, too early to assess the outcomes of the process, 11


but far along to assess the interactions within the process. Each roundtable is required to han specific representatives of different affiliations, such as agricultural, enYironmental, county, etc. Each roundtable is connected to the others through merlapping members, and also connected to a broader water policy community as the representati,es are expected to be communicating with the interests they represent. The statute that defines the membership limits some interests, including environmental and recreational, to only one mandatory appointment per roundtable. This single individual is expected to communicate back to the full range of environmental and recreational interests in their basin. The engagement of this broader network of stakeholders is an opportunity to examine beliefs much more broadly than just within the roundtable process itself. HoweYer, the roundtables represent an opportunity to look at how these core beliefs relate to the perceptions and interactions in the participatory process. The roundtable process faces many challenges due to the complexity of the policy issue and dinrsity of stakeholders. With the recent shift in Colorado's political climate, the result of a change in party in the Governor's Office, it is important for policymakers to understand how the roundtable process is progressing. Significant resources are inYested in the process, both state dollars and volunteer and staff time across the state. The process will not be continued if the stakeholders invoked do not sense it has the potential to succeed. Among the issues of greatest concern is whether or not the roundtables can identify ,iable solutions ginn the political and legal climate they operate within. In essence, the question is whether the process has the potential to 12


engage stakeholders across interests at such a leYcl that implementation of decisions can occur. The Colorado Basin Roundtable, one of nine roundtables around the state, was interested in participating in this dissertation study as its chair and members belieYed they would be better able to address the challenges at their table if they could understand the underlying dynamics occurring in the room. On the heels of multiple studies and reports documenting the physical em'ironment surrounding \Vater (CBl\1, 2004), the role of Yalues and beliefs m water (Colorado Institute of Public Policy, 2006), and the historical context of water (Hobbs, 2004), it became increasingly eYident to water roundtable members that the political context surrounding the roundtables was more complex than could be easily understood. Though water conflicts \vere often discussed at roundtable meetings and in the reports in dichotomous terms, "consumpti,,e vs. non-consumptiye" and "east YS. west," the conflicts themselYes were far more in'Toh'ed, with many different interests approaching the same issue from a wide range of directions. With m'er 290 stakeholders participating in the roundtable process a process unable to legally challenge the complex political and legal environment of water tools \verc needed to better understand the opportunities that existed for the roundtables to influence water policy and management. In response to this need, this water roundtable study was designed and implemented, with the practical goal of increasing the water community's understanding of: (1) priorities held by water stakeholders and their specific desired outcomes from the roundtable process; (2) progress that water stakeholders belie,e the roundtables have 13


made; (3) commumcauon occurring between different interests at and outside of the roundtables; and (4) nenvorks of influence in the roundtable process. This practical goal fits well with the more academic goal of understanding how beliefs, as defined in the public policy literature, relate to participatory processes, as understood through the public participation literature. The Colorado Basin Roundtable's high level of interest and the broader policy emironment's need for additional information to inform their process created an opportunity for more extensive data collection than is usually possible through survey methods. Suncy participants throughout the state devoted significant time to answering complex and personal questions. The dissertation research included the collection of (1) demographic and core belief quantitative data, (2) qualitative supporting data, and (3) nenvork data to answer the questions. The study triangulated methods to create a complete picture and mme the understanding of participation and representation forward in both the academic literature and the applied settings. Additionally, due to the policy arena's interest in the research, the thesis was not onlY an academic exploration, but also an opportunity to directly meet the needs of a current and complex participatory policymaking process. To ensure the study was of use to the roundtable, the research methods employed strategies from participatory action research models, including steady involvement of key stakeholders across a wide range of interests from design through interpretation of results. 14


Organization of the Dissertation The study is organized into fiye chapters, each of which contributes to the combined understanding of the research setting, the Colorado water policy community and the Colorado Water Basin Roundtable, as well as the research purpose, to understand the role of beliefs and representation in a participatory process. Chapter nvo proYides background on the water policy issues most rele,ant to the work of the Colorado Basin Roundtable. Although the majority of the analysis and research questions are not related to any specific water policy problems, to understand the core beliefs component of the analysis, it is important to understand the complexity of water issues in Colorado. Certain terms are commonly used in water policy and will be used throughout the analysis. Chapter nvo introduces and defines these terms within their historical and policy context. Chapter three explores the academic that has gone dmvn nvo different paths: paruCipatory democracy, participation and deliberation strategtes, and representation of interests; and the importance of the role of beliefs in policymaking processes. The literature re,iew focuses on the strengrhs of both sides, highlighting the opportunity to connect the study of representation in participatory processes to the understanding of the role of beliefs in policy interactions. Chapter three concludes with the deYelopment of research propositions on the role of beliefs in a participatory policymaking process. 15


Chapter four outlines the research methods, including cluster analysis to develop the coalitions of beliefs, nenvork data to understand interactions with the clusters, and additional quantitati,-e analysis to understand engagement in the process. The chapter also explores the participatory action research approach that guided the design and implementation of the study, and the nlue it added to the research. Chapter five presents the findings of the study, beginning with the development of the belief coalitions, the core of the analysis. For this study, the belief coalitions sene as the independent ,-ariable for all of the remaining analysis, allowing beliefs to drive the research. The analnis uses net\vork data to describe the within-cluster and benveen cluster interactions of the participants in the Colorado Basin Roundtable. The final component to the analysis explores how these individuals perceive the success of their roundtable and the broader policy process. Chapter six concludes with a discussion on the research findings in the context of the research propositions, and with a discussion of both the theoretical and methodological learning that this study can contribute to the broader literature. It presents the limitations of the study and next steps to address the issues this study was unable to cover. Chapter six also provides the reader with the story of how the Colorado Basin Roundtable used the research to inform their continued work on water policy issues. 16


CHAPTER 2 WATER POLICY IN COLORADO Colorado water policy is facing a crisis. Limited natural resources and competing demands are making it imperatin that public policy addresses the looming social, economic, and emironmental consequences .. \s stated by the Water 2025 report issued by the Bureau of Land i'vlanagement, public policy must face "the reality that the demands for \Vater in manr basins of the \X1est exceed the available supply even in normal years" (Bureau of Reclamation, 2005, p. 1 ). Where once drought was the cause of \Vater crises, now it is only one of many 1ssues. Colorado's naturalh arid environment and limited water aYailabilitr arc all the more problematic as the state faces an increasing population, over-allocated watersheds, aging facilities, inadequate '':ater storage, and conflict over the \alues associated with water allocation (Bureau of Reclamation, 2005). The conflicting \'alues are partially the result of an increasing focus on the non-consumptive uses of water, specifically: the enYironment and recreational needs (CDl\1, 2004). As these new stakeholders take on more central positions in water policy, the underlying conflicts expand. \X'ater laws once focused on the use of water in homes, farms, and commercial acti,ities can no longer address today's needs (Pontius & SWCA, 1997). 17


An exploration of any \Vater policy issues in Colorado requires understanding the context: the natural and man-made water context in Colorado; the laws and management of water; the stakeholders invested in water law and management; the changing needs; and the current policy attempts to address the water crisis. The Lay of the Water Colorado's riYer system is composed of eight major water basins, connected by four major river systems, all of which begin in the center of the state, with their headwaters emerging from the mountains along the Continental Divide (CDl\'1, 2004). The state is unique from others in the nation, as all of its major rivers originate within Colorado; water is not captured from other states, but instead is required by a complex array of interstate agreements to be allowed to flow into other states. As can be seen in Figure 2.1, six of the eight basins are largely on the Western half of the state. Though the South Platte and Arkansas Basins represent over half of the total landmass in the state, and include 85/o of the state's population, the two basins combined represent only a tiny portion of the water flow in the state only 5% of the total water flmv leaving the state comes from these basins. In contrast, the Colorado basin alone accounts for almost 701o of the water leaving the state, yet has only a fraction of the population. Due to these contrasts, \Vater transfers between basins are a significant issue in Colorado, with the dry Front Range pulling water from the Colorado basin through a complex network of reservoirs, rivers, and underground pipes (CDl\1, 2004). 18


Each part of the state uses a different balance of groundwater and surface water. Ground\vater is held in aquifers underneath the land, \vhile surface water flO\vs through riYers. Although groundwater and surface \\"ater exist throughout the state, some areas like the \X' estern Slope rely primarily on surface water, while other parts of the state, such as the San Luis \'alley, dra\v on both types ohvater. South Platte Figure 2.1. HBOS-1177 Water Basins The large cities in the Front Range rely heaYily on surface water, much of it drawn from the water resources of the \'{1estern Slope, but smaller municipalities also access groundwater. Surface waters are more ,-ariable than ground waters, and depend hea,Tily on precipitation and snowpack melting 2004). Some surface and 19


groundwaters are connected to one other, with irrigation from wells in certain types of hydrologic zones haYing affects on the anilability of \ in rivers (Meyers, 2008). Two types of uses arc recognized as important in water policy and law: consumptiYe and non-consumptive. Con.wmpti!Je water use is defined as: "Water use that permanently withdraws water from its source; water that is no longer aYailable because it has enporated, been transpired by plants; incorporated into products or crops, consumed by people or liYestock, or otherwise removed from the immediate water emironment" (Hobbs, 2004: 31). In contrast, 11011-t'Oil.mtJJptil'l' water use is the use of water that does not remove it from the environment, such as water flowing through a stream, and thus supporting a native habitat and/ or recreational activities. Con.mmpti1e ll;'ater l!Je.r in Co/omdo Historically, water m Colorado had three purposes: agriculture, numng, and supporting human habitation. Communities grew up around water sources, along creek and river basins (Hoyt, 2005). Onr time though, the technology has allowed for water to flow to people instead of people migrating to the water. Consequently, consumptin use of water is not limited to areas where water is abundant, but in fact is quite the opposite, with consumptive usc throughout the State and hea'}' in the drier areas of the state. For most people in Colorado, water is first and foremost a municipal need. As of 2005, to keep water flowing from the taps in residential and commercial buildings throughout the state, 6.7 o of the water in Colorado that is diverted from nvers, 20


resetTous, and the ground is directed to municipal use. Eighty-senn percent of municipal and residential use of '.Vater comes from water managed by municipal utilities, \vhile printe wells account for the rest. .\s much as 70 o of municipal water is not directed to human consumption, howeYer, but to the \Vatering of lawns, parks, and trees (Grigg, 2005). The largest man-made consumptiYe use of water is agriculture, with combined production of food and other agricultural products accounting for oYer 86 o of the diversion of Colorado's waters from riYers, reservoirs, and the ground as of 2005. Colorado's agricultural industry generates fi,,e billion dollars a year in reYenuc, with two thirds from linstock and one-third from crops. Agricultural users divert water from the ground or ri,ers, but also return water after use Yia "irrigation return flows." Depending on the type of agricultural uses, return flows in each basin nry from 43'o of the total agricultural withdrawals to onr 80% (Grigg, 2005). Although municipal use of water is not as demanding as agriculture, the water use battles are often between municipalities and agriculture, as the growth in population is driving the transfer of water rights out of rural communities and into their urban counterparts. Although industrial use of water accounts for less than two percent of Colorado's available water, its demands are growing, particularly for energy production. \Vithin municipal water districts, commercial and industrial water usc accounts for a significant percentage of water use (Grigg, 2005). Outside municipal areas, significant water is needed to extract oil from oil shale, and oil companies ha\'e purchased both flow 21


and storage rights allowing them to build resenoirs specifically to meet their industrial needs (Hanel, 2008). Though many consumptiYe uses happen within a water basin, some consumptiYe uses of water require diYerting the water from its basin of origin to the basin where it is needed. When water is di,erted out of a basin to meet any of these consumptiYe needs, no return flow is left in the basin. Thus, the sale of agricultural water to municipalities or industrial uses outside the basin can ha,e consequences for downstream water rights. Currently, much of the Eastern Plain's municipal water is acquired from other basins, particularly the Colorado Basin, and future water diYersions are expected from the Colorado Basin to the Front Range. j\'oii-CO!WttllptitJe u,:-'ater {_ in Colorado Colorado's natural em-ironment is the largest consumptiYe water user m the state, using 78 o of all water in Colorado. agricultural, municipal, and industrial/ commercial use, emironmental use of water does not diYert water from its natural pathways, whether aboYe or below ground. Rather, enYironmental use of water feeds the habitats surrounding watenvays. Colorado has a di''erse ecosystem, and its 22 million acres of forest land and 30 million acres of plains (high and lowland) require significant water to sutTin. Within those forests and plains are many different species of fish and other wildlife that depend on the maintenance of the natural habitats (Grigg, 2005). 22


Management of water over the last century has resulted in millions of acre of feet being diYerted from riYcrs, changing the ecosystems around the riYer. As water is moYed around through man-made infrastructure, it changes the natural flows, temperatures of the water, \Vater quality, and unintentionally, can negatinly affect many species, from fish to birds. In fact, the four fish historically most present on the Colorado Rinr (Colorado pike minnow, razorback sucker, bonytail, and humpback chub) are listed as endangered species (Pontius & SWCA, 1997). The negative impact of artificial mmement and storage of water for human, agricultural, and industrial/ commercial consumption, combined with pollution resulting from the years of agriculture, industrial uses, and mining, haYe led to water quality issues throughout the state. .\ll eight of the Colorado water basins ha,e documented emironmental concerns including many incidents of pollution and degradation of natural habitats. Federal and state monitoring of the emironmental problems han led to many special projects attempting to mitigate water contamination from mining and agricultural runoff, create and maintain high quality water habitats to protect endangered species, and control propagation of non-natin species (CDl\1, 2004). Non-natiYe species can negati,ely affect water in Colorado through a Yariety of means. For example, the Russian Olive is an imasi,e species in Colorado that uses excessiYc water and the Eurasian \Vatennilfoil, another invasive species, grows on the surface of the water, preventing light from penetrating below (Owens-Yiani, 2001). In 2008, Zebra mussels were found for the first time in Colorado (Associated Press, 2008). In the Great Lakes


region, this species has propagated aggressiYely, causing both economic and ecological damage (Owens-\'iani, 2001). Addressing the penasin and persistent problems of inYasiYe, non-natiYe species in the watenvays is one of the emironmental challenges of the state, a challenge that has a direct affect on all uses of water. Non-consumptiYe users of water in Colorado include more than just the emironment. The recreation industry, including whitewater rafters and fishermen, have a suong interest in keeping water flowing in the rivers. While the two mterests, emironmental and recreational, are sometimes grouped together (e.g., such as in the legislation mandating the membership of the basin roundtables, explored later in this chapter), they do not always share the same priorities. For example, reintroduction of natiYe fish into the streams, a priority for many em-ironmental interests, may not meet the recreational needs of fishermen whose sport has benefited from the abundance of non-nauve spectes. However, the two interests have much to gain from partnering together, as the federal requirements on emironmental issues can be leveraged for instream flow programs that benefit recreational users, and the economic benefit of recreation in the rivers can be leveraged for implementation of costly or water-intensive emironmental restoration efforts. Case in point: in a study that explores the economic benefits of nonconsumpti,e uses, the authors emphasize recreation as the revenue generator, but define the issue as one of both emironmental and recreational needs (e.g. Sanders, Walsh, & Loomis, 1990). 24


The Colorado Bmin The broader description of Colorado's \Vater context also applies to the Colorado \Vater Basin. Howe,er, the nearly 10,000 square mile basin has its own unique issues as well. The Colorado Water Basin is home to Grand Junction, with nearly fifty thousand people, and Glenwood Springs, with under ten thousand people. Elention in the basin ranges from fourteen thousand feet down to four thousand. federal land comprises the majority of the upper basin of the Colorado, with forest and rangeland the predominant land uses. The federal lands allow for Yarying le,els of liYestock grazing, recreation, and timber harYesting, resulting in a wide range of stakeholders with ,-ested interests in these lands (IBCC, 2008). The Colorado River (and thus the basin) flows far beyond Colorado borders. It ts fed by senn states and eventually flows into l\Iexico. The water in the riYcr ts allocated Yia the Colorado River Compact of 1922, a negotiation between the l'.S. government and the seYen states. The Colorado Basin has an unusually large capacity of \Vater storage facilities associated \Vith it, allowing ability to diminish the negati,e impacts of droughts (Pontius & SWCA, 1997). In Colorado alone, water storage on the Colorado Basin includes the Blue Mesa, :1\IcPhee, l\Iorrow Point, Taylor Park, Ridge\vay, and Crystal reservoirs (Pontius & SWCA, 1997). Water in the Colorado RiYcr goes primarily to agricultural uses, with approximately 80 u across all seyen states it flows through currently diYertcd to agricultural needs. In Colorado, industrial, municipal, and recreational uses are also high 25


priorities. According to the lnterbasin Compact Committee's website, "TI1is complex dynamic of water uses makes the Colorado .tvlainstem Basin Roundtable one of the most diYerse and complex roundtables in the state. It also means that it faces some of the most difficult challenges" (IBCC, 2008). In October 2007, the Colorado Basin Roundtable reported to the IBCC an assessment of their major water issues. The Roundtable identified issues and problems for each county within the basin, and the common theme was the negatiYe impact of transbasin diYersions and interstate compacts on environmental, recreational, municipal, industrial, and agricultural needs 'vithin the Basin (Colorado Basin Roundtable, 2007). lVaterQuanli(J' and l:ltt11re Prf!jedion.f Colorado is the head\vater state for seYeral major western rtYers. For the Colorado River, 86 o of the water rn the river originates in Colorado, with seasonal patterns to the natural flow of water. o,er the years, dams and reservoirs have been created to manage water flow, controlling it and decreasing the differences between peak and low flow times. HoweYer, storage of water in above ground resenoirs results in enporative loss. One of the challenges of Colorado in general is the increased evaporation rates common to arid states (Pontius & SWCA, 1997). Municipal, industrial, agricultural, recreational, and environmental interests are all competing for the same limited of water (Corbridge, 1998), a supply that must also go downstream to the other states. Current estimates have found that when accounting for new projects and planning processes already undenvay, the State of 26


Colorado will only han approximately 80'o of the \Vater supply that it is estimated to need as of 2030 (CDt\1, 2004). If the projects included in the estimates are not completed, an eYen greater shortage will ha,e to be addressed. The statewide numbers can be misleading, however, as available water and water demand arc often not located in the same place, resulting in additional local shortages (CDt\1, 2004). ;\lthough new water supplies can sometimes be de,eloped by mming water between locations or increasing water storage, the expense of developing the water supplies is high enough to be prohibitive in many cases (Corbridge, 1998). Climate change predictions suggest additional challenges in the future. By 2050, some scientists are predicting a 10 30 o decrease in anilable water in the Colorado River basin. This decrease in runoff will result in shortfalls where water rights cannot be fully met (Barnett & Pierce, 2009). Climate change is affecting snmvpack in Colorado, decreasing it steadily. As most of Colorado's water is stored in snmvpack, changes in runoff patterns and decreases in o\erall snowpack can both ha,-e an impact on total water availability in the state (Pierce, 2008). Specific to the Colorado Basin Roundtable, projections suggest an increased need in both municipal and industrial water uses. Although conservation practices in the basin are recognized as one solution, they are seen as insufficient to meet the projected needs. The basin expects to meet the majority of the needs through existing water rights and implementing new projects, but shortfalls are anticipated. They arc not e\cnh distributed throughout the basin, but rather some areas are projected to experience more


significant shortfalls than others (JBCC, 2008). This matches the statewide pattern of water needs and future shortfalls being unequally dispersed around the state (CDJ\1, The Laws of the Water Colorado's complex legal structure does not tie water rights to the land, as is common elsewhere. Instead, water and land are legally distinct resources. Ownership of the land does not equate to ownership of the stream or the water in the stream (Hobbs, The Colorado Doctrine that legally binds water use in Colorado today is a consequence of the early choices that resulted in four core principles developed in the courts oYer the last 150 years: (1) water is a public resource intended for beneficial use; (2) water rights are permission to use a portion of this public resource; (3) holders of water rights han the additional right to build upon the land of others \vhen needed for diYersions, extractions, or moYement of water to where it will be used; and (4) water rights holders may use existing streams and aquifers as water moYement and storage resources (Hobbs, 2004). The combination of the four principles creates an interesting contrast. Water is a public resource, intended for beneficial use by public and printe entities, but it is controlled through a priYate market system that risks putting land and water interests in conflict. This market mechanism must haYe rules to guide it, and the ownership of rights in Colorado follows the rule of prior appropriation that declares that more senior 28


rights han first priority, with more junior rights left with access to the remaining water. The concept of "first in time, first in right" results in jeopardy for junior rights during times of drought (Corbridge, 1998). Rights are also complicated by the fact that they can be for storage or direct flow. Direct flmv rights, as the name implies, allow for the beneficial use of water taken directlY from its source. In contrast, storage rights allmv for water to be impounded for later beneficial use (Hobbs, 2004). The first two principles of Colorado water law declare it a public good and then proceed to privatize it Yia the market. The third and fourth principles take the water that has been separated from the land and made aYailable for purchase and reconnects it to the land and infrastructure on the land. Howenr, the principles do not connect the water to the land of the water rights holder, but to other priYately owned land that will aid in the diYersion, storage, and transport of water (Hobbs, 2004). The consequence of the water rights system is that despite water being a public good, water is all priYatcly mvned or owned by public entities such as municipalities and used for purposes that arc usually consumptiYe in nature (Corbridge, 1998). To further complicate water policy in Colorado, the state and its many \Vater rights holders do not "own" all of Colorado's water. As early as 1902 and as recently as 1969, national treaties, compacts, and congressional appropriates ha,e allocated water that comes from the head\vater state of Colorado to other political jurisdictions, including Mexico and Ne,ada, Nebraska, Kansas, Arizona, New l\Iexico, California, L'tah, and Wyoming. Each treaty, compact, and congressional appropriation places a 29


requirement on one or more basins to send water downstream, decreasing the water anilable to meet the growing demands of Colorado (Colorado Water Conservation Board, n.d.; Hobbs, waler !..Aw and Non-ConJiflli/Jiil'e ( TJeJ Emironmental interests are moredirectly engaged in water policy than in the past, in part due to Federal la\vs such as the National Emironmental Policy Act, Clean \X'ater .-\ct, Wilderness Act, Federal Land Policy and Management Act, and the Endangered Species Act (Hobbs, The Endangered Species Act, for example, prohibits actions that will harm the habitats of the listed species, including water management decisions that negatiYely impact habitat due to ne'v construction or changes in the flow and storage of water, whether on public or printe land (Public Law 93-205, 87 Stat. 884, 16 l' .S.C. 9). As the state has increasingly recognized and been unable to ignore the Yoices of emironmental and recreational interest groups, the current water law system has had to a\vkwardly adapt, finding wan within the prior appropriation system to address non-consumptiYe needs. The most significant adaptations of water law come through the Instream flow program of the Department of Natural Resources. Instream flow rights, 'vhich can only be held by the State of Colorado, ensure flow in specific riYers meets or exceeds defined leYels that Yary throughout the year in alignment with traditional peak and low flow pattems. The flow le,els respect both emironmental and recreational needs specific to each riYer. In practice, in-stream flow is a ,-ery controYersial program to implement. 30


Serious disagreements occur mer the determination of the needed leYcl and flow of water in the ri,ers, leading to consumpti,Te water users filing objections to specific instream water rights. One of limitations of an instream flo,v right is that it is generally a junior right to most of the rights on the riYer. Howenr, the flmv rights do require a base flow e\en during lmv flow conditions, and dictate the recoYery flow tight during higher flow conditions (Pontius & S\\'C\, 1997). By its ,-ery existence, the Instream Flow program has expanded the definition of beneficial use to include the benefits deriYed from leaving water in the rivers, not just remming it for another use. Another adaptation of water law to accommodate non-consumptiYe uses was adopted by the Colorado Legislature in 1001. The Recreational In-Channel Dinrsion Program allows local gonrnments to apply for water rights specifically for recreational uses, thus adding recreation to the definition of beneficial use. Similar to the instream flow program, the water rights define minimum stream flows, and for in-channel diversions, the flow leYel is defined by the amount needed for "a reasonable recreational experience in and on the water" (Section 3 7-92-101, C.R.S.). Conflict around recreational in-channel di,ersions has been significant and ongoing, with the Colorado Supreme Court hearing cases related to specific recreational in-channel diYersion rights as early as 2005 (Colorado Water ConserYation Board Y. l'pper Gunnison River Water ConsenTancy District, Colorado Supreme Court, 2005). Environmental interests have also been able to leYerage another part of Colorado law to meet their water needs. The Colorado Land L'se Act of 1970 gins many powers 31


to local communities, including powers that allow counties and municipalities to govern how water is denloped within their jurisdictions. Though not originally intended as an emironmental law, in practice, the Act has allowed for counties whose water is being remoYed to other jurisdictions to refuse to permit denlopment within their jurisdiction unless it meets their standards, including environmental standards. A notable recent example is the permit denial by Eagle County when Aurora and Colorado Springs sought to dinrt water from Eagle RiYer into their resenoir (Hoyt, 2005). In this context of competing interests and market forces, the management of water is complicated, time consuming, and can leave most, if not all, stakeholders dissatisfied with the results. Although the transfer of water rights has consequences for those beyond the buyer and seller, the current legal structure is not designed to accommodate these secondan Interests. A landowner, a ctty or town dependent on continued agricultural production in the region, recreational users of a riYer, or a habitat dependent on water can all experiencenegative consequences when a water right is transferred from one pri,ate party to another. Yet, water transfers are inevitable. l\Iunicipal demands are growing as population increases (CDl\f, 2006) and ranchers and farmers arc reaching retirement age throughout the state and selling their valuable water rights. Though stakeholders who are indirectly affected by water transfers have limited opportunity to stop them, the stakeholders who are directly affected have multiple opportunities to get their needs met. The legal structure for dealing with water issues 32


includes the Colorado water court, federal agencies, and small local water consenation boards. If stakeholders find one setting is not amenable to their needs, they can mme their water conflict into another (DaYis, 2001) \ll of the settings are complex and time consunung legal processes; a single negotiation of a water transfer can rake years to complete. The Lords of the Water :t\lanagement of the complex water la\v system and the infrastructure that captures and moYes water around the state requires significant resources and engages many people and institutions. Within Colorado, the public institutions responsible for maintaining the water system are many, each with distinct roles and responsibilities, and different stakeholders who ascribe expectations on them. In the ExecutiYe Branch at the state le,el, the Colorado \Vater Consenation Board, a Gonmor-appointed council, is the organization responsible for consening, deYeloping, protecting, and managing Colorado's water in the present and for future generations. The Colorado Department of Natural Resources administers programs that deal with water through a variety of approaches, from a focus on resource extraction (mining) to emironmental issues (parks, wildlife, forests). Within this Department, the State Engineer is responsible for the management of waters, and the Division Engineers under the State Engineer enforce water rights and water distribution throughout the state. The Colorado \Vater Resources and Power DeYelopment Authority focuses narrowly on the initiation, construction, and operation of water projects, such as wastewater treatment plants. The Colorado \Vildlife 33


Commission, a goyernor-appointed council, is responsible for wildlife management, which necessitates acquiring water resources to meet wildlife needs. The Colorado Department of Public Health and Emironment sets standards for water quality and monitors those standards. The Department's Water Quality Control Commission and DiYision, Board of Health, and Operators Certification Board deYelop, implement, and enforce regulations related to the treatment and management of drinking water and domestic sewage. Specific to groundwater, there is a State Groundwater Commission, which sets regulations, and local groundwater management districts, who regulate local use of wells in their districts. In the Judicial system, Water Judges and the Water Courts hear all water related legal cases, with Water Referees working with them to imestigate water rights issues (CDl\1, 2004; Frohardt, 2003). Colorado has many local leYel, quasi-gmernment organizations called special districts or Title 32 Districts (in reference to their authorizing statute), some of which prmide water senices. These districts include \Vater districts, which supply water for domestic use; water and sanitation districts, which manage water as well as sanitation; metropolitan districts, which prmide many senices, of which water supply can be one; drainage districts, designed to support agricultural landowners; ground water management districts, ,vhich manage water within a specific water basin; water consenancy districts, which can act to conserve riYer water. Additionally, Colorado also has water authorities, which contract with municipalities to sene multiple areas, and haYe the authority to own and operate water systems and facilities (Center for Systems 34


Integration, 201 0). The water consenancy districts can collect taxes to build and maintain water storage and exchange water \Vith other entities (CDl\I, 20tH). ln the Colorado River Basin alone, there are 10 water consenancy districts operating. Local gonrnments and utilities are responsible for the distribution of water to their constituents and are owners of water rights. The City and County of DenYer is the largest owner of water rights in the state, with both storage and flow rights that diYert water from the Colorado River Basin to the Front Range. Inter-governmental agencies arc responsible for much of the implementation of the Clean Water Act, including the Denver Regional Council of Go\ernments, North Front Range Water Quality Planning Association, Northwest Colorado Council of Gmernments, Pikes Peak ,\rca Council of Governments, Pueblo Area Council of Governments, and l"pper .\rkansas .\rca Council of Gonrnments (Frohardt, 2003). The local and state organizations must also interact with federal offices that hold responsibility for the regulation and management of water, including the Office of 1\.lanagement and Budget, the Department of the Interior, the EnYironmental Protection Agency, the Army Corps of Engineers, the Department of ,\griculture, and the Council on Emironmental Quality. The \'arious agencies perform a wide range of roles, including setting standards and guidelines, creating and implementing monitoring systems, managing water infrastructure such as damns and rcscnoirs, financing and building water projects, and partnering with local entities in the impron:ment of the natural em-ironment (CDM, 2004). 35


In addition to the gmernmental organizations at federal, state, and local leYels, there are also many non-goYernmental organizations who are actiYely inYohed in water policy. l\lany of the groups represent recreational interests, enYironmental interests, or a combination of both (CDJ\1, 2004). According to the Statewide Water Supply Initiative (SWSI), these non-goyernmental organizations focus their lobbying efforts around enYironmental policies and against policies that \Veaken current laws on water quality or limit opportunities for recreation (CDl\1, 2004). Many of these groups go beyond legal and policy actions to also engage in projects that address environmental or recreational issues on the riYers, such as restoration projects that engage local partners in repairing habitat harmed by human actions (e.g., Nature Conservancy, 2007). The Colorado Foundation for Water Education is a unique non-goYernmental entity in Colorado that stands apart from the adYocacy organizations because it seeks to provide non-biased information and education about \Vater to all stakeholders in Colorado. The final group of stakeholders is perhaps the most diYerse: water rights holders. Within this category fall eYeryone from small and large municipalities to energy or other industrial companies to individual agricultural producers and large agricultural enterprises. What they hold in common is ownership of water. What differs between them is their place in line to secure access to the water they own, that is, the seniority of their rights. They are supported by a legal industry that specializes in the sale and rental of water rights. For example, Colorado \Vater is a private firm that specializes in 36


facilitating the exchange of water rights and operates within and across most of Colorado's water basins (\Vater Colorado, 2008). How do the Stakeho!dm fnlerad? The growing stakeholder interests are side by side with the increasing crisis in water need ,-ersus anilabilin. The Water 2025 report identified the 1-70 corridor in Colorado's Front Range as an area highly likely to experience conflict around water, as well as the Western slope area around Grand Junction (Bureau of Reclamation, 2005). However, as we learned already, water needs and water sources are not always in the same places. \Vater conflict in the Denver metro area really means conflict throughout the state, and particularly with the Colorado Basin, the source of much of the municipal water in the Denver metro area. Not only are more stakeholders inYohed and more issues at stake than 111 the past, but conflicts 111 water are increasingly complex. An excellent example ts the ongoing lawsuit over water 111 the North Republican River basin. The basis of the conflict lies in interstate compact entitlements and junior rights to irrigation wells. Energy issues have become involved as the increasing production of biofuel crops that are heavy water users has ele,ated the need for agricultural water in the region. The Division of Wildlife has joined the lawsuit to protect a hatchery that produces approximately 40% of the warm water fish it uses to populate streams _throughout the state. Recreational interests are also engaged, as the landowners have fought back against the senior water rights holders by shutting their lands to pheasant hunters. The 37


town of Wray is also in the middle of this lawsuit, with city water coming from a well that taps the same \Vater as the irrigation wells surrounding the town. Finally, Nebraska has a Ycsted interest in the outcome, as the state is the recipient of water from this river, and compacts require dclinry of a specific amount per year (Meyers, 2008). Obviously, the water courts arc hard pressed to negotiate so many interests in a timely manner. As this complex \Vater litigation example shows, the movement and storage of water has consequences not only for land owners, but also for municipal users and emironmental and recreational interests. In response to the changing needs, the legal concept of beneficial use has gradually begun to recognize that water has value for more than just consumpti,e uses like agriculture and domestic use. Both emironmental and recreational uses of water are now recognized, but \Vith different restrictions on the rights than the restrictions on consumptive water rights. For example, in-stream flow rights, \vhich are intended to protect natural environments by maintaining necessary \Vater flow and levels, can only be purchased by a state agency and recreational 111channcl diversion rights are held only by local government entities (Hobbs, 2004). IF.hat mn Polio'maker.r do? The historical and legal oveniew is necessary to understand one of the core issues in Colorado water policy: Colorado water law is a public vision of how to manage a finite resource that is a public good, implemented in a market structure that brings many interests into direct competition. The legal structure and the limited water resources of the arid state ensure that water policy decisions in Colorado will engage 38


manv stakeholders and contain a great deal of conflict, conflict that crosses levels of government, generatlons, sectors, water basins, and ultimately, that invoh-es many passionate who are seeking to protect their interests. However, the conflict is not inherently problematic, as conflict is "a fundamental condition" of liberal democratic politics (Barber, 2004, p. 5). The challenge is to find a democratic process, instead of a rigid legal system, that is capable of dealing with the increasingly complex and conflicted policy arena. l\Iany policy options have been explored as solutions to basin by basin, municipality by municipality, use by use problems. However, each solution identified has a wide range of both benefits and problems associated with it. For example, although active consetTation measures by municipal, industrial, and agricultural users cost less to implement than new water storage projects and can stretch existing supplies, the structure of water law requires water to be put back into the river at specific (return flows) and conservation and reuse may decrease the water available to meet that legal need. Failure to put water back into the rivers has implications for whether Colorado can meet its multistate compact requirements. Other strategies, like permanent or temporary transfer of agricultural water rights to municipal or other uses the advantage of not increasing total depletions of water in a basin, but the disadvantages of drying up agricultural lands and consequently, negatively affecting the surrounding communities. No policy choice is without negatiYe consequences, some of which are emironmental, other social, and many econonuc 39


(CD:\1, 2004). To meet the agricultural, municipal, emironmental, recreational, and other \Vater needs, multiple solutions need to be implemented in concert, with careful thought on the design of each solution and its impact on local and statewide needs and populations (CDl\1, 2004). The possible policy options are many, and such things as \Vater conser..-ation including municipal and industrial conserntion and agricultural efficiency measures; deYelopment of additional storage, including new facilities and enlarging existing facilities; controlling non-natiYe plans; agricultural transfers, both permanent and interruptible, as well as rotating transfers; reuse of water, including municipal and industrial water reuse, non-potable reuse, and indirect potable reuse; and conjuncti,e use of surface water and groundwater, utilizing bedrock aquifers and allmial aquifers (CD:\1, 2004). With such a range of policy options anilable, each one harming one or more interests, policymakers in Colorado are facing a seemingly impossible challenge: to garner the political support needed to make changes in water policy that can address the looming water crisis. The Future of the Water In response to these many policy challenges, Colorado's legislature passed a law 111 ::W05 that began a regional, cooperatiYe planning model to engage stakeholders throughout the state in interbasin and intrabasin dialogues on water policy issues (Colorado House Bill 05-1177). Spearheaded by the ExecutiYe Director of the Department of Natural Resources in 2005, Russell George, the law represents a new way for Colorado to address water policies and the conflicts create. HB05-1177 created 0 40


nine water basin roundtables and one m-erarching roundtable, the Interbasin Compact Committee (IBCC). The nine roundtables represent the eight water basins and the DenYer-metro, technically part of another \Vater basin, but separated out as its own roundtable. l\lanagemcnt of water at the basin le,-el is appropriate for both consumptiYc and non-consumpti,-c uses (Mostert, Yan Beek, Bouman, Hey, Sa,-cnije, & Thissen, 1999), and in Colorado it separates decision-making into nine roundtables, each \Vith their own internal and interbasin challenges and opportunities. The \Vater roundtables' mandate is "to facilitate continued discussions within and between basins on water management issues, and to encourage locally driYcn collaborati,-e solutions to water supply challenges" (3 7-75-104, C.R.S.). The IBCC's mandate is "to facilitate the process of inter basin compact negotiations" (3 7-75-105, C.R.S.). According to HBOS-1177, each roundtable is required to include representatives appointed by each county, municipality, water consernncy district, and \Vater consenation district in the basin. One member is appointed by the two legislatiYe committees responsible for agricultural policy and 10 at large members, appointed by the previously mentioned roundtable members, represent agricultural, industrial, environmental, recreational, and domestic water interests as well as the interests of \Vater rights Nomoting members, mandated and Yoluntarily participating, arc also on the roundtables. As ofJuly 2006, the nine roundtables had membership ranging from 15 members to 53 members, with additional nomoting participation from interested parties, including eight state and federal offices. In total, the nine basins had 290 participants 41


bringing "a diYersity of interests, backgrounds, experiences, and geographic perspectives to Colorado's \Vater discussions" (Hecox, 2006, p. 1 ). The roundtable process has been given a number of tools to help it accomplish tts goal of de,eloping collaborative solutions to water problems. The contractor responsible for the 2004 Statewide Water Supply Initiative that preempted the water roundtables is conducting needs assessments in partnership with each basin. The assessments cmer basin specific consumptive needs, non-consumptive needs, water anilability, and potencial solutions from already proposed projects. A consumpcin workgroup and a non-consumpciYe workgroup are active at the statewide level to ensure a common technical platform is adopted between the basins, allowing for the comparison of assessments among all the basins. Additionally, each basin is participating in a statewide grants process, the \Vater Supply Resene Account, to direct funding to either water supply or environmental projects prioritized by the basins (IBCC, 2008). At the time of this study, only two years after the passage HBOS-1177, stakeholders were already questioning whether the roundtable process can succeed at making decisions that the many stakeholders with pmver are willing to implement. A new Governor and the ne'v Executive Director of the DeparU11ent of Natural Resources, Harris Sherman, \Vere asking for evidence of the processes value. Yet, even if the roundtables demonstrated the ability to identify solutions the member agree upon, without legal authority, it is questionable whether a collaboraciYe process can lead to legal 42


changes. A preYious Program Director with the LT.S. Fish and Wildlife Senice described the consequences of Colorado water law on collaboratin processes (Hamill, 1993, p. 5): .\nother negatiYe force is Colorado's water law. Some people belicn that Colorado's water law system is archaic. In this state, when there is a problem in this arena, it is common practice to go to court to resolYe the issue. We are trying to solYe these problems in a cooperatiYe manner, but frequently We are challenged legally \VheneYer we make a mmc. J;or example, the State of Colorado recently ftled for a water right and senral people objected to it. NO\v we han to go court and resohe those disputes. This adYersarial approach doesn't lend itself to this kind of adaptiYe management, negotiated-type of settlement. Within and between basin decisions are needed for the \Vater problems to be addressed. The conflicts at the roundtables, the worry that many water issues cannot be resohed through consensus, and the desire for a faster process are also leading some members of the water community to suggest a different solution is needed. The Colorado Basin Roundtable The Colorado Basin Roundtable is one of the larger m the state, with 50 members that not only include all of the mandated partners, but also include ''otmg representatiYes from the Eastern Plains (IBCC, 2008). The mandated members of the roundtable made a decision early on that \Vas unique from all other basins: they chose to actively incorporate the Yoices of those who were recipients of water, but did not reside in the Colorado Basin geographic area O. Pokrandt, personal communication, l\larch 17, 2008). As the Basin has mond fonvard \Vith planning efforts, they han engaged in multiple needs assessments, including a non-consumptiYc needs assessment that has 43


carefully documented all of the emironmental and recreational needs within each riYer and stream in the Basin. Additionally, the Basin is partnering with another roundtable (the Yampa/White/Green) to conduct an energy needs assessment, looking at non consumptiYe and consumpti,re issues. TI1e Roundtable reported to the IBCC in October of 2007 that the primary issues it is facing include (Colorado Basin Roundtable, 2007, p. 31 ): "Addressing shortages in headwater areas; Looking at the in1pacts of trans basin firming projects; Looking at compact deliYery impacts to existing and future in-basin water rights; Ensuring endangered species needs don't negatively impact future in-basin needs; Ensure adequate water supply for future needs: l\1&1 [Municipal and industrial], Energy deYelopment, "\gricultural, Emironmental, Recreational; Identifying the non consumptiYe needs for emrironmental and recreational flow." These needs coYer a mix of consumptiYe and non-consumpti,-e issues, with a recognition of emironmemal and recreational concerns, yet clear priority also on the consumptiYe needs in the basin. Again, it is the diversion of water out of the basin that is a priority concern, including compact delivery and transbasin projects. Conclusion To "minimize or aYoid water supply related crises," s1x principles have been proposed for policymakers and implementers to follow (Bureau of Reclamation, 2005, p. 2). These principles emphasize the market nature of water, including recognizing and respecting the rights of all of the water owners, water compacts, and other decrees related to allocation of water. They also focus on the deYelopment and maintenance of 44


water facilities, water supply infrastructure, and \Vater technology. The principles emphasize the need for water consetTation and monitoring of the resources to ensure effectiYe use. Though they do not emphasize emironmental or recreational concerns, these water principles fit Colorado's crisis \vell, in part because the fourth principle is to "use collaboratiYe approaches and market-based transfers to minimize conflicts" (Bureau of Reclamation, 2005, p. 2). In practice, the report describes this principle as when all interests come together to "plan for and meet long-term objectiYcs ... because ... endless litigation rarely, if eYer, achieves this goal" (Bureau of Reclamation, 2005, p. 4). The water roundtables arc an excellent example of Colorado seeking to put this principle into play: they do not change the nature of water ownership from a market base, but they ask that it be managed \Vithin a context that respects the complexity of the stakeholder enYironment and minimizes the ongoing between interests. The roundtables demonstrate the growing recognition that comprchensiYe water planning efforts must include a wide range of stakeholders, representing diYersc interests including emironment, agriculture, recreation, local domestic water prmiders, industry, and owners of water rights. CollaboratiYe decision-making structures like the roundtables han tremendous potential, but they also suffer from the complexity of interpersonal and interest-based stakeholder interactions, to be explored in more detail in the next chapter. 45


CHAPTER3 LITERATURE REVIEW In the west, where water is scarce and needs are many, public policy risks creating a situation of distributiYe injustice: a lack of fairness in how a good that affects indiYidual and social well being is distributed among those in need (Deutsch, 1985). In the water policy community, the trend is to use collaborative decision-making processes to address these Yalue-driYen problems, with the assumption that there is value to participation by those affected by problems and by the potential policy solutions (Leach & Pelkey 2001; Leach & Pelkey, 2002; 1\Ioore and Koontz 2003; Lubell 2004). Yet, the inclusion of these stakeholders leads to an expansin list of potential interests that need to participate in any collaborative water policy process. When the public participates in a process, democracy itself is served, as participation can address the failures of representatin democracy (deLeon 1992; Petts 2001 ). This expanding participation is necessary to address increasingly complex policy problems (Dryzek, 1996) and support fundamental democratic nlues (Renn, et al., 1995). The water roundtables described in the prenous chapter are explicitly collaborative structures that ex1st 111 this institutional Yoid, where water policy crosses existing structures, levels, and institutions of goYernment and society, typical of policy 46


problems that lead to participatory processes (Hajer & Wagenaar, 2003). These interactin policymaking processes haYe been explored in depth by the public participation theorists, who recognize the importance of the process and participants to making public decisions. The roundtables are an example of such a process, going far beyond the citizen as adYisor to a "citizen goycrnance" model where the elected officials and their representatins become coordinators instead of decision-makers (Box, 1998). When such a process is "authentic," it brings citizens into all aspects of the decision process, from problem defmition to solution identification to implementation (King, Feltey, & Susel, 1998). Howe,er, the public participation literature does not fully address the inherently political nature of this type of decision-making process, where different interests sit at the table, not just the public at large. The literature reYiew will follow a discussion of public participation theory with an introduction to a public policy theory that suggests that beliefs play an important role in how actors within a policy system work together. The research propositions are presented within this framework of participation processes being dependent on successful interactions of ind.i,-iduals who are representatiYe of interests. The argument made here relies upon two bodies of literan1re; fust, public participation literature, with its emphasis on process and the indiYidual participants. The process literature includes the goals and approaches to public participation and criteria for judging success, while participation literature explores the concept of representation; and second, public policy theory, with its emphasis on coalitions, where individuals 47


beliefs are important, but the individual actions are lost in the collectiH action of the coalitions. The Roundtables as a Public Participation Process To better understand who should participate in the water roundtable process, the public participation literature is an important road to follow. First, we investigate the theoretical justification for participatory processes, followed by a review of why public participation is necessary in the current, complex policy environment. An oven,iew of the history and study of participatory approaches prmides insight into their use and nlue to policymaking at all levels, including the local and state levels that are so relennt to water policy. An examination of participation models and criteria for measuring their effecti,eness helps to understand how the quality and type of interactions occurring in participation processes are ,ital to their success. From there, the literature review narrmvs in on the concept of representation as one component of a successful participation process, and explores the definitions of representation and typical strategies for creating representative processes. Pal1itipation a.r a j\'ece.r.rao' Component Philosophers and theorists examining the concept of democracy have engaged in a debate over the role of representation versus participation for centuries (Cunningham, 2002). Participatory democracy theorists argue for a form of democracy that "will expand opportunities for participation and democratic control not only in smaller units where the democratic process could be greatly strengthened, but in the larger units [of 48


gmernment] as well" (Dahl, 1989, p. 231 ). The argument is that "state and civil society are not distinct entities" and that "there is no line dividing a state that rules and citizens in civil society who are ruled" (Cunningham, 2002, p. 127). The emergence of a modern society where governance is complicated, pmblems are interconnected, cooperation and coordination are necessary for implementation, and government cannot lead alone (Hajer & Waagner, 2003; Milward & Provan. 2000) -has replaced the academic question of, "Should the public participate in gonrning?" with the practical reality that the public is participating. That said, the idealism of participatory democracy theorists is not so easily achieved. Participatory democracy is not just about the act of participating in decision-making, but the quality of that decision making process. A participatory democracy includes deliberati9n and discourse 1983), with representation of a broad cross-section of the affected public (Parkins & 2005); such dialogue, when conducted fairly and with competence, is proposed as the ideal form of democratic decision-making (\X'ebler & Tuler, 2000; Beierle, 1999). This space bet\veen the idealism of participatory democracy and the need for practical ways to engage the public in democratic decision-making is filled by the public participation literature. Public participation is a tool to build collective action bet\veen gmernment and citizens. It is a tool that raises awareness on the responsibilities of citizens, creates pmcesses that are more open, inclusiYe, and transparent, and recognizes that democracy is an inherently normati,-e process that requires inclusive processes to adequately capture both the depth and breadth of debate on public policy issues 49


(deLeon, 1997; Barber, 2004). When the public participates in a process, democracy itself is sen'ed, as participation addresses the failures of representatiYe democracy (deLeon, 1992; Petts, 2001). As more than just a theory but also a practice, public participation deriYes its nlue from the steadily increasing research results that identify the connections between successful policies and the participation that leads to them; these research results suggest that goYernance is imprond when the public is inYolved (Schneider & Ingram, 1997; Narayan, 2002). Some of the researchers exploring the participation of affected publics outside of electoral politics arc focusing on public participation processes as a place for representatins of public interests to shape policy (e.g. see Weber, 2000; Kenny, 2000; Petts & Leach, 2000; Curry, 2001; In'in & Stansburg, 2004; Rowe & Frewer, 2000; Rocldoff & l\loore, 2006). A high quality public participation process has tremendous potential, perhaps nmv more then eYer, as, in the context of complex modern society, public participation is changing. We are beginning to understand that deliberation and participation need not be the proYidence of elites, as public participation can be done well through thoughtfully constructed processes (Kathlene & Martin, 1991) and can engage the public in a deliberati\'e and democratic process without requiring "people to dcYote their li\'es to political participation" in order to han a ,-oice in the decisions (Guttman & Thompson, 1996). Calls for processes that go beyond public inYoh-ement and beyond interest group inYoh-ement propose that "participation must be collaboratin and it should incorporate 50


not onlv citizens, but also organized interests, profit-making and non-profit organizations, planners and public administrators in a common framework where all arc interacting and influencing one another and all are acting independently in the world as well" (Innes & Booher, 2004, p. 422). Participation in this new context is no longer about meeting minimum requirements for public hearings or conYening focus groups on targeted issues. Rather, it is about participatory policymaking, where stakeholders across many sectors interact with decision-makers in democratic processes that allow for authentic dialogue (Dryzek, 1990). 'This is not one-way communication from citizens to government or government to citizens. It is a multi-dimensional model where communication, learning and action are joined together and where the polity, interests and citizenry co-evoke" (Innes & Booher, 2004, p. 422). Innes and Booher (2004) are describing an interactin policymaking process like the water roundtables, where public and private, interests and indiYiduals, are all engaged in an ongoing dialogue of policymaking, supported by resources and gonrnment actors as coordinators instead of decision-makers. Whatever term is used, from citizen governance to authentic participation to interactin policymaking to water roundtables, the same questions can be asked: "With the realization that the process of policymaking has to be participatory comes the issue of who to participate, and how much to 0 0 "(\V! 1 ?000 'J?J) partlctpate ,., age, p. 0 51


A1ode!" Pmtitipation To address these questions of who and how, examples of public participation processes abound in both theory and practice. Before exploring models, however, it is important to understand that the goals of public participation can Yary greatly, driYing the selection and design of \videly Yarying processes. Understanding the goa!r and tmde!fring pou;er issues in public participation is an important foundation for exploring the structures through which citizens are brought into decision-making processes. Though the Yision of public participation is abstract and ideal, supporting democracy and good goYernance, the goals for participation can be very concrete. The International Association for Public Participation (2000) lists fi,e broad goals of participation, recognizing that when a participation process is undertaken, sometimes it may seck to achieYe the Yision of democracy, but other times, it does little to moYe beyond the traditional, institutional decision-making processes. The least empowering of the goals, "Informing/' refers to public participation efforts where the public is giYen information to help in their understanding of the issue, but not to help in making any decisions. Consulting, inYolving, and collaborating are goals of participation processes that increasingly engage citizens in decision-making. "Empowerment," is the opposite goal of informing and occurs \vhen a participation process matches the top of Arnstein's ( 1969) ladder where the public has the fmal say in the decision. It is worth noting that \vhile Empowerment is the top of Arnstein's (1969) seminal work in public participation, she also emphasizes hea,ily the failings of participation, much of which she sees as 52


driYen by the goals of the process. Her exploration focused on the power sharing that can but often does not occur when the public is brought into a decision process. She explained that sometimes citizens are brought into processes for legitimate imoh-emcnt, but other times they may be engaged for manipulatin or merely informatin reasons. Arnstein (1969) recognized, along with many theorists who followed her (e.g. Box, 1998; King et al., 1998) that unless we examine our goals and then our models for participation carefully, we cannot avoid our participatory processes from disempowering their publics, rather than empowering them. Many other authors cite goals of the opposite nature m the participation processes they haYe studied, with a positive focus on decision-makers seeking different levels of imolnment, the type of education and decision making occurring (Kathlene & 1991; Creighton, n.d.), or the stage of the policy process where participation is seen as necessary, or even ,ital, to good decision-making (\X'alters, Aydelotte, & 2000; Petts, 2001 ). To meet such goals, practitioners and theorists han collectinly undertaken the design and exploration of new models and the practice of participation has grown over the decades. The new models of citizen participation differ from the past in their emphasis on collaboration and representation. \X'here participatory processes were once seen as citizen engagement ,ia public testimony, focus groups, or other narrow mechanisms, participation is now understood by some as a multi-sector, multi-stage collaborati,e process. The learning that occurs as cmzens and interest groups work together to solve an Issue creates new capacity for innovation. 53


Consequently, participatory models are being used across sectors and for many reasons, including emironmental rule making, budgeting processes, creating alternatives to litigation, and enn statewide planning for limited water resources (Innes & Booher, 2004; Inin & Stansburg, 2004). The models of participation used in these Yarying sectors are categorized through many of the same concepts as the goals of participation, explored aboYe, including the leYels of empowerment inherent in the models (e.g. see l\lartin & Carson, 1999; Arnstein, 1969; Petts, 2001; International Association for Public Participation, 2000). and Carson (1999) prmide a laundry list of onr 50 participation approaches used in denloped nations around the world. They broadly fall into the categories of pa.oil'e and adi11e con.wltation.r, and pmtnenbip. The participation methods that fall into adi11e c"!ation begin to approach the desires of early theorists for empowerment, and the few models that fit into pmtner.rbip han the potential to truly achieve this goal. In some ways, the Colorado Water Roundtables fit best into the category of parluer.Jbip due to the cross-sectoral collaboratiYe decision-making design, though the lack of authority to implement decisions does decrease their position on Arnstein's (1969) ladder of empowerment. The models that fall under l\Iartin and Carson's (1999) categories of informing and passive consultation are frequently used. Some authors argue that their oyeruse is not arbitrary, rather a response to the many problems with participation processes. The problems include public administrators belie,ing citizens are too uninformed to be 54


helpful, ctUzens lacking time to become adequately informed on the issue, the susceptibility of citizens to manipulation by interest groups and politicians, meeting based participation that fa\'ors some citizens over others, and apathetic citizens who do not wish to be involnd, particularly when they are unsure if their participation will effect the final decision (\'\'hite, 1997; Arnstein, 1969; l\lartin & Carson, 1999; Busscmakcr & Voet, 1998; Abelson, Forest, Eyles, Smith, l\lartin, & Gauvin, 2003). Public administrators may also reject public participation approaches because are objecting to sharing power, are uncomfortable with the change in their role, or lack the resources or time to undertake the process (Walters, et al., 2000). The perception that the public is unwilling to participate may not be entirely accurate, as we know that citizens arc more likely to engage in public processes when their interests are affected (Abelson er al., 2003), indicating a successful process should include clients of the policy system who arc directly affected. In reading about the many limitations of public participation and the barriers to its success, it is impossible not to fear that many participation processes han little to do with citizens truly shaping policy. With decades of problems arising from public participation, resistance to am type, much less the complex, multi-sectoral models proposed by authors like Hajer and Wagenaar (2003) may well be expected. Howenr, it \vould be a mistake to focus on the problems to the exclusion of the strengths. Public participation raises awareness on the responsibilities of citizens, increases their decision-making capabilities, and creates processes that are more open, inclusive, and transparent (Fischer, 1993; Habermas, 55


1989). Such processes recognize that democracy is an inherently normative process and requires inclusive processes to adequately capture both the depth and breadth of debate on public policy issues (deLeon, 1997; Barber, 2004). l\lore pragmatic reasons for public itwolvement include engaging the knowledge citizens have of their communities to shape better policies and building commitment among citizens to the policy decision. Without "purposeful participation" in the policy process, a sense of ownership and support of the policy is unlikely to emerge in the conununity affected. Processes that engage in a two way reciprocal dialogue create the opportunity for both pragmatic reasons to be met (Curtain, 2003). Such processes are not easily created and the design of successful public participation is, even before its implementation, a significant undertaking. S!IL"L"eJ.f in Pm1itipation As participation changes and expands its role, it is increasingly important to understand what can make these new processes succeed. l\Iuch of the literature from classic public participation studies can help in understanding the new models of participation. Yet, despite the work deYoted to this topic, the knowledge of specific components of public participation processes come largely from theoretical and anah-tical models (e.g., Arnestein, 1969; 1\lajone, 1977; Dryzek, 1990; Fischer, 1990; dcLcon, 1990, 1992; Dunn, 1993; Healey, 1993, Valelly, 1993; Weiss, 2000) and examples of case studies (e.g., Crosby, Kelley, & Schaefer, 1986; Kathlene & l\lartin, 1991; Hajer, 1993; Petts, 2001) that largely lack comparative data or generalizability (e.g., Rosener, 1978; Lynn & Busenberg, 1995; Rowe & Frewer, 2000). The challenge left unans\vcred 56


by the literature is what leads to a successful process that produces the desired outcomes, and what must be included in the design of a process (\Vebler & Tuler, 2000). l\Iissing too are contextual models that proYide policymakers with direction as to what types of citizen inYohement are best suited for specific policy and planning processes, such as why and when are adYisory groups, neighborhood meetings, focus groups, or sutTeys best utilized (Rowe & Frewer, 2000; Inin & Stansbury, 2004; ,\bclson et a!., 2003); yer, if the need for participation is greater now than eYer, more successful, more democratic participation processes must be created. Further, if participation is about deliberati,e, ongoing dialogues, not just one time decision-making, successful models must be institutionalized, used in a wide Yariety of public decisions, and must meet the needs of not only the community, but also the public officials, be they office holders or administrators. Those in gmemment institutions are taking on job of engaging citizens (deLeon, 1992), creating or allowing the creation of participation processes. Thus, the future of participation includes public senants, often ,-iewcd as the barriers to participation, becoming the facilitators of it by undertaking the needed changes and shifting the approach to policymaking (King et a!., 1998). And, indeed, these public officers are seeking to become the facilitators, be it requirement or recognition of the value of participation that drins them. They are seeking information on not only how to engage citizens, but also on how to do it well (Petts, 2001). The criteria deYeloped by researchers to answer these practical questions haYc been heavily focused on process. A particularly compelling set of criteria for a successful 57


process is based on Habermas' theory of communicative rationality. The theory itself frames the concept of ideal speech, a level of discourse for which we can strive and that will transcend the institutions of society to address rapidly changing needs through a new framework (Habermas, 1981 as cited in Innes & Booher, 1999). The theory aligns well with the basic principles of collaboratin processes, making it a particularly strong fit for more dcliberatiYe models of participation. Authors building on Habermas' theory agree with the focus on levels and quality of interaction and argue that a deliberati,e process must be more than participatory: it must be fair and t"ompe!ml (Wehler, 1995; Beierle, 1999; Wehler & Tuler, 2000). A fair process includes equal distribution of opportunities to participate in all aspects of the process, to include agenda setting, procedural rule dcnlopmcnt, and the selection of information and expertise to inform the decision process. is not just the equal distribution, but ensuring everyone who "has a legitimate right to participate" is able to participate (Wehler, 1995, p. 52). Legitimate right, Wehler acknowledges, can be challenging to define, but should include all "individuals or groups whose interest or \'alues may be affected by the problem or proposed decision action" (p. 52). Beyond being able to participate,jaimm is also about the lenl of participation, such as the opportunity to influence the final decision, to challenge the beliefs of others, to bring information to the process, and to participate in the discourse (\'\lebler, 1995). The challenge to a fair process is the inability for a delibcrati,e process to include all possible affected parties, and so the authors give a 58


means for addressing this problem: when people must be excluded, at least do it randomly. Colllpelmt"e of a partlctpauon process focuses on the participants themsekes agam, asking whether each participant 1s able, cognitinly and linguistically, to fully paructpate, has the knowledge to paructpate, and whether the process uses reliable techniques for dealing \Vith conflicting knowledge (Webler, 1995). This assumes a lenl of sophistication for, at minimum, the facilitators of the process, though perhaps it also expects the same of all of the participants. One could question whether a fair and t"ompetenl process is possible, giYen the potentially conflicting needs for all affected parties to be able to access the process, and all participants in the process to be cognitiYely and linguistically capable as well as knowledgeable. It is important to note that neither criteria, fair or i'OIJ/peten/, addresses whether the final decisions resulting from the deliberative process are high quality; rather, they assume that processes that meet these criteria will result in high quality decisions. Despite these significant limitations, the two criteria are useful, for they define the process as requiring more than just deliberation, but well informed deliberation, both from the perspectiYe of the interests represented through the affected public and their ability to represent those interests. These two criteria, faime.r.r and t'ompetence, eYaluate the process of participation with a focus on the interactions of the participants in the process, beginning with their opportunity to be imohed, continuing through their opportunities in each stage of invohement, and inclusiYe of the characteristics of the discourse throughout their 59


inn>h-ement. Ginn the emphasis on the individual participant, Beierle (1999) opcrationalized the faimu.r and (ompetent"e criteria in a logical way, by identifying the type of representation as one of four critical dimensions. In his consideration of representation as a component of participation, Beierle explains the advantages of processes where the public is represented directly, rather than when the public is represented through representatives of interest groups. These processes arc "better at achieving the goals of education and trust formation than those where the general public is represented by "representative" members or professionals" (Beierle, 1999, p. 89). Beierle (2002) more clearly articulates the difference between public and professional \vhen he informs the planner of a participatory process that one can either use demographics or interest groups (Beierle, 2002). One can only conclude that Beierle (1999, 2002) sees demographics as the preferable means for securing "public" paruc1pation in a participatory process. Other authors using the faime.r.r and compe!en(e model have also emphasized representation heavily, arguing its importance based on Habermas' statement that "ideal speech," a central part of his theory of communicative rationality, cannot occur without all interests beio.g engaged in the discourse (Innes & Booher, 2000; Bickerstaff & Walker, 2005; Guttman, 2007). The emphasis these authors place on representation is not unique. The body of literature on public participation includes an extensive exploration on representation in policymaking, much of which has matched Beierle's (2002) interest in demographic representation. 60


RepreJenlation in Pm1iLipa!01J' PmaJJeJ Researchers exploring representation outside of electoral politics are focusing on public participation processes as a place for rcpresentati,es of public interests to shape policy (e.g. see Weber, 2000; Kenny, 2000; Petts & Leach, 2000; Rowe & hewer, 2000; Curry, 2001; Inin & Stansburg, 2004; Rockloff & t\loore, 2006; ). There is implicit recognition in the study of representation in participatory processes that the ideal of a participatory democracy is impossible in the modern democracy; that is, despite the potential \'alue of a fully participatory democracy, society and its problems haYe become too large for eYery citizen to participate (Dahl, 1989). For this reason, eYen a participatory process is composed not of the public at large, but of indi,iduals representing the public (Beierle, 2002). With the recognition that enn in the larger settings of open-meetings such as town halls, participation is not extensiYc, processes arc not democratic (Webler & Renn, 1995), and outcomes do not come from deliberati,e dialogue, if for no other reason than the practical limitations on how much time each participant can realistically speak (Dahl, 1989), the question of representation is of immediate concern. The more intensi,e participatory processes may be most likely to solve problems through discourse, but are also the processes that most exclude the wider public (Beierle, 2002). The study of representation has found that participatory processes arc often not representative even when intended to be (?\loote & 1997), and that the public may be aware of this problem (Inin & Stansbury, 2004). Curry's (2001) case study of 61


communitY parttctpatton tn rural commumttcs found that some participants openly admitted to pursuing their personal agendas, rather than the agenda of the broader group they represented. Other studies haYe compared the demographics of participants to non-participants, finding that participatory processes draw more educated and higher mcome members of the community (Weber, 2000). The picture painted by these authors ts one of participation processes dominated bv actors who neither demographically resemble the interests they are expected to represent nor are likely to be pursuing outcomes desired by others with similar demographics or interests. The lack of processes where interests are represented has consequences for the success of the new participatory models of goyernance as representation is not a practice in inclusi,eness for the sake of inclusiYeness. Rather, excluded stakeholders pose a threat to a participatory decision-making process. If the stakeholders haYe influence and/ or resources, they can, and indeed often do, undermine the decisions made by the participatory body (Susskind, l\.fcKearnan, & Thomas-Larmer, 1999). E'Ten when stakeholders do not ha,-e the ability to undermine a process purposefully, their exclusion may still undermine the legitimacy of the process in the eyes of the public (Innes & Booher, 2000). From the perspectin of public participation literature, representation of different interests is not just about the credibility of the process, but also about the quality of the decision. DiYerse participation brings more expertise, including liYed experience expertise, to the decision-making process. EYen the most technical policy 62


issues include decision points that are nlue laden, yet another reason to be inclusiYe in the participation process (Rowe & Frewer, 2000). The authors \vho have found that participatory processes are not demonstrably representati,-e haYe not ahvays defined representation in the same way. There arc two broad, merarching approaches to defining representation, one tl1at comes largely from the literarure on public participation in narural resource planning where the focus is on representation of all affected interests, and the other from the literature on deliberatiYc democracy and public participation more broadly, where the focus is on inclusion of all concerned citizens, not just those whose personal or professional welfare may be directly affected by the decision-making process (Parkins & 2005). Natural resource planning theorists who argue for balanced interests prmide recommendations of limited help, such as asking participatory processes to "actiYely ensure that certain groups and stakeholders show up at meetings" through aggressin marketing approaches (i\IcCool & Guthrie, 2001, p.322; Tuler & Wehler, 1999). Such suggestions say nothing about how the process identifies who the interests are or who can adequate represent them. Deliberati,-e democran theorists focused on understanding participarory processes complain that "traditional agency public participation procedures tend to foster participation by organized interest groups while limiting participation by the general public" (Moote, :McClaran, & Chickering, 1997, p. 878) and ask that participation processes engage the "general, non-acti,ist" public (p. 879), but do not han a definition for full inclusion that is practical or goes beyond ngue recommendations such as 63


"identification of all members of the public who may han concern" (Portman, 2009, p. 335). Regardless of which body of literature is followed, there is little guidance on how to identify and select the desired participants in order to ensure representation of all interests or all concerned citizens. The literature instead assumes "that stakeholders are sclf-eYident and self-construed, and has focused on categorizing existing stakeholders to understand their interests and relationships" instead of developing strategies for identifying stakeholders (Reed, 2008, p. 2424). Perhaps this limited focus is a consequence of the sheer complexity of the issue: when it has been considered, researchers ha,e proposed significant efforts to identify the interests, such as focus groups, semi-structured interviews, sno\v-ball sampling, (Reed, 2008), and questionnaires to determine the range of interests (Crosby, et al., 1986) without outlining the next step of identifying the stakeholders who can best represent the newly identified interests. Although they do not soln this dilemma, Petts & Leach (2000) expand the debate by defining representation through clarifying the difference between "repre.renting intere.r/.r and being reprmntaliJe intereJt/' (p. 21). In the design of representative processes, the latter definition is the easier to address. For example, the legislation authorizing the Colorado water roundtable process sought to create a repreJf!lllation inlt:J!!J/.f by mandating participation of necessary interests at eYery roundtable. The process went beyond the commonly represented interests, the municipalities and historical water leaders to include the unusual suspects, such as the counties within a ginn water basin. The legislation also explicitly included recreational and emironmental 64


actors. While the water roundtable process, with its emphasis on interest groups and alignment with the public participation literature from the natural resources arena, can be fairly argued to be reprexenlalil'e interex!J, it would be harder to demonstrate that it npremrlx intereJ/J. Petts and Leach (2000) define repre.rmting inlereJIJ, not as relating to demographics of participants, but instead as the specific agendas the participants bring to the participatory process and whether they match those of the interests they arc charged to represent. Petts and Leach's (2000) concept of truly representing mterests 1s more comprehensive than simple representation by being part of the interest group and more specific than arguing for inclusion of the general public. But is it sufficientlr comprehensive and specific to truly understand how representation works in a participatory process? There is e,idence that the public has more trust in participatory processes and the people participating m them when they believe there arc representatiYes of their interests involved in the process (Smith &, McDonough, 2001; Hoppner, 2009; Rockloff & 2006; l\IcCool & Guthrie, 2001). Rockloff and (2006) learned that representatins of the public interests were expected by the public to actinly participate, be competent in both skills and knowledge of the issue, adopt and demonstrate commitment to the identity of the group they are representing, communicate with those outside the collaborative process, use their established social networks to learn and communicate, and be capable of functioning in multiple roles, wearing "different hats," so to speak. These criteria are defined in ways that emphasize 65


usmg network relationships to share information, both inside and outside the collaborative emironment. The ability to accurately represent the interests of the individuals outside the collaborative process is defmed as when the "representation of the group in the wider conununity suggests presence of group identity" (Rockloff & l\Ioore, p. 656). The authors attached the term "credibility" to their criteria: the participant who can meet the expectations of their conununity and deli,er tangible results has greater credibility as a representative. From Rockloff and 1\loore (2006) comes the concept of networks and information sharing to help achien the results. Howenr, this might be extended further. Critics of the faimeJJ and (0111pe1ence evaluative framework, based on Habermas' theory of communicative rationality, han noted that the framework fails to take into account pmver in the participatory process (Bickerstaff & Walker, 2005; Santos & Chess, 2003). In a similar ,ein, Inin & Stansbury (2004) note that ideally the "community representatives with particularly strong influence are willing to sen'e as representati,es" of the different interests (p. 62). Thus, beyond representing the interest, sharing information, and having active networks, the credible representative must also have influence in the policymaking arena. This is consistent with the emphasis Arnstein (1969) places on pO\ver as a defining feature of a participatory process. If truly representatiYe processes are necessary for successful outcomes, then the abilitY to identify when a process includes credible representation is needed. Many authors haYe proposed representation as an important variable to consider when 66


eYaluating a participation process. HoweYer, most define representation narrowly, or enn when it is defined more broadly, proYide little guidance on how to e,aluate whether the representation is credible. It is important to note that many authors studying the issue of representation in participatory processes are not narrowly examining the type of collaborative processes that are emerging in the context of complex policy emironments. For that reason, then discussions of representation often focus on public participation processes such as surveying, focus groups, public opinion polls, public meetings, panels, and other non-collaboratin participation models (e.g., see Smith & McDonough, 2001; Kathlene & Martin, 1991). For example, Rowe and Frewer (2000) suggest that representation is an important part of ensuring a participation process and results in a decision that will be widely accepted. They state that "the public participants should comprise a broadly representatin sample of the population of the affected public" (p. 12). Their discussion focuses hea,ily on how to pull a sample of participants who represent the public and it makes the assumption that if representatiYes of the public interests are included, representation of interests \vill occur. They include no discussion of how the representatins behaYe within the participatory process. Another study that focused on non-collaborati,e participation processes explained representation in similar terms: "A public participation program designed to meet interest representation objectiYes would actiYely ensure that certain groups and stakeholders show up at meetings" & Guthrie, 2001, p. 322). 67


Although many researchers use demographics as theu prunary measure of representation m a participatory process (Smith & McDonough, 2001; Kathlene & l\lartin, 1991; Rowe & Frewer, 2000; McCool & Guthrie, 2001; Marshall & Jones, 2005; Jackson & Shade, 1973; Gundry & Heberlein, 1984; Checkoway, 1981), some researchers han begun to explore other measures. Beierle (2002) proposes that representation is either (1) demographically-based, or (2) interest-group based, and though he proYides some guidance to a participation planning on when to use the two types, he gi'es no insight into what it means to select participants based on the interests they represent. Pctts and Leach (2000) haYe proposed criteria for measuring whether a participation process is sufficiently representati,e of the interests. Their three criteria are: (1) whether participants represent both the people and interests that are potentially affected by the decision; (2) whether the process of selecting the participants was objective and open to scrutiny; and (3) whether participation remains open as the process moves forward, in both the formal rules of the process and the perception of those imolYed as to the Yalue of adding new participants. The criteria arc focused heaYily on the representation of interests, but also mention the representation of both the people and interests. L nfortunately, the authors proYide no examples of measures to get at this potentially broader definition of representation. Similarly, Leach (2006) has described a representatiYe process in terms of both perspectiYes (urban and rural, environmental and economic) as well as demographics, but does not prmide any measures for capturing the perspectiYcs of representati,es and those they represent. Rather, the measures of 68


whether the participants arc representing the interests of their faction arc limited to general questions like "The partnership represents the interests of most people in the local community." While the question may help to get at satisfaction with representation, it does not allow for any comparison of the perspectiYes represented. If public participation theorists are arguing for processes with representation of interests, howeyer broadly or narrowly that is defined, it is worth noting they arc not alone. Increasingly, network goyernance literature, which also argues for public decision-making through participatory structures, joins in this assumption by arguing that whether goYernment likes it or not, the "public" includes interests \vho arc needed for decisions to not only be made, but also implemented. And these interests arc not just the traditional interest groups, but rather indiYidual members of the public who may not come from any formally recognized interest group, but represent an interest with a stake in the outcome of the process (Hajer, 2003). Individuals with a vested interest (not to be confused with indiYiduals representing an interest group) are, by definition, the people that public participation processes must engage. The Roundtables as a Public Policy Process The \Vater Roundtables are not just a participation processthey are intended to address the unique complexities of the current political and policy enYironmcnt. As Hajer and Wagenaar (2003) explain, in the age of the complexity we currently li\-c within, policymaking occurs in an emironment in "hich organizational boundaries arc fluid, adhoc structures are an important part of finding successful policies, and participation 69


moYcs far beyond the usual suspects. The structure of the roundtables respects that politics no longer occur within the boundaries of organizations and nor do policies. With policy problems and solutions existing m an "institutional void" (Hajer & Wagenaar, 2003, p. 9) and a context of "radical uncertainty" (p. 11 ), there are no predetermined rules, defined membership, or even shared language and discourse. Policymaking and its participants han become a practice of interdependence among the usual and unusual suspects, where collaboration across differences is required for solutions to be found -\s explored previously, the literature from the participatory democracy theorists has focused on the "public" most broadly defined, and has rejected the interest-based focus of other authors. Yet, when we look at examples from literature focused on policy and governance (e.g. Hajer, 2003), eYen the concept of the "public" is comprised of those with a vested interest in the policy issue. They are "interested groups," rather than the general public or the formal interest groups. When a public participation process selects its representati,es demographically or otherwise seeks a more general public, the process is ignoring the importance of vested interests. Here, we dinrge somewhat from the public participation literature, which, as we have explored, has often gone down the road of random sampling, citizen panels, and an oYerall focus on the general public or the affected public. The public policy literature takes a much different ,iew, particularly from the aforementioned deliberatiYe process criteria of fairness and competence. L'nlike Beierle (1999) and other public participation theorists, public policy theorists do 70


not assume that when a deliberatiYe em-ironment is created with equal representation, a good structure, and good quality of information, that it will logically lead to a good quality of outcomes and decisions. Rather, it assumes that people arc not reprcscntatiYes of their community, but rather members of coalitions of interests, and these coalitions are where decisions can be made. Tbe Cout1ilmtion Pub/it' Po lit]' Theon' Compared to public participation theorists, policy theorists take a much less egalitarian look at how public decisions are made and implemented. Although policy theorists haYe traYeled down many different roads of thought, the theories that focus on the actors outside of gmernment and their interest groups, interactions, and shared action are the most releYant to a study of public participation. They arc the theories most able to incorporate the concept of participatory democracy, for though they do not explicitly include the concept, they describe a policymaking enYironmcnt where the participation of the public with a ,ested interest (interested groups) and the organized public (interest groups) are the driYing force behind policy change. The concept of the more formal intert'JI gro11p was de,cloped 111 the 1960s (rruman, 1960) and established as a central component of a policy change process by theorists in the 1970s, including its articulation as a component of the irou where goYernment and interests groups exist in a stable relationship by which policy is influenced. Heclo (1978) placed interest groups into a broader concept of an issue net\vork, composed of informal alliances benveen actors, where participants may mme 71


between groups, and coordinate to achieYe change through collecti,e acuon. In a rejection of the imn li7.allglf theory of gonrnance, Heclo established the connection between the interest groups and the elected officials by not separating them, but rather combining the two into the issue networks. Though Heclo is not speaking about the general public, his theory has a commonality with participatory democracy theorists in that he is not establishing a separation between gonrnment and ciYil society, but rather articulating a model of how society organizes to influence and interact with gonrnment. i\fter Heclo came many other policy theories that also enYisioned a policy environment influenced by interests groups engaged in shared actions. Rhodes (1985) kept the network vision of policymaking, but abandoned Hecla's (1978) open network in fayor of defining a policy community as a closed environment with stable relationships that influenced policy while remaining insulated from the broader public and enn elected officials. Later theorists maintained the role of interest groups and began to attach the concept of shared beliefs playing a driving role in forming the coalitions that inform the decision-making process. The different theorists han defined Yery differently. Some go all the way to deep core beliefs (Sabatier & Jenkins-Smith, 1993), while others focus largely on beliefs about recipients of policies (Schneider & Ingram, 1997). The Advocacy Coalition Framework (ACF) is a policy process framework that contributes a unique combination of networks of actors coming together into coalitions to influence policy decisions and participant beliefs influencing the creation of those coalitions (Sabatier & Jenkins-Smith, 1993). By incorporating levels of beliefs into the 72


extstence of groups that take shared actton to change decisions, ACF adds a new definition to the concept of who represents an interest, tgnonng the idea of demographics in fa,or of the importance of beliefs. According to Sabatier and Jenkins Smith (1993), indi,iduals "relate to the \vorld thro{Igh a set of perceptual filters composed of preexisting beliefs that are difficult to alter" (Sabatier & Weible, 2007, p. 194), and advocacy coalitions are composed of such indiYiduals. In the context of ACf policymaking, the indiYiduals in advocacy coalitions han three types of beliefs: deep core, policy core, and secondary aspects. Deep con cross policy issues and address such fundamental issues as definitions of liberty and equality. These beliefs are seen as nry difficult to change. Polio core are the consequence of applying deep core beliefs to a specific policy system, and sene as the basis of setting priorities and assumptions about what will address a policy problem. Finally, a.rpl'd.r are defined as the specific approaches to decisions and information within the coalitions' approaches to the subsystem. Policy learning is a component of how change occurs in secondary aspects, and is the long-term process of policy analysis becoming part of the subsystem as well as the more immediate influences through indiYiduals, including policy brokers (Sabatier & Jenkins-Smith, 1993). Though coalitions and their competition are at the core of ACF and its explanation of policy change, it would be remise not to acknowledge the additional components of the theory. They haYe less relennce to the exploration of a specific participatory process, but haYc merit as part of the larger theory, and explain policy change as a process that occurs onr time through a 73


combination of policy learning, external factors such as econonuc conditions, and constraints and resources within the policy subsystem. Studies on ACF haYe contributed to understanding the policy process in a ''ariety of different arenas since the late 1980s, with oYer 80 published ACF case studies (Sabatier & Weible, 2007). Nearly half of the policy case studies since 1999 han been emi.J:onmental or energy issues (Sabatier & Weible, 2007) with regulatory needs, not unlike the Colorado water policy enYironment. For example, ACF scholars have analyzed L'.S. energy policy (Jenkins-Smith & St. Clair 1993), nuclear waste disposal (Jenkins-Smith 1991 ), national forest wildness policy (Davis, 2009), auto pollution control (Jenkins-Smith, Clair, & Woods, 1991; Zafonte & Saba tier 2004), and many different water policy issues, including San Francisco Bay water policy (Zafonte & Sabatier 1998), the Life Protection Act (Weible, Sabatier, & Lubell, 2004), emironmental policy at Lake Tahoe (Sabatier, Hunter, & McLaughlin, 1987; Sabatier & Pelkey 1990; Sabatier & Brasher 1993), and agricultural water policy (Lubell, 2007) The 1993 energy study (Jenkins-Smith & St. Clair 1993) holds particular significance in the ACF literature due to its originating use of cluster analyses based on scales developed out of the content analysis of public meeting testimony. Prior to this study, research using ACf had taken a more qualitative approach with narratin descriptions of the changing advocacy coalition beliefs. The authors of the energy study note that the\' \vere able to determine the timing of the emergence of coalitions within the system, a phenomenon they had pre,iously been unable to identify in the subsystem, 74


due to the methods employed. The t\larine Life Protection Act study by Weible, ct al. (2004) is also worth noting, as it focused on two types of processes: top down with elite and technocratic participation and bottom up with a broader collection of stakeholders. This local level study, set in two California communities, found that the preferences of stakeholders for one approach over the other was tied to secondary beliefs, as defined by ACF. While this study did not explore the result of broader participation, it at least opened up the possibility of the advocacy coalitions being recepti,e to participation from non-elites. Other A.CF scholars haYe addressed emironmental regulation subsystems including the Endangered Species Act (Ellison 1998), public lands policy (Da,is & Da,is 1988), Canadian forest policy (Lertzman, Rayner, & Wilson, 1996), and L'.S. National Forest timber harnsts (Burnett & DaYis 2002). The forest timber hanest studr included an exploration of coalition beliefs on policy inclusiYeness (being responsiYe to experts nrsus responsi,e to the public). They found two coalitions changed in this regard over time, but the authors were unable to determine the cause of the change. They treated support for public opinion shaping policy as part of the core beliefs of the coalition (Burnett & DaYis 2002), thus suggesting that even core beliefs are amenable to change given enough time. Similar to the Life Protection .Act study noted aboYe, though they took an interest in public imolnment in the subsystem, it was not to explore the role of the public, but rather the willingness of elites to include the public. 75


From the explorations of prenous authors, .\CF has shown itself to be a worthwhile model for exploring a policy subsystem, with an important contribution in its focus on beliefs and the shared action of coalitions created based on those beliefs. HoweYer, ACF suffers from the weakness in assuming that coalitions always haYe those two things in common: shared beliefs and shared actions. It additionally fails to take into account the role of individuals within these coalitions, and the potential of indi,iduals with common interests to group together based on shared actions, but not belief, or fail to act together despite shared belief (Kim & Roh, 2008). As we mme from the macro level of policymaking, where coalitions of actors are the subject of examination, to the micro level of policymaking, such as the water roundtables, it is insufficient to depend on a theory that makes assumptions about collective action with little exploration of indiYiduals \Vi thin the coalitions. Kim and Roh (2008) seek to address this limitation through a model that allows for differing ievels of shared beliefs and active coordination, without rejecting coalitions that fail to have both. Based on a re,ie\v of the policy theory literature, they deYelop their own typologies of both the indiYidual members of coalitions and of the coalitions themselns. They describe three types of coalition members, based on the extent to which they share policy interests and choose to engage in collective action: substantial collaborators han both; potential collaborators share only policy interests, but no action; and reciprocal collaborators are engaged in action without sharing the same policy interests. This latter category is of interest in the context of a participatory process, as it suggests participants 76


representing different adYocacy coalitions han the potential to engage in sharcJ action during the participatory process. For those coalition members engaged in shared action with low or medium le,els of shared policy interests, the potential for cooptation and conflict are recognized in the Csing the concepts of trust and resources as t\vo potential dri,ers of coordination, they hypothesize a typology of coalitions that haYe a mixture of leYcls of coordination and shared beliefs (Table 3. 1). Of interest is how limited the circumstance is where coalitions 'vill engage in high cooperation: only when trust is high and the leYcl of resources as \veil. Table 3.1. Kim and Roh's (2008) hypothesized effects of resources and trust on the strength of cooperatiYe relationships between policy actors LeYel of Trust High l\Iedium J,ow High High i\Iedium Low Cooperation Cooperation Cooperation Lenlof l\Iedium Medium 1\ledium Low Resources Cooperation Cooperation Cooperation Low Low Low J,ow Cooperation Cooperation Cooperation Kim and Roh's (2008) model builds from the strength of A.Cr, the focus on how beliefs bring coalitions together. They propose that collcctiYe action can occur without shared beliefs, but 'vhen this only occurs when there is an established relationship benveen the policy actors, leading to sufficient trust in the likelihood of reciprocity. The combination of coalitions built from beliefs, typologies of shared action that take into 77


account levels of shared belief, the role of trust and resources in generating action as a coalition all contribute to a broader understanding of the role that indi,,iduals might play in a policymaking process, including a participatory process. 5 bared and Public Pm1itipatioll Returning to public participation processes and the emphasis on the decision forum, if the learning from policy theory is to be applied to public participation processes like the Colorado Water Roundtables, it becomes important to explore the extent to which beliefs arc drivers of interaction in the participatory process, and thus drivers of deliberation (for which jaime.rJ and t"ompetma are criteria for success). If you examine a public participation process flrst through the beliefs that are at the table, and how they connect back to the broader community (rather than demographics at the table and how they reflect broader demographics), what can you learn about how the process can and will function? Is a fair and competent process possible when the combination of common and divergent beliefs results in coalitions? Will a fair process lead to more or fewer coalitions? Are coalitions belief focused alone or is there \'alue in looking at coordination activities combined with beliefs? Starting from the concept of beliefs, and then looking at process and interaction, we can in\"Cstigate different types of coalitions and their individual actors. As noted abon, Kim and Roh's (2008) approach makes this connection, referring to coalitions by the types of collaborators within them: potential, reciprocal, and substantial collaborators. Interaction at the indi,,idual level goes beyond where policy theory has 78


focused because the indiYidual actor is more than just a policy cog in their models. Theories like collectiH action suggest greater interaction when it directly benefits the individual participants. Kim and Roh (2008) extend this to "The relationship of policy actor A with policy actor B can be predicted with the following critical factors the proximity of policy interests benveen policy actor A and B, policy actor B's resources, and trust of policy actor A toward policy actor B" (p. 682), leading to either cooperation, cooptation, or conflict. Cooperation occurs when policy beliefs are very similar, while cooptation occurs when policy beliefs are only some\vhat similar, and actors may attempt to influence each other to better align with their own views. Conflict occurs when there is a lack of shared policy interests. These interpersonal interactions will han an impact on the deliberative process, and the engagement of the public the participation. The Research Propositions The discussion of public participation revealed a weakness 111 the definition of representation and failure to account for interests or provide any meaningful guidance on how to include interests. The Advocacy Coalition Framework's approach to coalitions overcame some of this shortcoming, particularly the expansion of the concept of interests to the concept of shared beliefs, but assumes that sharing beliefs logically leads to sharing actions in a policy arena. Bridging ACF with Kim and Roh's (2008) model of individual interaction within coalitions introduced the concept of coalitions of shared beliefs that nry in their levels of shared action. Now, we have a model of a public participation process that is composed of (1) representatives appointed to a 79


participation process whose (2) beliefs align to create (3) coalitions with others both at the table and outside the table, and whose (4) leYel of shared actions within those coalitions may ,ary greatly, in part based on trust. In essence, we are suggesting that public participation literature has excelled at understanding processes for participation and how indiYiduals should have opportunities to interact and make decisions within these processes, but has stopped short of understanding the role of interests and representation in a participatory process in part because it cannot define what it means to be representatiYe in any comprehensiYe, useful way. In its distaste for interest-based decision-making, where authors go as far as to state that "interest-based politics symbolizes much that deliberati,e procedures seek to onrcome," (Hendricks, 2006, p. 571), this line of reasoning fails to incorporate the literature in public policy that expands the understanding of interests. Where public participation fails, policy literature can add in its understanding of interests and coalitions, and the introduction of shared beliefs and shared actions; yet, ACF largely ignores the importance of process, including such things as deliberation, trust, and interactions, and with it, the potential of nriability in shared actions. If we are to understand paructpatory policymaking processes, and more importantly, understand how to improYe them, these concepts need to be explored together. Arnstein (1969) introduced the importance of power and disempowerment; Beierle (1999) brought the concepts of opportunities to participate and representation into the criteria of a successful process; Sabatier and Jenkins-Smith (1993) emphasized 80

PAGE 100

the importance of shared beliefs; and, most recently, Kim and Roh (2008) explored the role of the indiYidual and the decision to act in concert with others in the policy coalition. NO\v, \Ve need to pull together each of these important concepts to form an understanding of hO\v coalitions of actors interact in deliberatiYe and participatory processes, both as indhiduals and as coalitions. To do so, belO\v are three propositions which form the basis of a case study. They are propositions which require a depth of exploration that only a case study can prmide (Flpbjerg, 2001). Rather than asking comparative questions, they encourage imestigation of the concepts in hopes of understanding the connections between these bodies of theory and their arguments for how policy decisions can be made, should be made, and are actually made. PropoJition 1 Participatory processes are composed of indiYiduals who represent policy interests, also known as adYocacy coalitions, and these coalitions are defmed by shared beliefs rather than demographic characteristics. As previously discussed, representation is an important feature of a participatory process, and can influence the extent to which the process is fair and t'o111pt'1enl. The Advocacy Coalition framework argues that policy decisions are the result of the shared actions of coalitions, \vith the coalitions composed of indiYiduals \Vith shared beliefs (Saba tier & Jenkins-Smith, 1993). If a participatory process like the water roundtables is composed of representatiYes of interests, they must also be representatiYes of beliefs that exist within and outside of the roundtable. Proposition 1 seeks to establish that representation of interests depends on more than demographics such as age, income, 81

PAGE 101

and genderand uses the public policy lens and its interest in beliefs as the focus. To study this proposition, it will be necessary to explore the beliefs and demographics of participants in a participatory process, such as the water roundtables, but also participants outside the participatory process, or one cannot argue the beliefs at the table han any relationship to coalitions in the policy subsystem. Propoxition 2 The coalitions represented in a policy process will engage in shared actions, as seen by their shared information and ttust in each other. Proposition 2 focuses on the internal functioning of the coalitions irwolnd in the participatory process. If research can identify the distinct interests represented in a participatory process like the water roundtables, then the next step is to understand how the interests function at the roundtable. Kim and Roh argue that it is a mistake to assume that coalitions based on shared beliefs \Vill also undertake shared action. If we follow .\CF's focus on shared beliefs equating to shared actions, we could expect to find multiple coalitions on the water roundtable, each of which is undertaking shared actions. If we use Kim and Roh's focus on differently functioning coalitions, we can expect to find multiple coalitions composed of indiYiduals with shared beliefs or shared action, but not necessarily both. Based on Rockloff and 1\foore (2006) and Innes & Booher (2004), action might include something as straight-forward as information sharing, an important role of a representatiYe in a participatory process, or as complex as influencing one another. It may also include Yarying leYels of trust, arguably a necessary component of shared action as researchers interested in public participation in natural resources have 82

PAGE 102

identified "a clear relationship between trust and successful participatory processes" (Parkins & Mitchell, 2005, p. 535). PropoJition 3 In a participatory process, "representation of interests" may not equate to "representing interests" if the stakeholders are not equally engaged in the process. One of the assumptions made m the public paructpation literature is that participation 1s equal to representation. Even if the participants of a participatory process are selected to be equally representatiYc of the beliefs in the broader public or among vested interests, is it accurate to assume all interests are equally represente& Susskind et al. (1999) explains that excluded stakeholders pose a threat to a participatory decision-making process, with the potential to undermine the decisions it makes. The assumption being made is that exclusion is non-,-oluntary and refers to being physically outside the decision-making process. Howenr, Kim and Roh (2008) describe a policymaking environment where coalitions 'vho share beliefs do not necessarily share acuons. As a fair and competent process assumes equal participation, the lc,-cl of engagement of the different interests will contribute to the extent to which the process is fair and competent. If coalitions participating in the process han nrying lenls of shared beliefs, shared actions, and trust, might they also haYing ,-arying b-els of shared engagement in the participatory process? Could their Yarying levels of trust in each other contribute to varying lenls of engagement? If they are not fully engaged, is the process still fair and competent? 83

PAGE 103

Conclusion As public participation theory has developed and empirical studies have begun to amass e\idence to support the nlue of a delibcrati,e process, the definition of adequate representation in this deliberation has been under debate. On the side of purposeful samples drawn from demographic information, Rowe and Frewer (2000) have ignored the importance of interests and beliefs. In contrast, Leach (2006) has sought to go beyond mere demographics (though he does still value them as a selection factor) to also considering varying perspectiYes. l\Iodng another step away from demographics are the findings of lrvan and Stansbury (2004) who identify representatives as those with influence and willingness to serve as representati,es of interests. What public participation does not yet offer is an understanding of how to identify the spectrum of interests that must be represented in a gi,en participatory process. Where public policy theory contributes is the understanding of beliefs and the role they play in influencing policy. ACF establishes the importance of the coalitions of actors driven by beliefs, and the .\CF scholars haYe explored methods for understanding those coalitions and their beliefs. Lacking from ACF is the individual focus, something that cannot be ignored in a participatory process like the roundtables. With Kim and Roh's (2008) expansion on ,\CF to consider the indi,idual level actors and the potential of coalitions with shared beliefs, but no shared action, or alternatively, shared action, but no shared beliefs, the .\CF literature has a bridge to the public participation literature. It is on this bridge that 84

PAGE 104

the three propositions aboYe belong, exploring concurrently the concepts of policy beliefs, coalitions, participatory processes, and shared actions. 85

PAGE 105

CHAPTER4 METHODS This chapter lays out the research design, including the participatory action research approach used to design and execute the research. The propositions m the prenous chapter are expanded upon to include the data needed to answer them, the creation of the surny instrument, collection of data, and the study population. A careful description of the analytic techniques used in the research is followed by a discussion of the limitations of the studr. Research Approach: Participatory .\ction Research The design of the study is not only the methods it uses, but the overarching strategies that mon it forward. From the beginning, the research was driven by the self expressed needs of the water community, as heard at meetings and in one-on-one discussions with key stakeholders. A single roundtable, the Colorado Basin Roundtable, was selected as the setting for the study due to its size and position within the state. The Colorado Basin Roundtable is particularly important, as the majority of transbasin diYersions and the most significant compact requirements all come from the Colorado Rinr. After nearly a year of building relationships and additional discussion of the 86

PAGE 106

study \Vith key water policy stakeholders, the direction of the studY was set and the political support for the study was secured. A participatory action research (PAR) approach was used to ensure the study accurately reflected the policy context of the \\"ater roundtables and resulted in a report that could meaningfully inform the roundtable process. P;\R is a term that has been used broadly and includes many different participatory, empowerment, and collaboratiYe research approaches (O'Brien, 2001), including Community Based Participatory Research and Action Research. The basic practice of PAR is: to fully engage the stakeholders in all stages of the research, leading to knowledge that has the potential to meaningfully affect the community (O'Brien, 2001; O'Fallon & Dearry, 2002). The approach guided the concrete steps taken to design and implement the study. The understanding of action research contributed to ensuring each step of the research remained focused on the ultimate goal: to generate knowledge that has sufficient credibility, accuracy, and value to the stakeholders that it can lead to meaningful change. Additionally, the action research approach emphasizes the iterati,-e nature of research, where each cycle of data collection and analysis informs the next (O'Brien, 2001 ). As the study began with the agreement among the stakeholders and funder that if the findings praYed useful, further iterations of the study should be completed with other roundtables, the participatory action research focus on using knowledge gained in one cycle to improYe the next \Vas an appropriate and necessary approach. 87

PAGE 107

PAR is an approach that engage the subjects of the research in the design and implementation of the research. It is based on the assumption that research and findings are more effectin when they bridge the gap between knowledge producers and the community being studied, recognizing that community as a unique social entity with a shared sense of identity and fate (O'Brien, 2001; O'Fallon & Dearry, 2002). PAR attempts to do more than understand: it seeks to build the capacity of the community to address problems identified during the research process. Action research goes a step beyond building capacity to actively seeking to solve the problem. PAR assumes that the unique strengths of the community will enhance the understanding of those problems, allO\ving for the research knowledge to be more accurate and integrated into improving the well-being of the community. Specifically, PAR has a core set of principles to guide research actiYities that promote collaboration, co-learning, community-driYen projects, and dissemination of results in useful ways (O'Fallon & Dearry, 2002). A study designed using PAR first and foremost recognizes the centrality of the social, political, cultural, and econom..ic systems of the community. It engages community members in deYeloping not only the research tools, but the research questions. It will also engage the community in interpreting results. l'ltimately, such methods are collaborative processes that establish trust and create opportunities for knowledge to be useful in impro\'ing community problems and addressing community-identified needs (Cornwall & Jewkes, 1995; \'iswanathan, .\mmerman, Eng, Garttehner, Lohr, & Griffith, 2004). To implement these types of research, the researcher shares or even releases the traditional 88

PAGE 108

"control" of the study to the participants, recognizing the \alue of their non-academic sources of knowledge and increasing the potential of the research to lead to meaningful change (Marti & Villasante, 2009). The sections that follow describe the four types of data (belief data, demographics, network data, and success data), as well as the participatory process of developing the research questions and suney instruments, collecting the data, and interpreting the results. Although interpretation of results is not usually included in a methodology chapter, in a participatory research study, it is important to explicitly plan for and report on how stakeholders participated in the interpretation process. Identifying Data Needed The research propositions are grounded in academic literature and built from gaps in the understanding of participatory processes. This dissertation also sought to answer the question posed by participants at the roundtables and leading stakeholders in the water community: The goal of the roundtable process is to make decisions within and across basins, decisions that haYe sufficient support to be implemented by water rights holders, not just policymakers; giYen this, we further ask how can the roundtable process leYerage the complex interpersonal and interest-based stakeholder interactions to moYe fonvard with consensus-based decision-making? To answer questions generated by both the roundtable and academic demands, the design of the study brought together two merarching strategies: using academic literature to identify data needs and de\clop draft measures for collecting the data; and 89

PAGE 109

using extensi,e piloting, appropriate within a participatory research approach, to create the survey instrument and thus the data collected. Thus, this section includes a detailed exploration of the types of data included on this study and the description of the piloting process as it heavily influenced the specific data collected. l\lany types of data are needed to investigate the propositions posed in Chapter .'>. One type needed for all three propositions is the belief data, which are designed to establish the structure of the belief coalitions at the roundtable and in the water policy community connected to the roundtable. The belief data sene as the independent nriablc in all three propositions, as it is assumed to capture policy core beliefs that are not easily changed. Proposition 1 states, "Pmticipa!OIJ' are t"omposed q/ indiT,idlla/.f who repre.renl poliq inlere.rl.r, k11own a.r adT-'Ot
PAGE 110

the following types of data are needed: (1) data to describe the belief coalitions; and (2) net\vork data on the interactions at the roundtable among the participants and their relationships within and bet\veen belief coalitions. The net\vork data were the dependent yariables in this analysis. This net\vork data allow for an examination of the interactions that indi,iduals han with other indiYiduals at the roundtable, and the interactions that coalitions of indiYiduals han with other coalitions of indiYiduals. Proposition 3 states, "!11 a pmtiapatoo' proce..r. "repre.renta/ion q/ in/t!J't!.f!J" lllt!J' not eq11a/e to "!t!preJenling intemlJ" if tbe Jtakeboldn:f are no/ equal(y mgaged in the proce.rJ. To address proposition 3, the following types of data are needed: (1) data to describe the belief coalitions; and (2) data on the participants' perceptions of the participatory process. The percepuon data were the dependent nriables in addressing the third proposition. Data /o Under.rland Beliejj In 2006, a study at Colorado State Uni,ersity of nlues and beliefs in Colorado's water policy community employed q-methodology and k-means cluster analysis to deYelop clusters of beliefs. Q-methodology is a research approach used to study people's subjectiYe views, and allows for prioritization of beliefs. The results of the study show the Yalue of selecting analysis techniques that allow for multiple groupings of indiYiduals based on their beliefs (Colorado Institute of Public Policy [CIPP], 2006). Although preYious water policy studies employing net\vork analysis and looking at beliefs had found a dichotomous scale to be sufficient (\Veible & Sabatier, 2005), the ClPP 91

PAGE 111

(2006) study made it clear that the issues are far too complex in this setting to be boiled down to a binary analysis. This aligns with Sabatier & Weible's (2007) assertion that while it is possible to identify at least two coalitions based on only measuring two or three specific policy core beliefs, when more beliefs are measured, third coalitions may be found \Vho disagree with both of the original two. The CIPP (2006) study outlined how interests in the water community can both ha,e conflicting and shared beliefs with other interests, highlighting the complexity of the policy issue. This study sought to capture this complexity, though through fewer measures than a full q-methodology array. Statements that helped to define one or more of the CIPP clusters, but not all of the clusters, were pulled out of the two q-sorts in the aboYe mentioned study. Additionally, the waler S11pp(y and 1\'eed.r Repottjor tbe Colorado Basin (CDl\1, 2006) included a survey of basin roundtable member's indiYidual preferences. Preferences prioritizing municipal, agricultural, emironmental, recreational, and economic issues all had high variability among participants, suggesting these are areas where differences in stakeholder preferences could provide meaningful distinctions bet\veen factions. An additional statement on cooperation among basins was added to the survey, as this is the purpose of the water roundtables, as described in their authorizing legislation. As the focus of the study is on the roundtable process, all of the belief statements were written as "priorities of the roundtable process," rather than being priorities in \Vater policy more generally. 92

PAGE 112

To imestigate proposition one, belief data are needed across all of the Colorado Basin Roundtable (CBRT) participants and the stakeholders more broadly inYolYcd in the water community, identified by roundtable members. With onlY 50 members of the CBRT, there were not enough respondents from the roundtable alone to deYclop clusters of beliefs and feel confident they represented the interests at the water table. By engaging oYer 100 additional peripheral stakeholders, identified by the CBRT surYcy participants, the belief clusters can represent the water community more broadly, \Vith the roundtable members acting as representatiYes of those interests. Data to U ndmtand Demogruphit".J' In addition to traditional demographics such as age, gender, and socio-economic status, the demographics include many questions specific to water policy and the roles of individuals within the policy arena. The water specific demographics include such variables as the statutorily mandated role of the roundtable member, the interest/ affiliation with which he/she self-identifies, the length of time in the \Vater community and the ownership of water rights. The demographic question specific to affiliations is similar to the questions in both Zafonte and Sabatier (1998) and Weible and Sabatier's (2005) studies of policy networks. The two studies range in their number of affiliations, from 34 to only 14, and both are measuring a similar size network of policy stakeholders. In this study, the total number of affiliations began at less than 10defined by the legislation mandating the membership of the water roundtables-and ended after the piloting with 31 affiliations plus the option of writing in an "other" affiliation. To allow 93

PAGE 113

more flexibility in the analysis, the 31 affiliations arc grouped into seven categories, with "other" options in each relevant category, allowing even answers that are "other" to be included in an analysis that groups the affiliations. The fuller range of demographic questions allow for a complex understanding of affiliations and role in the water community, using variables identified in two ways: through informal discussions with \Vater stakeholders and drawn from a previous study of water policy stakeholders in Colorado (CIPP, 2006). j'\'etwork Data to U nderJtand the Relation.rhipJ at the Roundtable To investigate proposition two, it is necessary to collect network data from the Colorado Basin Roundtable participants. Net\vork data are one of the principal ways of analyzing stakeholder relationships, particularly as they relate to communication, trust, and influence (Reed, 2008). It has been used in many different policy and inter organizational settings, including water issues (\'Veible & Sabatier, 2005). Among the outcomes of net\vork studies, this type of research can be used to improve levels of collaboration by identifying peripheral stakeholders whose untapped interests and expertise could be drawn into the process (Cross & Parker, 2004; i\Iilward & Provan, 1998) and help understand where and why alliances form (\'Veible & Sabatier, 2005). Network data arc a particularly appropriate for this dissertation. As the method has developed, so too has the body of theories that incorporate a network component (i\Iongc & Contractor, 2003). Increasingly, policy research has used network data, allowing for the understanding of relationships between policy actors including the 94

PAGE 114

patterns and structures of their interactions (Wasserman & Faust, 1994). Studies in the early 1990s demonstrated different types of policy networks, the central roles of key actors such as entrepreneurs, and fragmentation of policy subsystems (Dowding, 1995). In addition to studying the different relational ties benveen actors, net\vork analysis can also include specific attribute information about each actor, capturing the "'the size, shape, and flaYor' of the actors constituting the net\vork" (\X'asserman & Faust, 1994, p. 39). In a study that focuses on indiYidual representatins of different interests, the attribute information is particularly important. To gather this attribute information, network researchers have used a 'Tariety of different methods. For example, in studies of public sector interorganizational and intersectoral networks, it is common to see qualitatiYe inteniew data (Provan & l\Iilward, 1995; FosterFishman, et al., 2001 ), organizational demographics (Vogenbeck, 2005), and suney questions on policy beliefs and affiliations (Schneider, Scholz, Lubell, :\lindruta, & Edwardsen, 2003; Weible & Sabatier, 2005). Important information on nlues and beliefs can also be collected and analyzed as a compliment to the net\vork measures in a study. For example, some studies haYe captured core policy beliefs by incorporating a series of Likcrt-scale questions into the surny (Weible & Sabatier, 2005; Zafonte & Sabatier, 1998). In this study, the approach of Weible and Sabatier (2005) is followed, but instead of creating a dichotomous measure of water policy beliefs from the Likert-scale questions, a cluster analysis approach allows for a more complex result that captures the Yariation in belief. Nt'lwork Data 95

PAGE 115

Network data and analysis is an emerging area of methodology; it continues to undergo significant flux with new de,elopments eYery year. The description of network data and analysis below is drawn from T\longe & Contractor (2003) and Wasserman & Faust (1994), and the examples are drawn from this thesis. In application, network analysis approaches the study of people in a fundamentally different way than other statistical techniques. The standard quantitatiYe method focuses on the indiYidual, aggregates the indiYidual answers of all members of a sample, and infers the findings to a population. In contrast, social network analysis focuses on the interactions between indiYiduals (or groups), uses their to help understand those relations, and aggregates the relations to study the full network of relations. As random sampling is often not used, the statistical emphasis on inferring the results to a larger population is less of a focus. Instead, researchers are interested in the driYers of network structure, the structures themsehcs, and the consequences of the network structures. In this dissertation, the driYers of network structure are presumed to be the beliefs, while the interactions \Vithin that structure are studied through measures of information exchange, trust, and perceiYed influence. Network analysis labels each unit in the network as a node. Thus, an organization, country, or indi,idual, as we use in this study, will all appear as a node in the network. Each node maY haYe attributes that can be included in the analysis, such as age, affiliation, language, economic power, policy beliefs, or any other characteristic of interest to the researcher. The connections between individuals are the relalion.f and 96

PAGE 116

many relations can exist between any two nodes. Relations can be actions, such as loans between organizations, marriage ties between families, or information exchange between individuals. Relations can also measure more intangible things such as trust or friendship between individuals & Contractor, 2003; Wasserman & faust, 1994). Regardless of the measure, the relations haYc many properties. They can han dirrdion, e.g., when one individual trusts the other, or be non-directional, e.g., shared funding of a program by two organizations. \X!hen a directional relation between t\vo nodes goes both ways, e.g., node A trusts node B and node B trusts node .\, it is a reciprocal relationship. In this study, \Ve are particularly interested in aggregating the directional relationships go from one cluster of beliefs (the interest group) tmvard other clusters (interests). Relations can also ha,e .f!rength, e.g., the frequency of information exchange bet\veen individuals. When multiple network relations are measured, e.g., trust and information exchange, they can be analyzed in isolation or in relation to each other, resulting in a uniplex or multiplex analysis, respectively (Monge & Contractor, 2003; Wasserman & Faust, 1994). The analysis of relations can occur at many levels: the individual node, the tics between t\vo or three nodes, subgroups within the net\vork, or the whole net\vork. In this study, the focus is on the directional relationship bet\veen t\vo nodes (often referred to as a dyadic level of analysis). Just as frequencies, ANO\' AS, regressions, and factor analysis all allow researchers to answer different types of research questions and test different theories, so too do each of the Ienis and types of analysis ha,e their unique 97

PAGE 117

roles for a researcher utilizing social network analysis (Monge & Contractor, 2003; Wasserman & Faust, 1994). In this dissertation, the relations were measured between iudit,idtta!J, and the relations include actions like information exchange and more intangible concepts such as trust and influence. The relations were measured diredionai!J' between the individuals and the .rlrenglb of the tie was measured. The multiple network relations were analyzed at the dyadic lenl together, resulting in a m11lliple.:..: analysis. The relations were also aggregated within and between each of the clusters (interests) so as to compare relations across clusters, not just across indi,,iduals. I\'e/tvork AleaJ!II"N The measures for the surYeys include si..'i: network measures coYering information exchange, information importance, beliefs, two dimensions of trust, and influence (see Appendix A for the suryey instrument). The data allow for an exploration of the different tics that might exist among actors and interests in a participatory process: information sharing, influence, and trust. The final network questions asked respondents to indicate \Vho in the water community (and at the roundtable, if they were CBRT participants), they: (1) consider the most influential statewide; (2) exchange information with; (3) recei,,e important information from; (4) belie,e share their goals; (5) trust to keep their interests in mind, regardless of whether they share their goals; and (()) can depend on to follow through on a commitment. 98

PAGE 118

To develop each of these measures, a three-step process was used. First, measures were identified in the literarure, drawn wheneYer possible from the studies of well-known authors respected for their work in the net\vork field, e.g., the nct\\ork measures used by Weible and Sabatier (2005) in a study of water policy nct\vorks. Second, the measures were adapted to fit the water policy setting of the roundtables. Third, the measures were extenshely pilot tested with actors in the water policy community whose roles, years of experience, and knowledge of both water and research nry greatly. The revisions during the pilot process arc described more extensi,-cly later in this chapter. One revision from the piloting process that affected all of the net\vork questions was the choice to design measures that haYe ,alues indicating such things as the strength or frequency of the tie. An issue to consider when de,eloping the net\vork questions in these areas is whether the question can be better answered by a dichotomous Yariable the tie is present or it is absent -or by a question that measures both the existence of the tie and its value (Wasserman & Faust, 1994). Tie nluc can be measured by including a scale of possible answers for each tie that extsts. For example, a question on the exchange of information may ask whether information is exchanged or not. But, to caprure the Yalue of the relationships, it would instead ask whether information is exchanged on a daily basis, weekly, monthly, yearly, or neYer. Capturing the \alue of the relationship is a way of understanding its "depth" (Scott, 1991 ). Additionally, nct\vork questions that use these types of scales haYe been found to have greater reliability than 99

PAGE 119

fixed choice designs (Wasserman & f-aust, 1994). It became nry clear early in the piloting process that haYing the ability to differentiate between different strengths of relations made the survey feel more accurate and comfortable for the water stakeholders. The scales for each question \vere piloted and revised until respondents consistently indicated the scale felt useful and accurate to them. The influence measure explicitly asks the sutTey respondent to assess the influence of their peers in water policy and management issues statewide using a scale of nry influential to not ,ery influential. This may seem to undermine the goal of measuring influence at the water roundtable. However, it has advantages over asking for the influence of water roundtable members at the table only. First, the roundtables were designed to move toward interbasin dialogues. While influence at the table is important initially, influence between tables may be more important as the process mo\es forward. Second, the information exchange measure can be analyzed to look at the prestige of the actors to capture the concept of influence within the roundtable, making it less necessary for the influence question to specifically focus on the roundtable or Basin. The Colorado Basin Roundtable respondents \vere asked to answer the extent to which they agreed with each question for all other members of the roundtable (Figure 4.1) and to identify one or more stakeholders outside their roundtable for each question. External respondents, those who were invited to take the suney because they were identified on one of the six questions by a Colorado Basin Roundtable member, were only asked to identify one or more stakeholders for each question. They were not given 100

PAGE 120

a roster of stakeholders, as had been done within the Colorado Basin Roundtable. The external respondents identified by the Colorado Basin Roundtable members were included in the belief analysis to establish the broader network of water policy beliefs. 2A. On a scale of 1 -.5 (dUly to nevet), please ind1cate how often in !be last yeu you h4ve infomution alated to watet policy and issues with othet membets of the toundtable. f Rowtduble 0 e f Rowtduble 0 e Members f ::t .fi "' Gl ::: 0 Gl Members f ::t .fi Gl ... 0 ... "' Gl Gl Ci :!!! .....l Ci :!!! Gl .....l 1 RTmembet 1 1 2. 4 5 11 RTmembez: 11 1 2. 3 4 2. RTmembet2. 1 2. 4 .5 12. RT membet: 12. 1 2. 3 4 3 RTmembet3 1 2. 3 4 5 13 RTmembet 13 1 2. 3 4 4 RTmembet4 1 2. 3 4 5 14 RTmembez:14 1 2. 3 4 5 RTmembet:5 1 2. 3 4 5 15 RTmembet 15 1 2. 3 4 6 RTmembet6 1 2. 3 4 5 liS RTmember liS 1 2. 3 4 7 RTmembet7 1 2. 4 5 17 RTmembez: 17 1 2. 3 4 0 "Co"T"'' ......... 1 ..... 0 ., 'I t o 'D-'r' .......... 1 .... 0 ., 'I Figure 4.1. Example roster for answering the information exchange network question, with names removed. Data to Under.rland the Pemptioll.f qltbe Ro1111dtab/e Prot"e.l.t" 5 5 5 .5 5 5 5 t At the time of the surny, the water roundtable process was still a rclatiYely new undertaking for Colorado. l\Iany different acti,ities had begun over the course of the first 18 months of the roundtable process, including dc\eloping a grants process and awarding grants, setting priorities within roundtables, preparing for the needs assessment, educating roundtable members, and dialogue both within and between tables. Based on input from roundtable chairs and the Department of Natural 101

PAGE 121

Resources, a set of acti,ities underway in the Colorado Basin Roundtable and the statewide water roundtable process were explored as part of the study. The extent to which respondents report the roundtables were successful at each activity, as well as successful at meeting their own goals for the process, helps us to understand hmv the process is ,ie\ved at a particular point in time. While the activities cover many of the required and desired outcomes from the roundtable, they do not cover the merarching goals of the roundtables, including developing "locally driven, collaborative solutions to water supply challenges" (37 -75-1 04(1 )a, C.R.S.). In summary, the data collected for the dissertation included the following: data on beliefs and demographics (collected and analyzed across all participants); data on network relationships (collected and analyzed within the Colorado Basin Roundtable; data collected to identify external stakeholders for participation in the belief analysis); and data on success (collected and analyzed within the Colorado Basin Roundtable). Piloting tbe Sumf!J' The data \vere collected through the usc of two suney instruments: (1) a suney specifically designed for the Colorado Basin Roundtable members (CBRT); and (2) a second survey designed for the external network identified by the CBRT members. Onh the roundtable member smTe\' included the full list of CBRT members for each -network question, as the network questions \Vere needed for the analysis at the roundtable level, while the belief questions \vere needed for the analysis of all stakeholders. 102

PAGE 122

Denlopment of the survey tools was done in partnership with leaders in the water policy community. The survey was designed and then piloted in three phases with multiple revisions during and after the pilot phases. Piloting addressed both the appropriateness of each question on the survey as well as the overall purpose, content, and structure of the survey. Additions were made to meet the needs identified by \Vater policy leaders, and some questions were to address concerns about participant discomfort when answering the questions. of the initial suney questions \vere drawn from previous studies in other states on similar issues (Doney & Cannon, 1997; Cross & Parker, 2004; Weible & Sabatier, 2005; Dietz & Hartog, 2006) and previous studies in Colorado on the roundtable process (CDi\1, 2004; CIPP, 2006). Other suney questions were developed in partnership with the Public Education, Participation, and Outreach W'ork Group of the lnterbasin Compact Committee (IBCC) and other key stakeholders in the roundtable process. The suney was piloted \Vith I\ water stakeholders, representing participants and non-participants on roundtables and the IBCC, representing a range of interests and perspectives. The questions were substantially modified after discussion with the pilot partlclpants. The final sutYey design reflects the input of the Public Education, Participation, and Outreach \Vork Group and pilot participants, and questions dra\vn from research cited in this dissertation. The first piloting round was conducted with three other researchers to ensure the suney format and questions were easily accessible for the external pilot and had no 103

PAGE 123

ob,ious problems. The second round of piloting was with water policy stakeholders representing the funding body for the study. They were particularly Yaluable for this first external pilot due to their dinrsity. The participants included members of roundtables, rcpresentati,es on the IBCC, a water researcher, and a facilitator imolved with multiple water roundtables. .\ custom suney was created for each pilot participant with names relennt to the piloter (e.g., participants from the roundtable they attend) in the network questions that utilized a roster. In response to feedback, most of the network questions were adjusted, the influence network question was added, and the belief questions were completely redone from only quantitative to a mix of qualitatiYe and quantitatiYe. A new question on participation in subcommittees was added to give participants a place to describe their imoh-ement in the roundtable process in a less quantitative, structured manner, as was requested participants. Many of these changes resulted in questions that were part of the analysis for the report (explained in the Preface), but were not used in the analYsis for this thesis. The third round of piloting sened to finalize the sun'ey as well as establishing its credibility \Vith key water stakeholders. Chairs, ,-ice-chairs, and IBCC representatiYes from roundtables throughout the state were provided with a sune) that listed 10 randomly selected members from their own roundtable. The six participants in this phase of the piloting had little to no prior exposure to the study. The pilot test began with a ten minute introduction and instructions by phone, followed by the participant taking the sutTey, and ending with a second phone call of approximately 20 minutes 104

PAGE 124

immediately after the sur..-ey was completed to debrief the suney questions. .\ custom suney was created for each pilot participant with names rele,ant to the piloter (e.g., participants from the roundtable they attend) in the network questions that utilized a roster. The feedback ensured statewide input and a more diYerse perspcctiYc as participants included both long-term water policy leaders and stakeholders newer to the water arena. The belief questions did not change as a result of this phase of piloting. The wording of the net\vork questions were changed in a few cases, but the content largely remained the same. The demographics were changed more substantially, particularly in the area of water rights, as the new pilot testers had \'ery different types of rights than preYious testers, reYealing problems with the measure. 0Yerall, the surycy questions were consistently interpreted and easily understood by the third round of pilot testers. The original measures included in the study were drawn from highly releYant studies by respected authors, as discussed earlier. .\s the 11 representatiYes of the many different parts of the Colorado water community piloted the suney and participated in a detailed debrief of their experience, the suney questions undenvent substantial re,ision, as laid out in Table 4.1. Though tested and Yalidated measures from other settings was the original sunrey design approach, the pilot participants' insight into the Colorado water policy setting changed this significantly. The context of the water policy arena was seen as too nuanced to be understood through measures drawn from elsewhere. Also, measures such as the influence, trust, and information sharing proYed to be too general 105

PAGE 125

to capture the expenence of the \Vater stakeholders who joined in the pilot. .-\s the measures have been increasingly refmed, the interpretations by pilot testers have become ,ery consistent, suggesting the ne'v language is accomplishing its purpose. Table 4.1. Adaptation of network measures Original Source Reasons for revision Final measure measure "vou ha,e relied Weible The first round of pilot "how often in the last on most heavih and testers felt that "information rear you have for information Saba tier exchange" was a more exchanged information or advice on (2005) accurate description of their related to water policy water policy interactions. Additionally, and management issues and information exchange with other members of measures are used often in the roundtable" 0 Issues the network analnis literature. A second measure of "please indicate the information exchange was extent to which the created in response to information you receive multiple pilot testers concern from each roundtable that the frequency of member is important, exchange of information was that is, helps you important, but did not achieve your goals for capture the full importance of water policy and the information exchange management issues." relationship. "you regard as Weible The first round of pilot "the extent to which allies on and testers interpreted this \"OU belien the important \Vater Saba tier question in many ways. After following roundtable policy and (2005) re,iewing the data needed, members' agree with or management the measure was split into disagree with your goals 0 lSSUCS two questions, one focused for the roundtable OJ). shared goals and the other process" focused on trust, the next measure listed. 106

PAGE 126

Original Source Reasons for reYision Final measure measure "the extent to Cross and The trust measure did not "you would trust the which YOU Parker originally emphasize that following roundtable would trust the (2004) beliefs could differ, e\en if members to keep your following trust was present. The interests in mind on roundtable addition of this language water policy and members to resulted in more consistent management Issues, keep your interpretations of the regardless of whether he/ she shares your mterests m quesuon. mind" interests" Table 4.1 (Continued). Adaptation of net\vork measures Original Source Reasons for reYision Final measure measure Dietz and A second trust measure was "please indicate the Hartog added during the ftrst round extent to which you are (2006) of pilots in response to confident each of the piloters' comments that they following roundtable trust the person, but would members will follow not tum to them for help on through on a a water policy and commitment" management issue. The trust measure is intended to get at trust in action, not just words. 107

PAGE 127

Original Source Reasons for revision Final measure measure "Please indicate Cross and The influence measure was "please indicate the the extent to Parker clarified to be influence in extent to which you which ,-ou (2004) Colorado water policy and consider each person consider each management issues, not just listed below to be person listed Colorado Basin issues. influential in water below to be Otherwise, no change was policy and management influential in made to the measure. issues in the STATE water policy 0Yerall, this was the network OF COLORADO; that and measure that was most is, the people who seem management consistently interpreted by the to han pull, weight, or issues; that is, piloters in its original form. clout with others on the people who these issues." seem to ha,-e pull, weight, or clout with others on these lSSUeS. Collecting the Data Once the surYeY \Vas fully designed with input from water stakeholders, two approYals were secured from Institutional ReYiew Boards (IRB), one at Colorado State L:niYersity and the other at the Uninrsity of Colorado, Denver. Two IRB approYals were necessary due to the Principal Imestigator's dual affiliations as a student at one uni,ersity and a researcher at the other, where the funding for this study was housed -\fter the two IRB approYals were secured, the data collection process was undertaken \s detailed abon, it included a combination of Colorado Basin Roundtable members and additional \Vater stakeholders. As the water stakeholders were identified by the roundtable members, it required a multi-stage data collection strategy that took OYer six months to complete. 108

PAGE 128

DeJ.t'liption q/S11bjed Pop11lation The sample of participants came from the roundtable and the water stakeholders they identified. Social network studies often chose not to use random sampling. h>r example, in Weible and Sabatier's (200S) study of ,,ater policy networks, collected names of the key stakeholders in the policy subsystem of interest through snowball sampling and sent surYeys to the stakeholders identified. This approach is appropriate as standard network studies utilize a whole network approach where all actors in the network and all relationships between actors are included in the dataset. Though ego centric network analysis is also possible, where information is collected on pairs \Vithout connection to the rest of the study population, the analysis possible \vith such a study is limited (Wasserman & Faust, 1994). If random sampling \Vere used, the study would be limited to this ego-centric approach, thereby decreasing the analytic options. A purposeful non-random sample followed by snowball sampling through the network question was used. All members of the Colorado Basin Roundtable \vere invited to participate in the first phase of the study, for a total of SO possible sur\"Cy participants. Of the SO possible participants, 46 participated. The suryey participants were contacted and suneyed between l\lay and August in 2007. The 46 \vho completed the sutTey identified an additional 191 water stakeholders, each of whom was invited to participate in the study. As the external stakeholders were identified through net\vork questions that focus on water policy and management, they were active members of the water policy commumty and 109

PAGE 129

representative of many interests in the basin and outside of it. The external stakeholders each received an imitation by telephone as well as email im-itations. Those who could not be reached or chose not to return the telephone calls did not receive an email invitation with the suney link. One hundred and eighteen of the external water stakeholders participated in the suney and 95 of those participating completed all questions in the suney (81 o). The total response rate was 68% for suneys returned by the stakeholders. In all, 164 indi,iduals 46 basin members plus 118 external stakeholders -participated in the research. Data Co//edio11 At a Colorado Basin Roundtable meeting, 26 of the 50 roundtable members completed the suney. It took each member between 45 and 60 minutes. During the subsequent discussion, the participants asked a variety of questions about the use of the information including if the study would be conducted with other roundtables and whether the study could be redone in 18 months to track progress in their networks. The members also shared some insights from taking the suney. One member said, "After taking this thing, I realized there are folks at my roundtable that I want to know." Others pointed out that many of the people on the roundtable would ha,e \'ery different networks because they are "people whose job is not water" and "the roundtables are composed of folks who aren't traditionally in the [water policy] community." The remaining roundtable members received letters and recruitment telephone calls. In alignment with the participatory action research approach, key water stakeholders on the 110

PAGE 130

roundtable participated in recrwung participation from other roundtable members, through emails and calls to those who had not taken the survey. ,\ttempts to secure smYeys from roundtable members resulted in an additional10 participants. The contact information for the external stakeholders identified on the original roundtable surYeys came primarily from two sources: a database of onr 1600 \Vater stakeholders maintained by the Colorado Institute of Public Policy and follow-up telephone calls as needed to survey participants who named specific contacts. be external stakeholders in the second round of the SUJTey were told only that one or more roundtable members identified them, but not which roundtable member(s). Analyzing the Data The analysis strategy was designed to take advantage of the unique collection of data; the combination of net\vork data, extensi,e demographic data, data on beliefs, and data on perceptions of success; and the inclusion of detailed whole net\vork information specifically among the Colorado Basin Roundtable members. Some data were analyzed primarily for the report that was delinred to the funder and the water roundtable stakeholders throughout the state. The discussion below more narrowly focuses on the analvsis most relevant to this dissertation. The first step of the analysis was to denlop the belief coalitions based on the nine variables from rhe survey that asked about beliefs related to the water roundtable process. Cluster analysis is ideal for this purpose as it is a multinriate analytical strategy that seeks to find homogeneity among cases (Aldenderfer & 13lashficld, 1984). A K111

PAGE 131

means cluster analYsis usmg SPSS was conducted on the belief statements across all part1c1pants 1n the study to identify clusters of beliefs. A second analYsis was run to identify the statements that most distinguished the clusters from one another. ANOVA was run on the nine statements, comparing the average response on each statement within each cluster. These clusters are treated as the belief coalitions described by the AdnJCacy Coalition Framework (Sabatier & Jenkins-Smith, 1993). All of the analyses used to explore the propositions began with these clusters as the independent variables. The second step in the analysis was to analyze membership in the clusters compared to affiliations and other demographics. Chi-Squares were used to analyze the demographics within clusters in a series of separate analysis, with the cluster assigned to each case from the k-means cluster analysis as the independent variable and the demographic measures as the dependent variables. Chi-Square is one of the few analytic techniques appropriate when both the independent ,ariable and dependent variable are nominal and the independent nriable represents independent groups (Leech, Barrett, & l\lorgan, 2005). This analysis was used to explore proposition one. Propo.rition 1: Participatory processes are composed of individuals who represent policy interests, also known as advocacy coalitions, and these coalitions are defined by shared beliefs rather than demographic characteristics. The third step in the analysis was to explore the relationships on the Colorado Basin Roundtable within and between clusters. This analysis grouped the network ties of cluster members within their cluster, to each of the other clusters, and to the roundtable 112

PAGE 132

as a whole. t-.Iore detail on this analysis is pro,ided later in the chapter. This analYsis was used to explore the second proposition. PropoJilion 2: The coalitions represented in a policy process will engage in shared actions, as seen by their shared information and trust in each other. The final step in the analysis were .\NO\'.\s to analyze the perceptions held by cluster members of the success of the roundtable process. ,\.s the Likert scale questions for the perceptions of success can be treated as scale data if they arc normally distributed, ANOVAs are an appropriate analytic technique (Leech, et al., 2005). The analysis \Vas used to explore the third proposition. In a participatory process, "representation of interests" may not equate to "representing interests" if the stakeholders are not equally engaged in the process. The sections below explore each analysis strategy in more detail, and in particular focus on the analysis of the network data, as it utilized a different approach than many network studies. Where most network studies begin with network analysis and then usc attribute information to understand the resulting network structure, this used the clusters that were created from non-network data as the basis of the study, with the network data analyzed \Vithin the context of the clusters to understand the relations among clusters. This strategy was appropriate as the three propositions are interested in the belief coalitions, \vhich are defined by beliefs as opposed to network relations. 113

PAGE 133

. Ja!)'Ji.r S tratrgJ.for the Bdie/ C!I!Jlt'l:f SutTcy participants were asked to respond with their level of agreement to nine statements coYering a nriety of priorities for the HB 1177 process. One hundred and sixty suryey respondents answered all nine of these questions. Cluster analysis was used to create the belief coalitions from the nine belief statements. K-means cluster analysis is appropriate because it utilizes a partitional clustering algorithm that defines discrete groups (lain, l\lurty, & Flynn, 1999), with each survey respondent in only one group, matching the theoretical model of belief coalitions in the advocacy coalition framework. The data were initially analyzed using hierarchical cluster analysis, Yisually allowing for an exploration of the grouping. However, the flllal analysis of the data was done using K means cluster analysis to produce "typologies" among the respondents to the suryey. K means data analysis has been used in previous studies seeking to identify advocacy coalitions in complex policy enYironments (Svihula & Estes, 2007) and has the advantage of producing clusters which are both discrete and easy to interpret (Grant, n.d.) and not assuming hierarchy exists within the data. K-means cluster analysis is an iterative method that seeks to minimize distance between clusters using an Euclidean distance measure. f\lultiple subsequent iterations based upon the nearest Euclidean distance to the mean of the cluster arc conducted until the cluster means no longer shift cases (Garson, 2006). Interpretation of k-means clusters begins by identifying the mean and standard de,iation for each statement in each cluster, looking for the themes that bring the 114

PAGE 134

clusters together. Once a set of five clusters was determined to be the "best fit" using kmeans, ANO\'.\ and Games-Howell Post Hoc Tests were used to identify the statements where each cluster differed significantly from the others. The .\NO\',\ and post hoc test were not intended as a nlidation strategy and, in fact, would be inappropriate as such, as significant differences are inc,itably found in any cluster analysis groupings (Aldenderfer & Blashfield, 19R4) as the cluster method is designed to create significantly different clusters. Instead, the ANOVA and post hoc tests \Vere used as an explanatory de,ice, helping to identify defining features of each cluster. Anab'JiJ f!fDemograpbio Demographic data, most of which is nominal, were analyzed using descriptiH techniques and chi-squares to determine significance. The data were analyzed to better understand the clusters, thus the chi-square analyses was done with the belief clusters as the independent ,ariable and the demographic nriables as the dependent nriables. Given that both the independent and dependent variables are nominal, chi-square is one of the few measures of significance that can be appropriately used with this data (Leech, et al., 2005). Anab.reJ l'.ring the j\'e!JJJork Data The cluster data from the k-means cluster analysis were combined with the network data to understand the relationships within and between the fiye clusters. .-\s described aboYe, the Colorado Basin Roundtable members were asked which members on their roundtable they: consider the most influential statewide; exchange information 115

PAGE 135

with; receiYe important informacion from; belieYc most share their goals; trust the most to keep their interests in mind, regardless of whether they share their goals; and can depend on to follow through on a commitment. To analyze the network data captured from these questions, the first step was to separate out the members of the roundtable into the five belief clusters. L'nlikely many network studies, where the grouping of members is done based on the network data and the relationships it shows, for this study the belief data are the agent used to create the clusters, as this matches the theoretical argument that beliefs are an important dri,er of interaction in participatory processes. Tlus study sought to understand how members of each cluster interact with and members of their O\Vn cluster and of other clusters. This is different from a network study that seeks to establish clusters based on the network relationships. Thus, to use net\vork data when the clusters in the nenvork are pre-defined and it is the interactions of cluster members that are of interest, the nenvork data had to be manipulated to create a new set of variables that help to understand each respondent and his/her interactions and perceptions of each of the clusters. Nenvork methodologists recognize net\vork measures as being attributes of nodes, e.g., centrality is often described as a "structural" attribute of the node (Borgatci, EYerett, & Freeman, 1992). In this study, each network measure was treated sinlllarly, as an attribute of the node that helped to describe the nodes structural posicion \Vithin its cluster and within the broader roundtable nenvork. This attribute deri,ed from nenvork data becomes a continuous variable that is manipulated in SPSS in a vanetY of ways. A combination of data 116

PAGE 136

preparation m Excel and SPSS data anahsis was used to create and analyze the new ,ariables created from the network data. For each of the stx network data suney questions, the same analytic techniques were used to create the new Yariables. The description to follo\V uses the influence question as an example. This same process was used with each of the other network data questions. Creating the New l 'miableJFive matrices were created for each network question using ExceL Each matrix was specific to one of the clusters generated from the belief analysis, with the ro\vs representing only those cases that were in the cluster and the columns representing all of the cases on the roundtable. The numbers 111 each cell represent the reported relationships from each of the cluster members (in the rows) to all of the roundtable members (Table 4.2). Thus, fiye different matrixes were created for the influence relation that each had onhthose rmvs that matched the cases that were in each of the clusters. Each cell was filled with the reported strength of the tie, from zero to Empt) cells resulted from either missing data (no report from the cluster member for a specific roundtable member) or a self-referential cell (the cluster member was not expected to proYide a network tie to him/herselfl). 117

PAGE 137

Table 4.2. Example Nenvork i\Iatrix: Influence Question for Cluster 1 tC 2C 3C 4C sc 6C 7C sc 9C tOC llC t2C 13C t4C 9C () () I) 2 0 0 0 2 0 0 0 () toe 0 0 0 2 0 0 ") () 0 2 0 2 () <13C 2 () 0 () () 0 0 () () 0 0 I 22C () () () 0 () () 0 3 3 3 -+ () 3 25C 0 () () 3 u 0 0 3 3 0 3 () 2 31C 0 0 () 0 () 0 0 () 0 () () 0 0 0 33C () 0 () 3 0 0 () (I 3 () 2 () 0 2 34C () 0 3 3 0 0 () 3 3 () 3 0 (I 2 48C 0 (I 3 2 0 0 0 3 3 0 3 0 0 2 l'sing this matrix, an aYerage \Vas created for each respondents reported relationship with all of the members of their cluster (Table 4.3). Table 4.3. Example calculation of aYerages for the Influence Question from Cluster 1 members to all other members of Cluster 1. Cluster 9C toe 13C 22C 25C 3tC 33C 34C 48C t 9C 0 0 0 2 0 0 3 2 .\verage for 9C: 0.88 tOC 0 2 3 0 () 3 0 0 .\ verage for 1 OC: 1.00 13C 1 0 2 3 2 () 0 () .\\-erage for 13C: 1.00 22C 3 3 3 -+ 0 3 -+ 0 .h-erage for 22C: 2.50 25C 3 0 1 0 2 0 -+ 0 .\verage for 25C: 1.25 3tC () 0 0 () 3 0 0 0 .\ verage for 31 C: (J.38 33C 3 0 0 0 3 () 0 2 .\,erage for 33C: 1.00 34C 3 0 0 0 2 3 0 3 .\ verage for 3-+C: 1.38 48C 3 0 () 0 -+ () .-\verage for 48C: 1.50 Creating an a\erage across the ues reported for each respondent IS an appropriate use of descriptive statistics on the relations within the net\vork. The anrage helps to aggregate the information known about the relations of each individual in the net\vork (Hanneman & Riddle, 2005). For example, for the matrix above, an average was created using the numbers reported by 9C for all of the cluster members (listed in 118

PAGE 138

the flrst column). This a\erage was put into SPSS as a nriable that represents how the indi,idual survey respondent views other of his or her own cluster on the influence network question. The same calculation was completed for each of the remaining clusters, so that an aYerage score was created for how each member of Cluster 1 ,iews the members of Cluster 2 (Table 4.4), Cluster 3, Cluster 4, etc. As the final analrsis included flve clusters, for each network question, flye new nriables were created in SPSS, representing the individual surrey respondents' aYerage responses on the network question for the members of each of the flve clusters. Table 4.4. Example calculation of aYerages for the Influence Question from Cluster members to Cluster 2 members. lC 2C 4C llC 35C 42C Cluster 2 9C 0 0 2 2 0 :herage for 9C: 0.83 toe () 0 .., 1 3 () \ wrage for 1 OC: 1.17 13C 2 0 0 () 3 0 .\ ,erage for 13C: 0.83 22C 0 0 () 0 :\ ,erage for 22C: 1.33 25C 0 0 3 3 3 0 :\xerage for 25C: !.SO 31C 0 0 0 0 0 () :\Yerage for 31C: 0.00 33C 0 0 3 2 .3 0 :\ Yerage for 33C: 1.33 34C 0 0 3 3 3 0 :\Yerage for !.SO 48C () 0 2 .3 2 0 _\ ,-erage for 1.17 A sixth variable \vas created for each network question that included the oYerall average for the survey respondents for their ,-iews of all Roundtable l\lcmbers included in the analysis. In total, this resulted in an additional six Yariables per network question, for a total of thirtr-six nc\v variables. 119

PAGE 139

. the Ne!Jvork I 'miab/e.r The thirtY-six new network variables were analvzed usmg one-way ANO\' As. :\lthough traditionally network data are analyzed using analysis techniques specific to networks, more recently researchers ha,e acknowledged that descriptive and inferential statistics are important tools in understanding networks. Inferential statistics can be used to assess whether or not the patterns found in network data are random. Appropriate questions for inferential statistics with network data include such things as the extent to which the strength of ties is related to an actors' attributes, such as the centrality of the actor in the network (Hanneman & Riddle, 2005). In this study, the strength of ties is of interest based on the attribute deYeloped through the cluster analysis. Prior to analyzing with the ANOVAs, a Shapiro-Wilk test for normalitY of distribution was used to confirm that the majority of the new variables were distributed normally within their clusters. Overall, 85.4% (82 of 96) of the variables created from the network data \Vere distributed normally within each of the clusters (see Appendix B for details). This analysis excluded the ,-ariables specific to the Current System Broken cluster due to the small size (11 = 2) of the cluster. "\!though normal distribution may not be typical in social science data, in this case, the normal distribution was found within clusters that were grouped based on shared beliefs on questions that asked them about individuals who were members of other clusters also grouped based on shared beliefs. For the ANO\' As, the Independent \' ariable was the k-means cluster membership dcYeloped using the water roundtable goals quesuons. The dependent 120

PAGE 140

variables were the different a\erages for each survey respondent, representing their responses on each network question to all of their 0\Vn cluster members and the cluster members of each of the other clusters, as well as to the roundtable as a \Vholc. ANOVAs are an appropriate analytic measure as the independent variable is nominal with fin categories and the dependent nriables are scale (Leech, et al., 2005). Ana!J.ti. q/ Pemptiom q/Sm-ce.r.r The perceptions of success data were collected using a five-point I .ikert scale. This data helped explore Proposition 3. When normally distributed, a five-point Likert scale variable can be treated as scale data, prmided the distance between each point on the scale is thought to be equal. f.'or this reason, ANOVAs are the most appropriate statistical technique, as they explore the differences between the means on the scale for each subgroup (Leech, et al., 2005). .\s the goal of the analysis is to compare the clusters, the cluster was the independent variable and each question of the perception of success \vas the dependent variable. lntnpreling !be l{e.r;t/!J The research results were interpreted in a three-step process to ensure full stakeholder engagement, consistent with the community-based participatory research approach. After the data \Vere analyzed and interpreted, a draft of the results was prepared and disseminated to leaders in the water policy community. The leaders included representatives of many different interests, indi,iduals who responded to the survey, individuals in leadership roles on the Colorado Basin Roundtable, individuals 121

PAGE 141

who piloted the study, and the funders of the study. In addition to receiYing the results, the key stakeholders \vere asked specific questions about the interpretation. The focus \Vas on ensuring the interpretation of the belief clusters reflected their understanding and experience in the water policy community. This was an important step to ensure the first step in the data analysis, creating the clusters that served as the basis of all the remaining steps, was interpreted accurately. The reviewers were also asked about the interpretation of the network findings. They were asked to help explore the potential reasons for the results, as this interpretation was important to ensure the information was useful when it was prmided to the water policy community. Based on email and telephone call responses to the original findings, revisions were made to more clearly and accurately describe the findings in the context of water policy issues in Colorado. A second version of the fmdings was sent out to a smaller group of key stakeholders who had provided more detailed feedback on the original interpretations. With only minor revisions, the interpretations were finalized and used as the basis of the report for the water community and this dissertation. Validity and Limitations In a study whose design is driven by the participants, with measures adapted to answer their questions, the issue of validity changes somewhat. Some researchers have even asked whether the construct of "validity" as construed by academics coincides \Vith the concept of "nlid" research for the participants (l\larti & Villasante, 2009). For this dissertation, the research questions emerged from dialogues with stakeholders who were 122

PAGE 142

questioning the ability of the roundtable process to lead to new approaches for \Vater management and interbasin transfers. The choice to focus the study on the Colorado Basin Roundtable was for pragmatic reasons: political will was present and the funding \vas sufficient to undertake the work with one roundtable as a starting point. HoweYer, the stakeholders inYolYed expressed interest in thinking through how this study could inform the remaining roundtables. Thus, external Yalidity as defined by researchers is also a form of Yalidity that the participants valued. For the researcher, external nlidity is important though the study does not seek to generalize its fmdings to a specific, larger population, because it does seek to examine public participation theories and the concept of representation within a small group context that may be releYant to other settings. Both internal and external nlidity were addressed through three steps: first, the use of measures from preYious respected studies; second, the imoh-ement of stakeholders 111 the iterati,-e process of deYeloping the final suryey; third, the inYolvement of the same stakeholders in the iterative process of interpreting the findings, helping internal validity by considering alternatin explanations for the findings and narrowing to those explanations that most resonated with the stakeholders. The nine statements on priorities for the water roundtable, the basis of the belief coalition analysis, were drawn from two previous studies conducted in the same Colorado water policy environment, the Colorado Institute of Public Policy's (2006) report that used g methodology and the CDJ\1 (2006) report that had measures showing high ,-ariabil.ity among participants, suggesting these are areas where differences in stakeholder 123

PAGE 143

preferences could prmide meaningful distinctions between factions. As noted earlier, the network analysis questions were initially drawn from pre,ious network studies (e.g. Weible & Sabatier, 2005). The reYisions of the questions were tested repeatedly with pilot participants. Each pilot test included a detailed debrief to ensure the interpretation of the question matched the intent of the original question, even after it was customized to fit within the water policy em-ironment in Colorado. The internal validity of the measures was imprmed by this careful piloting process. For example, the decision on how to measure information sharing and trust was reYised during piloting, with each of this concepts split into two questions that more clearly captured the types of relationships that exist in the water community and had Yalue to the study and its desired outcome. Consistent with a participatory research approach, after the data were collected and an initial analysis completed, the stakeholders helped to interpret the fmdings (O'Brien, 2001; O'Fallon & Dearry, 2002). The belief coalitions described in chapter fiye were defined and named in partnership with stakeholders who reviewed the findings and considered their implications. The conclusions in chapter six are based on the interpretations discussed by stakeholders who represent the different belief coalitions found in the studY. Some of the stakeholders helping to review and interpret the findings had taken the surYey themselves, and learned which belief coalitions they fell into. Their self-reflection was part of the reYiew process. 124

PAGE 144

Due to this learning process and the increased credibility of the study as a result of the stakeholder engagement, not only are internal and external ,alidity increased, but so to is the "catalytic Yalidin." Catalrtic nlidin-, a term coined br action researchers .. .. "' refers to the empowering process of action research by which change can occur (Anderson, Herr, & Nihlen, 1994). In essence, this type of nlidity is primarily demonstrated not through the design of the research, but through the outcomes of the research the eYidence of change (Hope & Waterman, 2003). In this case, the presentation of the results to the full Colorado Basin Roundtable resulted in: (I) a discussion by roundtable members that began \vith unprompted expressions of agreement with the findings, demonstrating face Yalidity in the interpretations of the findings; and (2) a facilitated process with the roundtable members leading to new tools to address the disconnects benveen different belief coalitions paructpating in and peripheral to their roundtable process. The research sened as a catalYst for a ne\v approach to making, implementing, and documenting decisions at the roundtable to increase transparency and participation of interest groups. LimitalioiiJ As with any study, the research design has a number of limitations. The external stakeholders who were surveyed to capture the broader belief coalitions of the water policy community in Colorado are being identified through snowball sampling, Yia the net\vork questions. They are already connected to the Colorado \Vater Roundtable Yia at least one participant on the roundtable. It is not unrealistic to suspect that some belief 125

PAGE 145

coalitions that are present in Colorado could be left out of the sample as a consequence of the design. Ideally, other water stakeholders would be identified through alternative mechanisms, such as surYeying a sample of citizens in the region. HoweYer, given the complexity of the suney and the limit of the research budget, sunreying the stakeholders identified by the roundtable participants is more realistic. Additionally, the examination of representation goes beyond whether interests are at the table, to explore how they sit at the table (influence and information exchange). For that reason, even if an interest is unknown and thus missing, the information gathered on the remaining interests is still meaningful and can help to understand the representation issues in a participatory process. Another weakness of the study is its inability to determine any casual relationships. This is a weakness conunon in many participation studies, where little research has been done to establish a connection bet\veen the design of a participatory process and its outcomes. Casual inference concerning the relationship bet\veen outcomes and representation would require studying multiple roundtables to determine if their differing net\vork structures are associated with different outcomes. It would also benefit from studying the roundtable process when it has had more time to accomplish its goals. In this studY, the need for immediate information to help the political process makes it difficult to study the final outcomes of the roundtables. A third weakness of the study is that it is drawn upon a single point in time. The cutting edge of net\vork theory includes the study of dynamic net\vorks, the 126

PAGE 146

understanding of how networks denlop and change over time (Carley, 2003), and it would han been ideal to be able to add this study to that emerging body of literature. By studying relationships at a single point in time, this study cannot capture how the participatory process may affect those relationships and either mitigate the impact of belief coalitions or the opposite. The survey captures a single point in time during the HB 1177 process during which the roundtable process was still relati,ely new. The roundtable process \Vas too nascent at the time of the research for an analysis of its outcomes, thus limiting the study to a focus on the process and interactions within it. Each roundtable has a great deal of latitude in designing its structure, number of members, by-laws, etc. This analysis reYcals the state of affairs at one point in time, starting from one roundtable, the Colorado Basin Roundtable. The results cannot be inferred to all of the roundtables. Howenr, the roundtable process is not only a policy process, but also a political process still in its infancy. With a new political administration elected and appointed as of January 2007, immediate and useful results for policymakers \vere needed to help inform whether the roundtable process should continue and what adaptations may need to be made. Due to this practical need that the research was intended to meet, a policy analysis that focuses on the immediate need and recognizes that the roundtables may be discontinued or significantly changed in format in the coming years is more useful than a longitudinal study. 127

PAGE 147

It is important to remember that one of the limitations of the non-random sample is that the results cannot be inferred to water stakeholders throughout Colorado, to say nothing of outside Colorado. 1l1e design of the suryey, using a network analysis approach, docs not rely on random sampling of stakeholders. Rather, each member of the Colorado Basin Roundtable who took the suryey had the opportunity to identify key indiYiduals in water policy and management issues in Colorado with whom they exchange information with or to whom they are otherwise connected. The suney im-itation only went to those indiYiduals named by members of the Colorado Basin Roundtable. They are not representative of all stakeholders. However, the results are able to paint a picture of a group of stakeholders who are interconnected in many ways and hea,ily inYolved in water policy and management issues in Colorado. Finally, though the recruitment methods were very thorough, there were some limitations to the recruitment process. In four cases, contact information was not a\ailable despite using multiple means to track down the water stakeholders identified by roundtable members. lndi,iduals external to the suney who were frequently identified as influential in the water community were the least likely to agree to participate in the suney. Recruitment methods that relied upon the research team, rather than a well respected water stakeholder, to imite these leading members of the water community may be partially responsible for the lower response rate among this group. 128

PAGE 148

Conclusion This dissertation is much more than an addition to the academic literature on representation and democracy. It is applied policy research that works directly with stakeholders in a policy setting that is in transition. As a network goYernance approach begins in Colorado's water community, both the participants and other stakeholders benefit from better understanding how the roundtable model is functioning. \re there barriers within the coalitions, or alternatiYely, opportunities to create a more successful process using key indiYiduals and their networks? 129

PAGE 149

CHAPTER 5 ANALYSIS One hundred and SL'\:tY people responded to a sufficient number of sutTeY questions to be included in the analysis, taking the time to answer the questions on their goals for the \Vater roundtable process, their relationships with others on the roundtable or in the \Vater community, and their demographics. Tlus information was used to complete the four types of analysis described in the methods chapter: deYeloping the belief coalitions, comparing demographics to belief coalition membership, exploring belief coalitions through the use of network data, and comparing perceptions of success across belief coalitions. The four types of analysis allow for an exploration of the three propositions. Some of the analysis was only possible among water roundtable members, resulting in a small population for the analysis and limited ability to find significance. Despite that, the three comparatiYe analyses conducted all had significant findings. DeYeloping Belief Coalitions .-\ll three propositions focus on adYocacy coalitions, defined by the beliefs of the coalition members. For this, the first analysis undertaken was to deYelop the clusters that represent belief coalitions in this study. One hundred and sixty of the suney respondents answered all nine Likert questions on their beliefs related to water policy 130

PAGE 150

and management in Colorado. As the goal of the quesuons was to identifv belief coalitions, the nine questions were analyzed together using k-means cluster analysis. The respondents fell into five distinct clusters, each with its own characteristic beliefs and desired outcomes from the Colorado Basin Roundtable and HB 1177 process. The description below provides the detail of the analysis as well as an meniew of each cluster and a comparison between them. These clusters will be used throughout the remainder of the chapter to examine the three propositions. As discussed in the methods chapter, respondents were asked for their le,el of support, from strongly agree to strongly disagree, for the nine statements related to priorities for the HB 1177 roundtable process to address: (1) allocation and management of water resources through the market; (2) protecting existing individual water rights; (3) water transfers to high growth areas/sectors; (4) balancing water supplies and demands; (5) increasing cooperation among water basins; (6) protecting the agricultural economy and way of life; (7) protecting the recreational economy and its water needs; (8) protecting ecosystems and non-human species (or just the environment); and (9) balancing the water demands between consumpti,e uses and non-consumptive uses. Answers to the nine statements were first analyzed using k-means cluster analysis, then analyzed using ANOVAs. Re.rult.r oft be C/11.rler /lnafpi.r The responses to the nine belief questions were initially analyzed us111g a hierarchical cluster analysis to ,isualize the cases in a dendogram and identify how many 131

PAGE 151

possible clusters might exist within the data. Yisually, the dendogram showed fiYC distinct groups and one loose group at the end. For this reason, when the k-means analysis was run in SPSS, a five cluster analysis was selected. The resulting five cluster model balanced the need for clearly defined clusters and the desire to limit the total number of clusters to a meaningful number of "interests." The five clusters varied in size, and contained 35, 19, 50, 47, and 9 respondents. Once the five cluster analysis was run, the mean and standard deviation for how the members of each cluster responded to each statement was identified to help understand which statements defined each cluster. The answer scale for all nine questions was from 1 = Strongly Agree to 5 = Strongly Disagree. Thus, the higher the mean, the more the respondents in the cluster di.ragreed with the statement. CIIIJler 1 (fable 5.1) has the greatest agreement with the statements related to increasing cooperation among basins and protecting indi,idual water rights (1.71 and 1.6 respectively), followed by the remaining three statements that largely focus on consumptive needs. This cluster was the least likely to support non-consumpuve priorities (recreational and em-ironmental), with means of 2.94 and 3.14 respectively. For this reason, cluster one is named Protecting Co11J11!11p1ille !\eedJ for the remainder of the analysis. The standard deviations are relatively lmv, but lowest on those statements where the cluster members disagreed with the statement. 132

PAGE 152

Table 5.1. Means and standard de,riations for Cluster 1, Protecting Consumptin Needs # Question :\lean SD (n= 35) 8 HB 11 77 should support protecting existing indiYidual water 1.6 0.775 rights. 5 HB 1177 should support increasing cooperation among 1.71 0.92() water basins. 2 HB 1177 should support protecting the agricultural way of 2.11 0.867 life. 1 HB 1177 should support water transfers to high growth 2.37 0.731 areas/ sectors. 7 HB 1177 should support balancing \Vater supplies and 2.37 0.808 demands. 9 HB 1177 should support allocation and management of 2.37 0.80R water resources through the market. 4 HB 1177 should support protecting ecosystems and non2.74 0.561 human species (the environment) 3 HB 1177 should support protecting the recreational 2.94 0.684 economy and its water needs. 6 HB 1177 should support balance the water demands 3.14 0.733 between consumptiYe uses and non-consumptiYe uses. CIIIJ!er 2 (fable 5.2) is similar to cluster one in its support of protecting indiYidual water rights and supporting cooperation among the basins (means of 1.21 and 1.47, respectinly). It differs from cluster one in its slightly stronger support for the agricultural way of life (1.58), and its very strong disagreement with prioritizing water transfers to high-growth sectors (4.53). Transfers to high growth sectors are largely the result of municipal water needs. With similar means on the remaining consumptiYe and non-consumptive statements, the distinguishing feature of cluster two from cluster one appears to be its lack of support for the statement most related to municipal needs. 1-'or 133

PAGE 153

this reason, cluster two is named Proteding ./1grimlt11ml 1'\eed.f for the remainder of the analYsis. Table 5.2. l\leans and standard deYiations for Cluster 2, Protecting Agricultural Needs. # Question Mean SD (n=19) 8 HB 1177 should support protecting existing indiYidual water 1.21 0.419 rights. 5 HB 1177 should support increasing cooperation among water 1.47 0.513 basins. 2 HB 1177 should support protecting the agricultural way of life. 1.58 0.507 7 HB 1177 should support balancing water supplies and demands. 1.84 0.602 6 HB 1177 should support balance the water demands between 2.89 1.150 consumpti,-e uses and non-consumptive uses. 9 HB 1177 should support allocation and management of water 3.11 1.100 resources through the market. 4 HB1177 should support protecting ecosystems and non-human 3.26 1.098 52ecies (the em-ironment) 3 HB 1177 should support protecting the recreational economy 3.26 0.872 and its water needs. 1 HB 1177 should support water transfers to high growth 4.53 0.612 areas/ sectors. C/11.rler 3 (Table 5.3) largely agrees with the statements, and is neutral on one statement, transfers to high growth sectors. The strongest agreement, with a ,-ery small standard deYiation, is on increasing cooperation among water basins. The means for balancing consumptive and non-consumptive uses, protecting individual rights, and balancing supply and demand are all around 1.5, with relatiYely low standard deviations. The means and standard deviations suggest this group has an interest in balancing the 134

PAGE 154

\'ariety of needs, increasing cooperation, and generally supporting all the different needs. For this reason, cluster three is named Balaming j\'eedJ for the remainder of the analysis. Table 5.3. t\leans and standard deviations for Cluster 3, Balancing Needs. # Question i\Iean SD (n=50) 5 HB 1177 should support increasing cooperation among 1.06 0.240 water basins. 7 HB 1177 should support balancing water supplies and 1.5 0.544 demands. 8 HB 1177 should support protecting existing individual 1.52 0.735 water rights. 6 HB 1177 should support balance the water demands 1.52 0.58 between consumptive uses and non-consumptiYe uses. 2 HB 1177 should support protecting the agricultural way of 1.84 0.817 life. 4 HB 1177 should support protecting ecosystems and non-1.86 0.670 human species (the emironment) 9 HB 1177 should support allocation and management of 2.12 0.746 water resources through the market. 3 HB 1177 should support protecting the recreational 2.12 0.558 economv and its water needs. 1 HB 1177 should support water transfers to high growth 2.9 0.839 areas/ sectors. Cl11.rter .J (rable 5.4) has the strongest agreement, with low standard de,iations, on the three statements that support non-consumptiYe uses with means of 1.4, 1.55, and 1.62. The areas of strong disagreement are focused on consumptive uses, including transfers to high growth sectors, and on using the market approach in the roundtable process. This cluster is moderate in its agreement for supporting the agricultural way of life, individual water rights, and balancing supply and demand. Due to its top 135

PAGE 155

three statements, this cluster will be named Protediug j"\'on-Con.I'111J1ptil't' Need." for the remainder of the analYsis. Table 5.4. 1\Ieans and standard de,iations for Cluster 4, Protecting Non-ConsumptiYe Needs # Question Mean SD {_n=47) 4 HB 1177 should support protecting ecosystems and non1.4 0.538 human species (the em-ironment) 3 HB 1177 should support protecting the recreational 1.55 0.544 economy and its water needs. 6 HB 1177 should support balance the water demands 1.62 0.491 between consumptive uses and non-consumptive uses. 5 HB 1177 should support increasing cooperation among 1.72 0.826 water basins. 2 HB 1177 should support protecting the agricultural way of 1.89 0.699 life. 8 HB 1177 should support protecting existing indiYidual 2 0.692 water rights. 7 HB 1177 should support balancing water supplies and 2.13 0.741 demands. 9 HB 1177 should support allocation and management of 3.36 0.942 water resources through the market. 1 HB 1177 should support water transfers to high growth 4.15 0.780 areas/ sectors. Cl/l.ller 5 (fable 5.5) is the smallest cluster, with only nine members among the 160 included in this analYsis. The four areas of strong agreement are about balancing conflicting issues (e.g., supply and demand; consumptiYe and non-consumptive uses). This cluster disagrees with more statements than any of the other clusters. The three strongest areas of disagreement are individual \Vater rights (3.67), market management of \Vater resources (4.44) and transfers to high growth sectors (4.67). These three 136

PAGE 156

statements strongly describe the current legal model for water management, \vhich is based on the market, protects individual water rights, and has led to many cross-basin transfers to moYe \Vater from the Western Slope to the Front Range to accommodate growth. For these reasons, this cluster will be named Clm-en/ S_plem Bmk.m for the remainder of the analysis. Table 5.5. !\leans and standard de,'iations for Cluster 5, Current System Broken. # Question i\Iean SD (n=9) 7 HB 1177 should support balancing water supplies and demands. 1.22 1 6 HB 1177 should support balance the water demands between 1.22 0.7 consumptive uses and non-consumptiYe uses. 4 HB 1177 should support protecting ecosystems and non-human 1.33 0.441 species (the environment) 5 HB1177 should support increasing cooperation among water 1.67 1 basins. 3 HB 1177 should support protecting the recreational economy and ') 0.441 its water needs. 2 HB 1177 should support protecting the agricultural way of life. 3.11 1 8 HB 1177 should support protecting existing indi,-idual water 3.67 0.928 rights. 9 HB 1177 should support allocation and management of water 4.44 0.5 resources through the market. 1 HB 1177 should support water transfers to high growth 4.67 0.882 areas/ sectors. Compming tbr CIIIJ/er.r The initial interpretation of the clusters was done using the means and standard deYiations of each cluster. To conftrm the interpretation and clearly define how the 137

PAGE 157

clusters differ from one another in beliefs, an ANOYA was used on each of the nine statements to identify the statements that significantly contributed to the differences between the clusters. As noted in the preYious chapter, an ANOVJ\ is not a technique for Yalidation, and in fact would be inappropriate for such purpose. However, it does help identify the statements that most distinguish the clusters from one another. The conclusion from the "\NOVA is that among the 160 respondents included in the fin clusters, there is no \alue statement on which all are generally in agreement or disagreement. Conflicting Yiews are significantly present between the fiye groups, with each group in conflict with every other group on multiple statements. A more thorough study of \alues may find more agreement (e.g. see CIPP, 2006), but on the key issues cmered by the nine statements, areas of disagreement are prominent. The nine tables below represent the nine statements and show the results of the ANOYA. Accompanying each table is a chart that shows the means, allowing for a \isual understanding of the data. Following the tables, a description of the means, standard de\iations, and the "\NOYA is combined to begin the interpretation of the clusters. For all nine tables below (one table for each question), is present if there is a significant difference (at 0.05 or better) between the mean answers of the two clusters being compared on the belief statement. Significance was determined using a Games-Howell Post Hoc Test, selected because the variables have unequal variance (l'.Iorgan, Leech, Gloeckner, & Barrett, 2004). 138

PAGE 158

Table 5.6. Significant Differences Berween Clusters on Question 1: HB 1177 Should Support Water Transfers to High Growth Areas/Sectors Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Mean Cluster 1 2.37 4.53 2.90 4.15 4.67 Strongh \grc.: .\grcc Ncithn \grco: or --Yes Yes Yes Yes Cluster 2 Cluster 3 Cluster 4 Yes Yes Yes --Yes Yes --Yes Yes --Yes [J 0 m Strongly I )i:-'agrct: ------------------j Figure 5.1. Means on Question 1 Cluster 5 Yes Yes --Cluster 1 Cluster 2 Cluster 3 Cluster-+ Cluster S As can be seen in Table 5.6 and Figure 5.1, only Cluster 1, Proteding CollJIIIlljJiiJe NeedJ, agreed with this statement, and Cluster 3, Bala11dng I\'eedJ. remained neutral. Clusters 2, Protedi11g /1g1imltllral Nt!edJ, and Cluster 5, Cl.m"l!nf 5_yJtem Brokm. strongly disagreed with this statement, with a mean over 4.5 on a scale of 1 -5. This statement becomes one of the distinguishing statements between Cluster 3 and Clusters 4 and 5, which haYe similar means on many other statements. 139

PAGE 159

Table 5.7. Significant Differences Between Clusters on Question 2: HBt 177 Should Support Protecting the :\gricultural Economy and Way of Life Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 I\Iean Cluster 1 2.11 ---1.58 Yes 1.84 1.89 3.11 Strong)\' \grcc .\grcc l\cithn .\grcc or Cluster 2 Cluster 3 Cluster 4 Yes ---------Yes Yes Yes Strongh ----------------.. --------------Figure 5.2. I\Ieans on Question 2 Cluster 5 Yes Yes Yes -- Cluster 1 g Cluster 2 0 Cluster 3 Cluster 4 ml Cluster 5 can be seen in Table 5.7 and Figure 5.2, most respondents agreed with the statement that the HB 1177 process (the roundtable process) should protect the agricultural economy and way of life. Cluster S's neutral response to this statement was a significant difference. 140

PAGE 160

Table 5.8. Significant Differences Between Clusters on Question 3: HB 1177 Should Support Protecting the Recreational Economy and its Water Needs Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Mean Cluster 1 2.94 ---3.26 2.12 Yes 1.55 Yes 2.00 Strongh \grcc .\grcc 1\cithcr \grcc or Cluster 2 Cluster 3 Cluster 4 Yes Yes --Yes Yes Yes --Yes Yes Yes --Yes Strongly I )i,;agrcc --------------------------------! Figure 5.3. l\leans on Question 3 Cluster 5 Yes -- Cluster I Cluster 2 D Cluster 3 fill Cluster S As can be seen in Table 5.8 and Figure 5.3, most respondents agreed or were neutral on the statement that the HB 1177 process (the roundtable process) should protect recreation and its water needs, but with more 'Tariation than the pre,Tious statement on agriculture. Cluster 2, Proll'ding .lgrimltlfral NeedJ, was in the least agreement with this statement, closely follm,ed by Cluster 1, Proteding CoJWtlllp!ille I\'eetl.r. These two clusters were significantly different than the other clusters based on this statement. 141

PAGE 161

Table 5.9. Significant Differences Between Clusters on Question 4: HB 1177 Should Support Protecting Ecosystems and Non-Human Species (or just the Environment) Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 l\lean Cluster 1 2.74 --3.26 1.86 Yes 1.4 Yes 1.33 Yes Strongh \gn:c .\grcc: '\cith<:r \grcl' or Figure 5.4. !\leans on Question 4 Cluster 2 ---Yes Yes Yes Cluster 3 Cluster 4 Cluster 5 Yes Yes Yes Yes Yes Yes ---Yes Yes ------ Cluster 1 lj Cluster 2 D Cluster 3 Cluster -t Em Cluster 5 .\s can be seen in Table 5.9 and Figure 5.4, Clusters 1 and 2, Proteding Con.mmptiJ,e l\eedJ and Proteding gJimllllral Need.l', are significantly different than the remaining clusters due to their neutral or negative response to the statement regarding protecting the emironment. Clusters 3, 4, and 5 are significantly different than Clusters 1 and 2 in their responses to this statement, with greater support for protecting em'ironmental needs. 142

PAGE 162

Table 5.10. Significant Differences Between Clusters on Question 5: HB 1177 Should Support Increasing Cooperation Among Water Basins Mean Answer Cluster 1 1.71 Cluster 2 1.47 Cluster 3 1.06 Cluster 4 1.72 Cluster 5 1.67 Strongh .\gn:c: .\gm: Nl'ithl'r \grl'l' or Cluster 1 ---Yes Cluster Cluster Cluster 2 3 4 Yes --Yes Yes --Yes Yes --; Strongh-------------------____ _____ j Figure 5.5. I\1eans on Question 5 Cluster 5 -- Cluster 1 11 Cluster 2 0 Cluster 3 Cluster -+ m ClusterS As can be seen in Table 5.10 and Figure 5.5, all clusters supported increasing cooperation among water basins. This statement is the greatest area of agreement among the nine statements, and Cluster 3, Balmtting J\'ml.r, is significantly more likely to agree with the statement than any other cluster. With a mean of 1.06 on a scale of one to five, almost all respondents in Cluster 3 strongly agree with this statement. .\s can be seen in Table 5.11 and Figure 5.6, as with the statements on recreation and environmental needs, Clusters 1 and 2, Pmtedillf!, ConJ/fmptire J\'eedr and 143

PAGE 163

. l,gJimllllml SadJ, are significantly different than the remammg clusters due to their neutral response to the statement on balancing water demands between consumpttYe (e.g. municipal, agricultural, and industrial) and non-consumptive uses (recreational and enYironmental). Table 5.11. Significant Differences Between Clusters on Question 6: HB1177 Should Support Balancing the Water Demands Between ConsumptiYe Uses and NonConsumptive l'ses l\Iean Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Cluster 1 3.14 --Yes Yes Yes Cluster 2 2.89 --Yes Yes Yes Cluster 3 1.52 Yes Yes ---Cluster 4 1.62 Yes Yes ---Cluster 5 1.22 Yes Yes --Stron).!ly \wee Cluster 1 .\grec li Cluster 2 D Cluster 3 !\:either .\grec Cluster 4 or 1m Cluster 5 J)j,;J)!fCl' Stron).!ly Di,;awn Figure 5.6. l\leans on Question 6 As can be seen in Table 5.12 and Figure 5.7, the response to this statement helps to distinguish Cluster 3, Balcmdng ,'\'eed.r, and Cluster 5, C11nmt 5_yxtem Broken. from Cluster Proteding Soii-CoiiJIIIIlfJiil'e t
PAGE 164

response to many of the previous statements, when it comes to balancing supply and demand, Clusters 3 and 5 are significantly different than Cluster 4, as \Veil as Cluster I, due to their greater agreement. Table 5.12. Significant Differences Between Clusters on Question 7: HB 1177 Should Support Balancing Water Supplies and Demands l\lean Cluster 1 2.37 Cluster 2 1.84 Cluster 3 1.50 Cluster 4 2.13 Cluster 5 1.22 .\gm: 1\ either \gn:c or Cluster 1 --Yes Yes Figure 5.7. Means on Question 7 Cluster 2 --Yes Cluster 3 Cluster 4 Cluster 5 Yes Yes Yes --Yes Yes --Yes Yes -- Cluster I IE Cluster 2 D Cluster 3 m ClusterS As can be seen in Table 5.13 and Figure 5.8, the statement on indiYidual water rights is Cluster 5's defining statement, Ctonn/ S_yJiem Brokm, separating it from all other clusters including Cluster 3, Balaming Need.r, and Cluster 4, Proteding :'\'on-Con.fto;;pti!'e Need.!', which haYe shared similar means to cluster fiye on many previous statements. "\s 145

PAGE 165

noted in chapter two, much of Colorado water law focuses on indiYidual rights. Cluster 5 is the only cluster to disagree with this statement. Table 5.13. Significant Differences Between Clusters on Question 8: HB 1177 Should Support Protecting Existing lndiYidual Water Rights Mean Cluster 1 Cluster 2 Cluster 1 1.60 ---Cluster 2 1.21 ---Cluster 3 1.52 Cluster 4 2.00 Yes Cluster 5 3.67 Yes Yes Strong!\' .\grn: r---------------Neither .\grcl" or Di>agrL"l' l)i,;agn:c Strongly Di,;agrcc Figure 5.8. Means on Question 8 Cluster 3 Cluster 4 Cluster 5 Yes Yes Yes --Yes Yes Yes --Yes Yes Yes -- Cluster 1 II Cluster 2 0 Cluster 3 Cluster 4 !Ill Cluster 5 As can be seen in Table 5.14 and Figure 5.9, the final belief statement asks the extent to which the roundtable process should support allocation through the current legal structure, the water market. Once again, the cluster most strongly disagreeing with this statement is Cluster 5, Cl!/1<'111 5_yJtm; Brokm, distinguishing it from the other clusters. HowcYer, Cluster 2, Proteding _-Jgriml111ral j\'eedJ, and Cluster 4, Proteding J\'on-CoJJJIIIJlplille j\"eedJ, also disagree with the statement, while Cluster 1, Protecting ConJIIlllplitJf ;'\,ieed.-. and Cluster 3, Balant"ing J'\'eed..-, agree with it, though not strongly. No cluster strongly 146

PAGE 166

supports the roundtable process focusing on the market-based model of water allocation. Table 5.14. Significant Differences Bet\veen Clusters on Question 9: HB 1177 Should Support Allocation and Management of Water Resources through the l\Iarkct Mean Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 1 2.37 --Yes Cluster 2 3.11 --Yes Cluster 3 2.12 Yes --Yes Cluster 4 3.36 Yes Yes ---Cluster 5 4.44 Yes Yes Yes Yes .\gn:c !\lith<.:r \grcc ur J)i,;agn.:c Strongly ()i,;agrcc Figure 5.9. !\leans on Question 9 5 1111/JJILIIJ' greement 011 talements rm D I?Z} 1m Cluster 5 Yes Yes Yes Yes ---Cluster 1 Cluster 2 Cluster .1 CJugter S As can be seen above, there are areas of agreement between all clusters. I ;or example, four of the five clusters are supportive of protecting existing water rights and the agricultural economy and way of life. Although Cluster 5, Cun"t!lll S.JJic'JII Brokm. 147

PAGE 167

disagreed with many statements, the cluster members share a desire for cooperation among basins and balancing \Vater supply and demand. S!lll!llltiD' DiJagreemml 011 lalemen/.r The areas of differences are also important to keep in inind. The allocation and management of water resources through the market is clearly a controversial issue, as are the different uses of water, with the exception of agricultural uses. These issues are defining features of the different interests at the table. The ANO\'A identifies each cluster as distinctly different from the others on one or more statements. Cluster 1, Proteding Con.mmpti11e Needr, is defined in part by its support for reallocating water to high growth areas .. Cluster 2, Proteding .1g1imlt11ml ]\}eedJ, is defined by many of the same statements as Cluster 1, with a focus on consumptive needs. However, this cluster does not agree with the statement on reallocating water to high growth areas and is defined by its difference from cluster one on this statement. Cluster 3, Balcmting j\'eedJ, is defined by its agreement on most statements related to balancing the different types of water uses, with a significantly higher interest in cooperation between basins than any other cluster. Cluster 4, Proteding Non-ConJ/Imptillf j\'eedJ, is defined by its interest in emironmental and recreational water needs, and by its lack of agreement on statements related to consumptive needs, including balancing supply demand. Finally, Cluster 5, Cumnl SyJkJJJ Broken, is defined by its lack of agreement \Vith Statements related tO the current legal S)'Stem that prioritizes the market and individual water rights. 148

PAGE 168

Demographics of Each Belief Coalition Proposition one explores the extent to \Vhich membership 111 each cluster, representing belief coalitions, aligns with demographics. l\luch of the public participation literature identified in Chapter Three focused on demographics as a means of deYeloping and measuring representation in a participatory process. Hmveyer, the policy process theory explored in this thesis (the A.d,,ocacy Coalition Framework), suggests otherwise, through a focus on interests and beliefs. To explore proposition one and determine whether the belief coalitions at the Colorado Basin Roundtable are identifiable through their demographics, the fin clusters deYeloped in the preYious section are analyzed against the demographics collected for all participants in the study. Tables 5.15 5.17 prmide information about the significant differences between the demographics of the clusters. The demographics chosen represent two types of demographics often mentioned in participation studies that have defined representation through demographics (gender and age). The demographic of owning water rights was selected as an example of a demographic specific to this policy issue that could be used as a means of defining representation. finally, the demographic of "affiliation" is included because it is the demographic included in the statutory mandate for the water roundtables as the method for selecting representatiYes for the roundtables throughout the state. The roundtables were selected based on affiliations, and were not intended to be representative of the population within each water basin. The analysis is not broken out by roundtable members ,-ersus external stakeholders, as the numbers within the 149

PAGE 169

roundtable arc low enough in the clusters to make it easy to identify specific individuals based on their demographics.1 Table 5.15. Age by cluster membership* Protecting Protecting Protecting Current All (n= Consumptive Agriculture Balancing NonSystem 154) Needs Needs Needs Consumptive Broken (11 = 32) (n = 19) (11 = .J.9) Needs (11 = 9) (11 = .J.5) ') 0 0 0 2 0 18-30 (1.3 o) (0.0'o) (0.0 o) (4.4%) (0.0%) 31-40 17 3 1 7 4 2 (11.0%) (9.4'o) (3.1%) (14.3'o) (8.9%) (22.2%) 41-50 41 7 5 14 14 1 (26.6o) (15.6/o) (31.1 %) (11.1%) 51-60 65 16 8 20 17 4 (42.2%) (50.0%) (25.0%) (40.81o) (37.8o) (44.4%) 61-70 19 4 3 4 7 1 (12.3o) (12.5'o) (9.4%) (8.21o) (15.6%) (11.1%) 10 ') 2 4 1 1 OYer 70 (6.5 o) (6.3%) (6.3%) (2.21o) (11.1%) The 11 of the total and the clusters does not match other tables due to a small number of respondents not completing all of their demographic survey questions. There is no significant relationship between the age of respondents and their membership in a belief cluster. Consequently, there is no reason to believe selecting a representative sample of the population by age to be members of a participatory process would result in any specific distribution across the belief clusters. 1 For example, if only two members of the Colorado Basin \\"ater Roundtable were her.veen the ages of 18 and 30 and both fell into the same cluster, it would reveal their identity and cluster affiliation to anyone familiar \\ith the roundtable members. 150

PAGE 170

Table 5.16. Gender by cluster membership* Protecting Protecting Protecting Current All (11= ConsumptiYe Agriculture Balancing NonSYstem 154) Needs Needs Needs ConsumptiYc Broken (11 = 32) (11 = 19) (n = .J-9) Needs (11 = 9) (11 = .J-5) 129 27 18 43 34 7 (83.8'o) (84.4%,) (94.7o) (87.8,o) _{75.6 o) (77.8 o) 25 5 1 6 11 .., Female (16.2%,) (15.6'o) (5.3'o) (12.2o) (24.4 o) _(22.2 u) The 11 of the total and the clusters does not match other tables due to a small number of respondents not completing all of their demographic suney questions. Similar to age, there is no significant relationship between the gender of respondents and their membership in a belief cluster. Though there is no significant relationship, potentially due to the small sample size and small number in each cell of the cross-tab, there is a potentially interesting descriptiH finding: the only belief cluster that does not ha,e 1 01o or more female respondents is Protedi11g Ag1imltm-al 1\.'eedJ. Table 5.17. Owns indiYidual water rights by cluster membership* Protecting Protecting Protecting Current All Balancing Non(n= Consumpti Agriculture Needs Consumptive System Ye Needs Needs Broken 154) (11 = 32) (11 = 19) (11 = .J-9) Needs (11 = 9) (11 = .J-5j_ Owns 37 Individual 7 9 10 9 .., Water (24.0 (21.9%) (47.4'o) (20.4o) (20.0'o) (22.2 u) Ri_ghts %) The 11 of the total and the clusters does not match other tables due to a small number of respondents not completing all of their demographic suncy questions. 151

PAGE 171

Ownership of indiYidual water rights also has no significant relationship to cluster membership, although descriptively, the percentages do show that the agricultural belief cluster has twice as high of a percentage of participants who own individual rights. Again, there are no statistically significant relationships for ownership of organizational water rights and belief cluster membership. Although these four variables align with the means used in public participation literature to define representation, there is no significant relationship to the belief clusters, suggesting a stratified sample of participants denloped based on these demographics would not ensure representation of the five belief coalitions. Table 5.18. Represents an organization who owns water rights by cluster membership* Protecting Protecting Protecting Current All Balancin Non(n= Consumptiv Agriculture g Needs Consumptiv System e Needs Needs Broken 152) (n = 31) (11 = 19) (n = .J-8) e Needs (11 = 9) (11 = .J-5) Organization 92 21 11 28 26 6 Owns Water (59.7 (67.7'o) (57.91o) (58.3%) (57.8o) (66.7%) Rights %) *" The 11 of the total and the clusters does not match other tables due to a small number of respondents not completing all of their demographic sun'ey questions. Compming Affiliation to Cluslen As can be seen in the tables above (fables 5.15 5.18), each of the fin belief clusters has members \Vho personally own \Vater rights, others who represent organizations \Vith \Vater rights, some with both, and some with no \Vater rights. Men and women are represented within each cluster, as are people of all age groups. 152

PAGE 172

The one area where demographics differ significantly across clusters is in the affiliations reported by cluster members. Respondents reported up to three affiliations from a list of over forty options.2 The options were grouped on the suney, allowing for an analysis that focuses on a smaller group of options while giving respondents more flexibility in how they answered. As Table 5.19 shows, respondents with affiliated with special districts, elected officials, government, and water industrv were distributed across the clusters with no significant difference found. The only affiliations \Vith statistically significant relationships to the cluster membership are agricultural and recreational/ emironmcntal. For the agricultural affiliation, a significant difference was found, indicating that respondents who have an agricultural affiliations arc more likely to fall into the Protedi11g Ag1imltmul NeedJ cluster (31.6o have agricultural affiliations) or, to a lesser extent, the 2 I :ach allowed to identify up to three from a list of -l4 To Cfl'pr.:cial district; Speci;ll Districts: Hur;ll \\'atn \X'ater Consultant: ( ;mernmcntal Other l'ril'ate Heslarch ln>titutl'/ ( )thcr l'ri\'lte. I l:nJ.,>incering I 'irm/ Con,;ultant: ( )rhcr l'ri\,ttc l.r.:gal 1-'irm/ < Consultant: Water Hcsource,;; ()titer l'ri\ate l:nterpriscs: Public L'tility. ( hher
PAGE 173

Pro!tding ConJIIIJJptite Seedr cluster (11.4% haYe agricultural affiliations), as compared to other clusters (X: = 14.25, df = 4, N = 160, p = 0.007). Howenr, the finding of significance is tempered by the fact that only 60'o of cells \Vere 5 or larger, which means the test of significance is too liberal. Specifically among the Colorado Basin Roundtable respondents, agricultural affiliations were represented in four of the five clusters total, excluding only the Gmc'lll S,;Jtw; Broken cluster. Table 5.19. Affiliations of sutTey respondents by their clusters. Protecting Protecting Protecting Current All Balancing Non-(n= Consumpti Agricultura All Needs Consumpti System ,.e Needs 1 Needs Broken 160) (n=35) (n=19) (n=SO) ,.e Needs (n=9) (n=47) Elected 17 ') 1 6 5 3 officials (10.8 (5.7'o) (5.3%) (12.0%) (10.6'o) (33.3%) 0 ) 58 13 5 23 13 4 GoYernment (36.3 (37.1 %) (26.3-'o) (46'o) (27.7%) (44.4%) %) Agricultural 15 4 6 3 2 0 (9.4'o *"' (11.41o) (31.6%) (6.0%) ( 4.3'o) (0.0%) ) Special 48 13 7 12 2 (30.0 14 (28.0%) district o;o) (37.1 %) (36.8%) (25.5%) (22.2%) EnYironment 37 1 2 9 21 4 / recreation*'* (23.1 (2.9 o) (1 0.5'o) (18.0%) (44.7%) (44.4%) o o) \Xlater 32 5 3 11 ') Industry (20.0 (14.3'o) (15.8%) 11 (22.0%) (23.4%) (22.2%) Oo) ** Statistically significant at the .005 level usrng a Chi-Square analysis. Among indiYiduals with recreational/ environmental affiliations, a significant difference found in their cluster membership (X: =25.11, df = 4, N = 160, p = 0.000), 154

PAGE 174

with 80% of the cells containing 5 or more cases, making the significance test more appropnate than with the agricultural affiliation. Almost half of the members of the Pmteding Non-Con.wlllptif.e NeedJ cluster han a recreational/ environmental affiliation (44.7%), and the same is true for the much small C11m111 5_yJtem Brokm cluster (44.t-t' o), while only one member of the Protecting ConJ11111ptil'f J"\'eedJ cluster has these affiliations (2.9%). On the Colorado Basin Roundtable, the 11 representatiYes with recreational or environmental affiliation \Vere distributed between only two clusters: Balancing 1\'eedr (n = 3) and Proteding Non-Consump!it'e NeedJ (n=8). The analysis of demographics by belief cluster membership has shown only nvo affiliations have statistically significant relationships to the belief clusters, and all of the remaining demographics do not have statistically significant relationships. Although the belief clusters were shmvn in the cluster analysis to have distinctly different means on the nine statements, their demographics are not similarly strongly divided. The lack of statistically significant findings may be a function of the low n. However, most clusters have members representing most affiliations, genders, age groups, and water rights ownership or lack thereof. Consequently, creating the participatory process by sampling based on any of these demographics would not ensure representation across all of the belief clusters. The one area with the greatest demographic alignment is in the affiliations of agriculture and recreational/ etwironmental. 155

PAGE 175

Relations Among and Between Belief Coalitions Proposition two suggests that different belief coalitions will han different leYels of shared action (e.g. information sharing), due to such things as differing levels of trust. To inYestigate this proposition, the next step in the analysis is the use of the net\vork data to understand the clusters, representing belief coalitions, in the Colorado Basin Roundtable, including the extent of interaction and trust among those with shared beliefs (within a cluster) and between each of the different clusters. Among Colorado Basin Roundtable members, all fin clusters are represented, but the numbers within each cluster are small. In order to exanune interactions bet\veen and within clusters, respondents were asked \vho in the water communitY they: (1) consider the most influentiai statewide; (2) exchange information with; (3) receive important information from; (4) belieYe most share their goals; (5) trust the most to keep their interests in mind, regardless of whether they share their goals; and (6) can depend on to follow through on a commitment. As described in the methods chapter, the analysis is based on means created by aYeraging the responses that members of each cluster reported regarding members of their own cluster and the remaining members of the roundtable. Each of the following tables represents one of the net\vork questions from the smTey. Each cell in the tables represents the aYerage response from members of the clusters to members of their own cluster or one of the remaining clusters. For each table, an ANOVA was used to identify whether any there was any significance difference bet\veen the averages and a 156

PAGE 176

Games Howell post hoc test was used to determine what differences were causing the significant findings from the ANO\'As. Esta?lishing whether a relationship between cluster membership and Yiews of members in other clusters exists is one means of exploring \vhether shared actions and trust exist within the belief coalitions. The goal of the analyses is to identify \vhether there is a statistically significant relationship bet\veen membership in a cluster and ,iews of other roundtable members by cluster. Network data were the basis of this analysis. As described in the prenous chapter, the network data \vere used as attribute data and the calculations resulted in each respondent haYing an average for their reported ties to all members in each of the fi,e clusters. For this reason, the independent Yariable in this analysis is respondents' membership in one of the fin clusters. The dependent 'ariable in the analysis is the respondent's aYerage responses on the network questions to members of each of the clusters, including their own. Other Roundtable Melllber.r Share Goa/.r As Table 5.20 shows, respondents from all fiye clusters haYe relati,ely neutral responses on the question of whether the roundtable members in each of the fiyc clusters share their goals for water policy and management, with most means between 0.10 and 0.40 on a scale of -2 (Strongly Disagree) to +2 (Strongly Agree). The only mean of 1.00 (/1gree) is from the members of the Protecting i\'on-Con.rttmptil1e i\'eer/.1 cluster in reference to others in their cluster. The only negatiYe mean is from the members of the Protecting j\'on-ConJ/1/Jlpti/ie NeedJ cluster in reference to the members of the Pmteding 157

PAGE 177

Con.mmpti;e I\'eed., cluster, suggesting the members of the Proteding Non-Con.rttmptiJ'e Need.r cluster belicn other members of the roundtable in their cluster share their goals, but the members of the roundtable in the Protecting ConJJf!llpti;;e j\'eed.r cluster do not share their goals. The Balallting j\'eed.r cluster has approximately the same response for its own cluster members as for most other cluster members, indicating the members of this cluster are no more likely to believe roundtable members in their cluster share their goals than roundtable members in other clusters. Table 5.20. 1herage Response on "extent to which you belie,e the following roundtable members agree with or disagree with your goals for the roundtable process" (Scale: = Strongly Disagree to +2 = Strongly .\gree) Protecting Consumptive Needs Average (n=9) Protecting Agricultural Needs Anrage (n=6) Balancing Needs Average (n= 12) Protecting Non-Consumptin Needs A,erage (n= 16) Current System Broken Axerage (n=2) Perceptions of roundtable members in the followin_g_ belief clusters: 0.46 0.35 0.35 0.19 0.44 0.36 0.12 0.42 0.40 -0.03 0.06 0.37 0.17 0.00 0.09 c u 0 > z '.;:::1 OJ) c.*"' c:: E "C '.;:::1 ::l ll) u "' I!J I!J c:: z ..... 0 8 u 0.00 0.07 0.47 1.00 0.13 0.17 0.17 0.05 0.34 0.00 ** Statistically significant ,-anauon at the 0.05 level 158

PAGE 178

The ANO\'A found that the members of all of the clusters excluding the Lt/171'11/ Sy.rlem Broken cluster nry significantly in their means for how they Yiew members of the Proleding Non-Comumplil'l: Necd.r cluster (Table 5.21), with members of the Son-Comumptit'l' J'\eedJ cluster significantly more likely than other roundtable members to report shared beliefs with members of tlus same cluster. Table 5.21. ANOYA results from responses on "extent to which you bclieYC the following roundtable members agree with or disagree \Vith your goals for the roundtable process" Dl MS F Significance All Roundtable l\Iembers AYerage 4 0.14 1.85 0.140 Protecting ConsumptiYe Needs Average 4 0.35 2.00 0.114 Protecting A.gricultural Needs AYerage 4 0.33 1.25 0.306 Balancing Needs iheragc 4 0.04 0.35 0.842 Protecting Non-Consumptive Needs A'-erage 4 1.91 10.07 0.000 Current System Broken Average 4 0.17 1.36 0.264 Frequenq E.\.dltlllge bl'luwn Ro11ndtable Memlm:1 When asked how frequently they exchange inf01mation, the members of the Proteding Con.wmptil'e l\'eedJ and Proteding j\1on-Co!Wt111ptite J\1eed.1 have their highest a\crage responses for the members of their own cluster, suggesting they arc most likely to exchange information with others who share their beliefs (Table 5.22). Members of the Ptvleding Con..-umpti!Je J\1eetf., cluster have the lowest average response for information sharing with the members of the j\'on-Con.mmptil'l' 1\1eedJ and Citm:nt .f)'Jiem Brokm clusters. Similarly, the members of the Proteding J\'on-Con.mmpti1e 1\1eed.f cluster 159

PAGE 179

ha,-e their lowest aYerage response for members of the Proteding Com11mpti1Je Needr and Cmn:nt 5_y.rtem Brok.m clusters. Both findings suggest that information exchange between clusters with strongly differing beliefs is less frequent than information exchange within clusters. The members of the Balancing Need.r cluster report almost the same leYel of information sharing with members of each of the other clusters. The members of the C11nV11/ 5_yJ/em Brok.m cluster report the least information exchange overall. The variation in how the members of the fiye clusters reported information sharing with the members of the Pro If ding lg1imltlfral j\'eed.r and Proteding l\'on-ConJumptit'e NeedJ clusters is statistically significant (Table 5.23). Table 5.22. iherage Response on "how often in the last year you han exchanged information related to water policy and management issues" (Scale: 0 = Never to 4 = Daily) Protecting ConsumptiYe Needs A \'erage (n=9) Protecting Agricultural Needs Anrage (n=6) Balancing Needs Anrage (11= 12) Protecting Non-Consumptin Needs .-\ \'erage (n= 16) Current System Broken AYerage (n=2) Perceptions of roundtable members in the following belief clusters: 0.88 0.69 0.82 0.49 0.22 0.78 1.00 1.06 0.87 0.33 1.40 1.49 1.46 1.42 1.09 0.92 1.02 1.14 1.58 0.75 0.28 0.00 0.25 0.10 0.00 ...-Statistically significant Yanatton at the 0.05 leYel 160

PAGE 180

Table 5.23. ,\NOVA results from responses on "hO\v often in the last year you han exchanged information related to water policy and management issues" Dl MS F Significance All Roundtable 1\Iembers Average 4 1.24 2.81 0.038 Protecting Consumptive Needs Anrage 4 0.83 1.56 0.203 Protecting Agricultural Needs Average 4 1.37 3.06 0.028 Balancing Needs ,\verage 4 0.92 2.08 0.102 Protecting Non-Consumptive Needs Average 4 2.59 5.77 0.001 Current System Broken A ,erage 4 1.35 2.03 0.108 11llp011tlll
PAGE 181

information receiYed from those who do share their beliefs. Overall, the members of Ba/a11ti11g I\eed.f cluster report relatively consistent aYerages across all clusters other than the Ctm-e11t 5_yJ/em Brokm. The Yariation in how the different clusters report the leYel of importance of information receind from the Proteding l'\o11-Cons11mpti1Je NeedJ cluster members is statistically significant (Table 5.25). Table 5.24. AYerage Response on "extent to which the information you receive from each roundtable member is important, that is, helps you achieYe your goals for water policy and management issues" (Scale: 0 =Not at all important to 4 =Very important) Protecting ConsumptiYe Needs AYerage (11=9) Protecting Agricultural Needs AYerage (11=6) Balancing Needs A nrage (n= 12) Protecting Non-Consumptin Needs A ,erage (n= 16) Current System Broken Average (n=2) Perceptions of roundtable members in the following belief clusters: 1.53 1.39 1.48 1.19 1.47 1.50 1.63 1.83 1.71 1.07 1.33 1.43 1.61 1.50 1.59 t: 0 > z p 0.0 c.. "t, c:: E '"0 p ::l U rn 1:) c:: z ...... 0 8u ....... 0.84 1.22 1.62 2.08 1.72 0.44 0.50 0.68 0.44 1.50 ** Statistically significant Yariation at the 0.05 level 162

PAGE 182

Table 5.25. ,-\NOV.-\ results from responses on "extent to which the information you recei,-e from each roundtable member is important, that is, helps you achieYe your goals for water policy and management issues" Dl MS F Significance All Roundtable Members A ,-erage 4 0.25 0.38 0.820 Protecting ConsumptiYe Needs Average 4 0.69 1.01 0.416 Protecting Agricultural Needs Anrage 4 0.45 0.59 0.671 Balancing Needs Anrage 4 0.14 0.22 0.926 Protecting Non-Consumptive Needs Average 4 2.43 2.76 0.041 Current System Broken Average 4 0.58 0.65 0.632 Tnul Ro11ndtab/e Member.r to Keep Intere.rl.f in Mind When asked the extent to which the\' would trust other members of the roundtable to keep their interests in mind, members of the Proteding -lgJimllllml Need.f and Prv/eding l\}on-Con.fllntplil'e Need.r clusters reported their highest lenl of trust for others in their own cluster (Table 5.26). Although their onrall means are lower, the members of the Proteding Con.rt!mptille I\'eed.r cluster also have their highest mean for members of their own cluster. The members of the Proleding ConJ111Jip!iJ'e ,'\1eer!J cluster ha,-e their lowest average response for members of the Proteding Non-Cowllmplil'l' J\rerlf cluster, and vice versa. These findings suggest that members of the Colorado Basin Roundtable are more likely to trust those who share their beliefs than others. The members of the Balmuing 1\'eed.r cluster han relatively little nriation in their average response for members of other clusters, other than a lower mean for C11rm1/ S.J'.flelll 163

PAGE 183

Brokm. n1e Yariation in roundtable members responses on the extent to which they trust the roundtable members in the Protecting /1gJit1tl!ttral l\1eedJ and Protedi11g l\'on-Con.mmptite I\'eed.r is statistically significant (fable 5.27). Table 5.26. Average Response on "Extent to which you would trust the following roundtable members to keep your interests in mind on water policy and management issues, regardless of whether he/she shares your interests" (Scale: 0 = Never to 4 = Completely) Perceptions of roundtable members in the followif!g_ belief clusters: '/. I E c;.l -o c:: QJ 0.0
PAGE 184

Table 5.27. ANOVA results from responses on "Extent to which you would trust the following roundtable members to keep your interests in mind on water policy and management issues, regardless of whether he/ she shares your interests" D/ MS All Roundtable l\Iembers AYerage 4 1.45 Protecting ConsumptiYe Needs Anrage 4 0.83 Protecting Agricultural Needs Average 4 3.04 Balancing Needs .Axerage 4 1.17 Protecting Non-Consumptin Needs Axerage 4 5.74 Current System Broken A,erage 4 1.55 Conjiden! Roundtable Memben will f-ollow tbroltgb Oil a Commit men/ F Significance 2.64 0.049 1.19 0.329 4.62 0.004 1.83 0.142 8.60 0.000 1.06 0.388 When asked how confident they were that other members of their roundtable would follow through on a commitment, the members of the Pro/eding. glimltural I\'eed.r, Balancing NeedJ, and Proleding I\'on-ConJ/fmptil't' 1\'eed.r clusters have their highest anrage responses for the members of their own cluster (Table 5.28). The members of the Proteding Con.rt!mp!ite Need.r cluster have their lmvest average response for members of the Proteding J\1on-Con.rumptil'e NeedJ and Cun-enl SyJieJJJ Broken clusters. Similarly, the members of the Proleding 1\'ou-Coll.fiiiJlp!il'e NeedJ cluster have their lowest average response for members of the Proleding Con.wmptiJJe j\,'eed. and Cmn:nl 5_y.rlem Broken clusters. The ,ariation in the means for how roundtable members responded on the confidence question in regards to the Proleding 1\'on-Co!lJIIIlljJiif'r J\'eedJ cluster members is statistically significant (Table 5.29). 165

PAGE 185

Table 5.28. Anrage Response on "confident each of the follmving roundtable members will follow through on a commitment" From each Cluster to Each Cluster (Scale: 0 = Not at all confident to 4 = \' ery confident) Perceptions of roundtable members in the belief clusters: rr; I E 101 ""0 = (i) o.c (i) c > (i) (i) .... c: ... z z p "' c: p c.. rr; B "' C..* u E "8 u -"'0 CD './),...:.:: B ::t ('.) E a = p ;:I (i) ..... 0 8 Cj z 0 ;:: z ...... u "' (i) c: ... u .... 0.0 c: 101 c: z 0 ..... .... 0 ... u ..... -;; 8u ;:I ::!:) u Protecting Consumptin Needs Average 1.54 1.50 1.59 0.95 0.56 (n=9) Protecting Agricultural Needs AYerage 1.73 2.43 2.04 1.59 0.75 (n=6) Balancing Needs ;herage 1.50 1.62 1.75 1.63 0.59 (n= 12) Protecting Non-ConsumptiYe Needs 1.10 1.21 1.47 2.48 0.84 Average (n= 16) Current System Broken AYerage 1.67 1.50 1.63 1.60 1.50 (n=2) *"' Statistically significant variation at the 0.05 lenl Table 5.29. ANO\' A results from responses on "confident each of the following roundtable members will follow through on a commitment" Dl MS F Significance All Roundtable i\Iembers Anrage 4 0.27 0.37 0.827 Protecting Consumptin Needs .\nrage 4 0.64 0.67 0.614 Protecting Agricultural Needs Average 4 1.66 2.02 0.111 Balancing Needs ;\yerage 4 0.39 0.56 0.691 Protecting Non-Consumptive Needs Average 4 3.66 3.90 0.009 D\'>Current System Broken .-\,erage 4 0.47 0.34 0.847 166

PAGE 186

!J!flllohe t?fEacb Rotmdtable Member in U:'ater Poliq and :' lanage!Jimt When asked how influential other members of the roundtable arc in \Vater policy and management issues statewide, the members of the Proteding ConJIIIJ/pti!e Seed.r, Protecting .' 1g1imltmul J\.'eedJ, and Proteding I\'on-ConJIIJJlplif'e j\'eed.r clusters haYe their highest average responses for the members of their O\Vn clusters. Roundtable members responses on how they view the statewide influence for the members of the .1g1icullllral J\!eed.r, Bulanting l\'eedJ and Protecting ConJII/Jlptille i\'eed.f clusters arc significantly different (Table 5.31 ). The members of the Proteding Con.wmpti!lt' l\'eedJ cluster han their lowest average response for members of the Proteding J\'on-Con.wmpti11e Xeer/.f and Cllnml SyJtem Broken clusters. Similarly, the members of the Proteding l\'on-Con.mlllptite I\!eerl.f cluster have their lowest aYerage response for members of the Protecting ConJIIIJlptire Need." and Cttmnt S)'Jtem Brokm clusters. There is more statistically significant nriation in how each cluster views the other clusters on the statewide influence question than on any of the pre,ious questions. This is partially due to the Protecting Consumptin Needs cluster reporting very low lenls of influence for all clusters, including their own, as compared to other clusters. 167

PAGE 187

Table 5.30. Average Response on "extent to which you consider each person listed belmv to be influential in water policy and management issues in the STATE OF COLORADO; that is, the people who seem to haYe pull, weight, or clout with others on these issues" (Scale: 0 = Not at all influential to 4 = Very influential) Perceptions of roundtable members in the following belief clusters: I E C) c::: C) OJJ "E OJ:;* c ;:;. CJ z ..., c.. Current System Broken Average 4 0.86 0.86 0.496 168

PAGE 188

In summary, across multiple questions 111 the sutYey, covenng information exchange and importance, trust, and perception of influence, roundtable members repeatedly reported more positive perceptions and greater amounts of interaction \\'ith those who share similar beliefs. The cluster with means that indicate the highest amount of within cluster interaction and trust is the Protecting j\'on-Con.mmptiJe ,\'eedJ cluster. The clusters that consistently showed low interaction and tmst with each other are the t\vo clusters with the greatest differences in beliefs: Proteding Con.wmptile i\'wiJ and i\'on-Con.mmp!il'e I'\eedJ. The Balancing I\'eedJ cluster consistently did not shmv great differences bet\veen how the members of the cluster interacted with and Yiewed others in their cluster versus those outside their cluster. Overall, there is some evidence for the accuracy of proposition t\vo's statement that there is a relationship bet\veen membership in a belief coalition and levels of interaction and trust. Gi,en the nry small number of roundtable members in the analysis (n = 47) and small size of the clusters on the roundtable (ranging from 2 -16 members in each), the frequency of significant findings (1/3 of the time) suggests that meaningful differences do exist bet\veen how members of clusters interact with and ,icw others with different beliefs and shared beliefs on the roundtable. Significant differences were found bet\veen how clusters viewed members of other clusters for at least one cluster on every question, and three clusters on one question (the influence question). 169

PAGE 189

Belief Coalitions and Perceptions of Success The final proposition explores the difference between representing interests and being representatin of interests. Two types of analysis were conducted to compare the fiye clusters to responses on the questions regarding the success of the roundtable process. Roundtable members were asked specifically about success of the Colorado Basin \Vater Roundtable, while remaining suryey respondents were asked about the success of the HB 1177 roundtable process as a whole. The two sets of questions were analyzed separately. Pero!ptioiiJ Colorado Ba.rin Roundtable Sut"t"e.r.r Colorado Basin Roundtable members were asked whether their roundtable has been successful at meeting their personal water management goals, their goals for the Colorado Basin Roundtable, and their oYerall statewide policy goals. On average, members reported that the roundtable has been less than moderately successful at meeting these goals (fable 5.32). Colorado Basin Roundtable members were also asked whether their roundtable has been successful at accomplishing a \'ariety of activities. For most questions, they responded on average that the roundtable has been only moderately successful at accomplishing the acti,ities. Roundtable members reported that the roundtable has had the least success at developing a needs assessment plan, fostering collaboration across the basins, and outreach to stakeholders outside of the roundtables. On the other hand, the most success at the time of the study was reported as deYeloping a grants process, 170

PAGE 190

educating its roundtable members, and fostering collaboration within the Colorado basin (Table 5.32). Table 5.32 shows the means for each belief cluster on each success question, including only the respondents on the Colorado Basin Roundtable. Table 5.33 shows the results of an ANOVA that found no significant interaction benveen belief cluster membership and responses on the success questions. The total number of roundtable members is too small to find any significance in the difference benveen belief clusters and their perceptions of the roundtable's success. Howenr, the Protecting Consumptin Needs Cluster did haYe means lower than the overall mean on all but one question, whether the CBRT is achieYing outcomes important to the HB 1177 process. 171

PAGE 191

Table 5.32. .1\lean Responses to the Success of the Colorado Basin Roundtable by Belief Clusters of Colorado Roundtable Basin Members (5 = Yery successful and 1 = not at all successful) "\chieYing your desired outcomes 2 8 6 2.44 3.00 2.82 3.06 3.00 (11 .J-1) ,\chieYing outcomes important for the 2 95 2 .44 3.00 3.27 3.00 3.00 CO Basin Roundtable (11-.Jl) AchieYing outcomes important to the 2 7 7 2.78 2.60 2.73 2.75 3.50 HB 1177 process (11 +3) Prioritizing key issues to address 3 .00 2.22 3.17 3.17 3.13 3.50 (11-.J5) DeYcloping a grants process 3.64 3.44 4.00 3.25 3.94 3.50 (11-.J5) Educating its members 3.89 3.56 4.17 4.00 3.88 4.00 (11-+5) DeYeloping needs assessment plans 2 _71 2.44 2.83 2.92 2.69 2.50 (11-.J5) Fostering collaboration within the basin 3 .77 3.22 3.40 4.00 4.06 3.50 (11 .J.J) Fostering collaboration across basins 2.77 2.33 3.00 2.83 2.80 3.50 (11 .J.J) Outreach outside of the Roundtables 67 2 36 80 3 00 2.57 2.33 2. (11 .J2) 172

PAGE 192

Table 5.33. Analysis of Variance for Responses to the Success of the Colorado Basin Roundtable Process as a Function of Belief Clusters Success Question Df MS F Significance Achie,ing your desired outcomes (11 = 4 0.60 1.02 .409 .J.J) Achieving outcomes important for the 4 0.88 1.45 .239 CO Basin Roundtable (n=.J2) Achie,ing outcomes important to the 4 0.31 0.41 .798 HB 11 77 process (11 = .J 3) Prioritizing key issues to address (11 = .J 5) 4 1.67 2.07 .104 Developing a grants process (11 = .J5) 4 1.10 1.47 .229 Educating its members (11 = .J 5) 4 0.41 0.98 .432 Developing needs assessment plans (11 = 4 0.33 0.75 .567 .J5) Fostering collaboration within the basin 4 1.38 1.79 .151 (11 = .J.J) Fostering collaboration across basins (11 = 4 0.79 1.01 .415 .J.J) Outreach outside of the Roundtables (11 4 0.50 0.83 .513 = .J2) PemptionJ of HB 1177 Proce.rJ S11ct:e.rJ The surny respondents who were identified by Colorado Basin Roundtable members, but are not part of that roundtable, were asked the same questions, but more broadly focused on the onrall HB 1177 process. This change was made to address the concern that many water stakeholders may not be aware of progress at a specific roundtable, but are more likely to be aware of progress across the roundtable process. Using an ANOV.\ with a Tukey HSD Post Hoc test, some significant differences were found between the clusters on some of the success questions. Table 5.34 shows the 173

PAGE 193

means for each belief cluster on each success question, including only the respondents external to the Colorado Basin Roundtable. Table 5.35 shows that there is a significant interaction between belief cluster membership and responses on the success questions for some of the questions. The Balanl'ing j\'eedJ respondents are the most likely to agree that the HB 1177 process has been successful at achieYing their indiYidual goals, organizational goals, their goals for the process, and completing most of the acti,,ities. In contrast, the Proteding Con.mmpti!'e j\'eed.r and Cm?-ent SyJ/em Broken clusters are the least likely to believe the process is achieYing their desired outcomes and their goals for the process. The Proteding Con.rlfmptil'e I\'eedJ Cluster was also the significantly less likely to belieYe the process has been successful at prioritizing key issues to address. Across the board, all belief clusters had means below three for success at outreach outside the roundtables, with Balaming 1\:eedx and Proteding Non-ConJ11111p1il'l' NeedJ slightly, but significantly, more positive in their mean response on this question. The statistically significant differences m perceptions of the different clusters, representing belief coalitions, suggests that stakeholders with beliefs focused on consumpttYe uses see less success in the roundtable process than stakeholders with beliefs focused on balancing uses and non-consumptiYe uses. 174

PAGE 194

Table 5.34. Mean Responses to the Success of the HB1177 Process by Belief Clusters of External Stakeholders (5 = Ycry successful and 1 = not at all successful) Achieving your desired outcomes (11-85_1** A.chie,ing outcomes important to the HB 1177 Process (11 = 99)** Prioritizing key issues to address (1195j_** oe,eloping a grants process (11911** Educating its members (1199)** Developing needs assessment plans (11 -92) Fostering collaboration within the basin (11 = 99_1 Fostering collaboration across basins _{11 = 92) Outreach outside of the Roundtables (1195)** II c:: '-" 2.74 2.71 2.85 3.51 3.54 2.85 3.74 2.53 2.34 2.13 2.44 2.26 2.83 2.35 2.70 3.44 3.20 3.32 3.64 2.89 2.89 3.62 3.75 2.10 2.82 1.89 2.20 3.25 2.84 3.12 2.67 3.23 2.93 3.84 3.28 3.89 I 3.27 3.00 2.77 3.94 3.69 2.59 2.63 2.61 2.61 E (,J ... '.1: r-:,..., E) 'J;...:<: ... c ;..... (,J,..,... .... ....... 2.14 2.00 2.40 3.43 3.25 2.17 3.00 2.80 2.00 ** Stgntficant dtfference at the .OS level between the means of the fi\c clusters on the statement, using an ANO\'A. 175

PAGE 195

Table 5.35. Analysis of Yariance for Responses to the Success of the HB1177 Process as a Function of Belief Clusters Success Question D/ MS F Significance Achie,ing your desired outcomes 4 4.21 5.49 .001 (n = 85)** Achie,ing outcomes important to the HB 4 3.30 4.38 .003 1177 Process (11 = 99)** Prioritizing key issues to address 4 2.70 3.62 .009 (11 = 95)** De,eloping a grants process (11 = 91 )** 4 1.44 2.52 .047 Educating its members (n = 99)** 4 2.00 3.23 .016 Developing needs assessment plans 4 0.94 1.56 .192 (11 :::: 92) Fostering collaboration \Vithin the basin (11 4 1.14 1.56 .192 :::: 99) Fostering collaboration across basins 4 1.33 1.71 .154 (11 :::: 92) Outreach outside of the Roundtables 4 2.23 3.27 .015 (11 :::: 9 5)** ** Stgmficant dtfference at the .05 le,el between the means of the five clusters on the statement, using an ANOYA. Exploring the Clusters The summary below re,-iews what the analysis revealed about each cluster and examines the type of belief coalition that the cluster might represent, using the typologies proposed by Kim and Roh (2008) and the concept of an advocacy coalition as defined by Sabatier and Jenkins-Smith (1993). In addition to interpreting the four types of analysis previously presented in this chapter, the discussion also includes information gathered through the participatory research process when stakeholders had an opportunity to share their thoughts on how to interpret the findings. As explained in the 176

PAGE 196

prenous chapter, the results of the study were presented to stakeholders who participated in designing the study as well as other stakeholders in the \Vater policy and management community who reviewed the draft report and provided insight into the interpretations. Finally, the results were presented to the Colorado Basin \Vater Roundtable. In addition to stakeholders responding that the findings felt intuitively correct, the stakeholder feedback included specific insights about the different clusters. Protecting Coll.tlllllptite j\"eed.t One of the statistically significant differences between the Coll.flflllptil't' NeedJ cluster and all other clusters was its support of the statement, "HB 1177 should support water transfers to high growth areas/sectors" (p < .05). Other defining statements included its statistically significant lack of agreement on the three statements related to non-consumptive needs (environmental and recreational) (p < .05). The Protecting CollmJJJptille i\'t'edi cluster also joined only one other cluster in support of the statement, "HB 1177 should support allocation and management of water resources through the market." Finally, this cluster agreed on all of the statements related to consumptive uses. The name, Protedi11g Cou.mmptil'e Need., was chosen based on this high support for consumpti"e uses, lack of support for non-consumptive uses, and unique support for transfers to high grO\vth areas/ sectors. l\1embers of this cluster statistically significantly less likely to han cmironmental/recreational interests (only one member reported this affiliation), and statistically significantly more likelv to have an agricultural affiliation (p < .005). 177

PAGE 197

On the Colorado Basin Roundtable, members of the Protecting ConJIImp!itJe NeedJ cluster reported a higher perception of shared goals for water policy and management with the other members of their cluster than with members of the other four clusters. They reported the least agreement on goals with the members of the Protecting Non ConsumptiYe Needs cluster. They also reported more information exchange and more Yaluc from the information received within their cluster than outside it. However, overall, they reported Yery little information exchange with anyone on the roundtable, and enn among members of their own cluster, they reported exchanging information less than monthly on average. n1ey trust members of their own cluster to keep their interests in mind more than members of the other clusters, however their a\erages for trust are low and on a scale of Jlel'el; rare!J', in limited Jilllatio!IJ, qjien, and completely, their means for their own cluster arc only slightly above rarely. Finally, they consider members of their own cluster to be more influential in water policy and management then members of other clusters, although overall they do not see members of the roundtable as being particularly influential, \vith most indicated as not influential at all. They consistently report the least information sharing, trust, shared beliefs, and perception of influence for the members of the Protecting Non-Consumptin Needs cluster, and many of these findings were significant as compared to how other clusters viewed the members of the Protecting Non-Consumptive Needs cluster. Although the total number of roundtable members was too small to find any significance between clusters on thei.J. perceptions of the roundtable's success, the 178

PAGE 198

Protecting Consumpti,e Needs Cluster did have means 10\ver than the merall mean on all but one question. "\dditionally, the ConJillllpti!e l\'eedJ CllfJier members outside the roundtable were significantly less likely to belie,e the HB 1177 process has been successful at prioritizing key issues to address. \Xlhen the findings \vere shared with stakeholders, their discussions and written comments in response suggested that the Pro!eding Con.mlllpti!e j\,'l'l'dJ cluster members, with their focus on municipal needs and transfers to high growth sectors, may represent those water stakeholders \Vho have the least need for the water roundtable process. Larger municipalities often have both the resources to secure water rights in the pri,ate market and the legal right to use the water resources they already have due to the legal structure of individual water rights. Their limited engagement through information sharing and perceptions of success may be hec:ause they do not need to be engaged in the water roundtable process in order to meet their interests. Kim and Roh (2008) explored different types of coalitions and cooperauon among policy actors based on shared policy interests, shared action, trust, and resources. The Conmmptil'e l\:eedJ cluster members on the roundtable percei,c other members of their cluster as somewhat sharing similar beliefs, more so than members of the other clusters. They are engaged in some actions together through information sharing, more so than with members of the other clusters, but not at a high frequency. They are somewhat trusting of each other, though only in comparison to their lack of trust in the Protecting l\'on-CoiiJIIJJJpti!Je NeedJ and C11nrnt S_plem Broken cluster members. 179

PAGE 199

They are as much defined, in terms of trust and information sharing, by their lack of both, with the members of the Colorado Basin Roundtable who represent protecting non-consumpti,c needs, as by their presence of both among their cluster members. Based on Kim and Roh's (2008) typology, the Protecting Consumptin Needs cluster is best described as a cluster of Potential Collaboraton, due to the low amount of information exchange. With only a low or medium level of trust (higher than with other groups, but not significantly so), according to Kim and Roh's (2008) typology, they are capable of medium cooperation at best, prmided they han sufficient resources for cooperation. As a coalition, they may ha,e less investment in the roundtable process both due to their lack of belief that the process is succeeding at its many goals and actiYities and due to their access to other means, the private market, to meet their water needs. Protedi11g Ag1icultllral J\'eedx One of the statistically significant differences between the Protedi11g AglimltmtJ! 1\'eed. cluster and the Protedi11g CollJIIIJJplille NeedJ cluster was the opposing Yiews on the statement, "HB 1177 should support water transfers to high grmvth areas/ sectors" (p < .05). In addition to opposing this statement, which is focused on growth, often a municipal concept, the Protrding / lgJimllura/ J\'eed.r cluster has the highest level of agreement on the statements, "HB 1177 should support protecting the agricultural economy and way of life" and "HB 1177 should support protecting existing individual water rights." This cluster was neutral on the statement on supporting allocation and management of water resources through the market and negative on all three statements 180

PAGE 200

that support non-consumptiYe needs. The members of the Protecting 1\'eedJ cluster are almost exclusiYely male and 47.4'o own indi,idual water rights, as opposed to 24.0/o overall. They are statistically significantly more likely to have agricultural affiliations and less likely to have environmental or recreational affiliations (p < .005). On the Colorado Basin Roundtable, members of the Proteding / Igric11ltllral J\'eedJ cluster reported a higher perception of shared goals for water policy and management with the other members of their cluster than with the members of the other four clusters. They are engaged in mo11tbb' information sharing on average with members of their own cluster and the Balancing 1\'eedJ cluster as well. Despite this limited information sharing, the members of the cluster han a statistically significant higher level of trust that fellow members of the roundtable would keep their interests in mind, particularly those members in their own cluster (an anrage that falls benveen and i11 limited .,-itlfaliollJ) and the Balancing NeedJ cluster (an aYerage near i11 limited Jilllalio/1.1). Similarly, they are Jomewbat t'OI!fident members of their own cluster will follow through on a commitment, as will members of the Balanci11g i\1eedJ cluster. !\!embers of this cluster do see each other as injlmntial in water policy and management issues, and also sec the Balancing NeedJ cluster members as JOllleJJJbat Similar to the Protedi11g C(nuumptil!f! J\'eedJ cluster, members of the Agrimi!Jtral NeedJ cluster ha\'e less positive perceptions of the HB 1177 process in terms of whether it is meeting their desired outcomes and developing priority issues to address. Howe,er, they see the roundtable process as having success at educating membus and 181

PAGE 201

fostering collaboration within the basin. They are the most likely to report the roundtable process is fostering collaboration across basins, though not at a statistically significant leYel. Stakeholders renewmg the findings of the study noted that the Protediug .lgli(Jfltto'tli I\'eed.r cluster members haYe historically benefited from the market-based, indiYidual rights structure of water policy and management. However, small agricultural interests haYe less fmancial influence in the market-based water management system than the larger municipalities. Agricultural interests in the past have not needed to form coalitions as they haYe operated through the market-based structure, thus one would not expect them to han high leYels of information sharing with each other. Their engagement in water roundtable process might be an attempt to fmd a new way to have a Yoice in water policy, particularly in a context where agricultural lands are "drying up" as rights are sold to meet municipal needs. Using Kim and Roh's (2008) typologies, this belief coalition might best be described as Potential Collaborators, due to little information sharing, but high levels of trust. The potential to coordinate may be just as high with members of their own coalition as with members of the Needs coalition, ginn how often they report similar perceptions across the two clusters. As a coalition, they do see some success in the roundtable process, though not necessarily for their own goals at this point. Ho,,eyer theY are not Yery acti,e in sharing information with members on the roundtable, so they may not be as actively engaged as others at the roundtable. 182

PAGE 202

Balan,ing NeedJ The Balandng 1\'eedJ cluster is defined by its lack of disagreement with any of the nine statements and statistically significant and strongest agreement of all clusters on the statement, "HB 1177 should support increasing cooperation among water basins" (p < .05). .l\Iembers of this cluster agree strongly with statements associated with consumptiYe needs including balancing water supplies and demands and protecting individual water rights. They are equally strong in their support for the HB 1177 process balancing the water demands between uses and non-consumptiYe uses. They are one of only two cluster who support allocation and management of \Vater resources through the market, the current legal structure. The members of this cluster han demographics almost identical to the overall demographics of the 164 respondents to the survey. The one exception, which is not statistically significant. is the higher percentage of government affiliations (461o vs. 36.3/o onrall). l\Iembers of the Balalltillg J\'eedJ cluster on the Colorado Basin Roundtable reported ,ery similar Ienis of agreement that members of their own cluster, the Proteding /lgliml!ttral J\'eedJ cluster, and the Proteding J\'eedJ cluster share their goals for water policy and management. They reported lower levels of agreement with the Pro/eding ConJNmp!itJe I\'eedJ and C!llnn/ Broken clusters. Members of this cluster reported similar leYels of information sharing across all of the clusters on the roundtable, with an average report of between molltbb and JJJeekfy information sharing. Thev were also most likely of any cluster to report the information they received from other 183

PAGE 203

roundtable members was important. Members of this cluster reported they could trust members of their own cluster and all other clusters except for Cunr?nt 5_y.rtem Brokm to keep their interests in mind in limited Jit11atio11J and were .romewbat amjident most roundtable members, other than the C11rrmt S.YJtem Broken cluster members, would follow through on a commitment. They reported that members of Pmteding Ag1icultmul Need.r were .romewbat il(j/!lentia/, as were members of their O\Vn cluster, though with a slightly lower aYerage. They reported members of Proteding ConJifmptitJe Need.r and Protecting N.on-Con.rumptitJe ;\'mf.r as slightly less influential and members of the Cunmt 5_y.rtem Broken cluster as the least influential. On seYen of the nine questions asking members of this cluster outside of the Colorado Basin Roundtable about the success of the HB 1177 process, they haYe the highest perception of success of any cluster, including a significantly higher perception of success for whether the roundtable is achieving their desired outcomes, achieving outcomes important to the process, prioritizing key issues to address, and educating its members (p < .05). The presentation on the Balancing Need.r cluster resulted in substantial discussion at the Colorado Basin Roundtable meeting where stakeholders explored the results of this study. The existence of a group of stakeholders who put balancing needs ahead of specific types of needs was noted by one stakeholder as an invaluable opportunity for the roundtable process. The common theme among stakeholders responding to this cluster was that it was unlikely the members of the cluster all equally supported both 184

PAGE 204

consumptive and non-consumptive needs and more likely that their approach to meeting the need they prioritized was based in the recognition that their nt:eds \Vould not be met unless a compromise with the conflicting needs \Vas created. Kim and Roh's (2008) typologies ha'T more ability to account for this cluster than the Advocacy Coalition Framework structure that expects policy actors to be primarily engaged in shared actions with others who share their beliefs (Sabaticr & Jenkins-Smith, 1993). of this cluster appear to be supportive of all interests, perhaps from a nry pragmatic standpoint of needing to collaborate with other interests to get their needs met. They are somewhat trusting of members whose priorities arc consumpti,-e as well as those who prioritize non-consumptin, and more highly engaged in the roundtable process than other clusters \Vith information sharing across clusters and a perception that the process is achieving successes. They are clearly substantial collaborators within their own cluster, gi,-en the existence of shared beliefs and information sharing together. However, they are collaborating with more than just the members of their own cluster. At the same time as being substantial collaborators in their own cluster, they are reciprocal collaborators with members of other clusters (collaboration without necessarily sharing policy beliefs). Proteding l\'on-CoJWtllJj>lil'e :.\'eedJ The Proteding 1'\on-CollJtflJlptir;e j\'eed.r cluster is defined by its statistically significantly stronger support for the HB 1177 supporting the protection of the environment and recreational water needs, as well as its strong support for balancing 185

PAGE 205

water demands between consumptin and non-consumptin uses (p < .05). To a lesser extent, this cluster is also defined by its lack of support for the idea that the HB 1177 process should support allocation and management of \Vater resources through the market. tvlembers of this cluster are more likely to be female, though not significantly, with 24.4% female members compared to overall. The cluster is significantly more likely to haYe members with environmental or recreational affiliations than the previous three clusters, with 44.7% having these affiliations compared to only 23.1% overall (p < .005). I\Iembers of this cluster had statistically significant agreement that others in their cluster shared their goals for the roundtable process (p < .05). This is the only cluster to have an average of 1.00 (equal to agree on the scale). They reported an a\erage of -0.03 on the shared goals statement with the members of the Proteding CoJwtmptiiJe NeedJ cluster. 1\Iembers of this cluster had the highest mean for within cluster information sharing of any cluster, bet\veen month()' and weekb and also reported receiving more important information from each other than any other cluster. Unlike the Balandng Need.r cluster, they reported less information sharing with other clusters, particularly the Proteding CoiiJIIIJJpti!e j\:eed.r and C11n-ent 5_wtem Brokw cluster. The members of this cluster report that other members of their cluster can be uusted to keep their interests in mind and they are t"OJ(jidmt other members of their cluster will follow through on a commitment. These are the highest levels of trust among any cluster and the difference benvecn how theY ,iew members of their cluster and how other clusters vie\v them is 186

PAGE 206

statistically significant. Finally, they perceiYe each other as being someUJbal il!f/uenlial in water policy and management, with lower a\erages on the influence question for their perception of members of all of the other clusters. Members of this cluster both on the roundtable and outside the roundtable had relatiYely aYerage responses on most success statements, falling between the ConJttmptil'e !\'eedJ cluster's low perception of success and the Ba!miLing ;\'eer/.r cluster's higher perception of success. HoweYer, they did report one of the higher perceptions of success on whether the roundtable has been successful at outreach outside of the roundtables. The ;"'\'on-ConsumptiJ'e j\'edr cluster members' high leYel of informacion sharing and trust was not surprising to the water stakeholders who reYiewed the study. Stakeholders explained that emironmental interests haYe largely depended on coalition building strategies to achieYe policy wins in water policy, due to the indi,-idual rights and market-based structure historically leaving no place for non-consumptiYe needs. This aligns with a previous study's findings that non-consumptive users in Colorado achieYe their goals not through legal precedence (as the Colorado water law disenfranchises nonconsumptin interests), but rather through collaboration (Hatmaker, 2008). The roundtable opportunity is a unique situation for the non-consumptive interests, as the process has prmided them with new access and legitimacy. Stakeholders reYiewing the study were surprised by how many of their water policy colleagues ha,-e enYironmental/recreational goals for the water roundtable process. The Colorado Basin 187

PAGE 207

Roundtable officially has only one emironmental/ recreational representative, but this cluster was the largest on the roundtable. Using l(im and Roh's (2008) typologies, the Protecting Non-Consumptive Needs members are Sub.l"tantial Collaboralol). with both shared beliefs and active information sharing. They have high levels of trust, suggesting that if they have resources available, they would be the only cluster so far to be able to achieve high cooperation. Of any of the clusters, they most represent the ad,ocacy coalition model, with a combination of shared beliefs and shared actions. They are sharing information with others outside their cluster, but not as frequently as within, and they do have a somewhat positive report of success for the roundtable and HB 1177 process, particularly as relates to outreach, suggested they are at least moderately engaged in the water roundtable process. CunT!nt S_)'Stem Broken The Cunrnt 5_yJtm; Broken cluster is the smallest, with only nine members, two of whom are on the Colorado Basin Roundtable. For this reason, some of the findings, particularly the network data which is limited to the Colorado Basin Roundtable members, may not be worth analyzing. Though small, the cluster has distinct characteristics. This cluster is defined by its statistically significant disagreement with three statements that describe how the current water policy and management system is run: individual water rights, market-based allocation, and transfers to high growth sectors (p < .05). This is also the only cluster not to agree that the HB 1177 should support the agricultural economy and way of life. The cluster is supportive of non-consumptin 188

PAGE 208

uses, and this support is statistically significantly different from the lack of support for non-consumptive uses found in the Proteding CoJIJ'IfiJ/pliJe 1\'eed..and Protedi1zg tgriotll11ml NeedJ clusters (p < .05). The cluster has no agricultural members, compared to 9.4 o overall, and 44.4% of its members ha,e emironmental/ recreational affiliations, compared to 23.1% overall. On the roundtable, the two members of this cluster reported some of the 10\vcst levels of shared beliefs, information sharing, information importance, trust, and perceptions of influence. Howenr, with only two members, it may not be appropriate to consider these results. For this reason, the network data are not explored in this section. It is important to note that the low membership on the roundtable and c,en among the total survey participants does not necessarily suggest a small number of people in the state hold these beliefs. Rather, it suggests a small number of people directly imohed or connected to someone imolved with the roundtables hold these beliefs. The seven members of this cluster outside of the Colorado Basin Roundtable are similar to the Proledillg CollJIIIJip!ilte NeedJ cluster members in their relati,ely low, though not significantly low, perceptions of the HB 1177 success at achieving its goals and undertaking its activities. Among the stakeholders who revie\ved the results of the study, less discussion occurred around the Current System Broken cluster, perhaps because of its lm, numbers. The main point made by stakeholders with regards to this cluster is its lack of 189

PAGE 209

realistic perspectiYc, as the market model and indiYidual water rights are supported by the language of HB 1177 that initiated the \Vater roundtable process. Using Kim and Roh's (2008) typologies, this small cluster of water stakeholders is neither trusting of one another nor likely to interact, howeYer the numbers are too small to interpret meaningfully. If this pattern were to hold true \vith larger numbers, they would be potential collaborators at best, with shared beliefs, but no shared actions. Their lack of perception of success in the roundtable and HB 1177 process suggests they are not very engaged in the participatory process. Conclusion The belief coalitions denloped using k-means cluster analysis represent five distinctly different sets of goals for water policy and management. The coalitions, as interpreted with stakeholders, represent interests that they recognize from their cxpenences in water policy and management. The stakeholders identified concrete, historicallY and legally-based reasons for the different levels of interaction and engagement, and helped to explain why different belief coalitions participate in the roundtable process. From this analysis and interpretation, a picture of the Colorado Basin Roundtable process emerges, with more stakeholders invested in non-consumptin issues than affiliations would suggest, and an unexpected group of stakeholders focused on balancing the many needs. The implications for the findings on these five belief coalition are explored in the following chapter. 190

PAGE 210

CHAPTER 6 This dissertation studied a single participatory process, the Colorado Basin Roundtable, through the context of the representatives participating in the process, The propositions presented in the literature review were analyzed using data prmided by participants of the roundtable and other stakeholders imoh-ed in Colorado water policy and management. E,en with very small sample sizes, the analysis found significant findings, which can now be explored in more depth through the context of the three propositions. The cluster analysis of the nine statements asking about priorities for the water roundtable process resulted in five clusters composed of individuals who shared priorities for the process. The nine priority statements covered the major competing interests 111 Colorado water policy (municipal, agricultural, emironmcntal, and recreational), the legal structure of water policy and management (individual rights and the market), the priorities of the roundtable process in statute (cooperation between basins and interbasin transfers), and opportunities for compromise (balancing supply and demand; balancing consumptive and non-consumptive). The five clusters of values are similar to what previous studies in the water community ha,e found. L. sing a much 191

PAGE 211

more complex set of questions, the Colorado Institute of Public Policy (2006) identified a group of respondents whose defining feature was dissatisfaction with the current system, similar to the Current System Broken cluster in this study. The report also found distinctions between those focused on consumpti,e uses and non-consumpti,e uses. The similarity in the findings suggests the clusters are a legitimate and accurate way to describe the nlues of stakeholders in the water community. Concrete differences exist across water stakeholders, but just as importantly, clear agreement exists between many of the stakeholders, particularly their support for cooperation across and balancing water supplies and demands. When the demographics of cluster members were analyzed, each of the clusters in this study \Vas found to haYe diYerse membership, not limited to stakeholders whose affiliations were the most obYious for inclusion in the cluster (e.g. only half of the Proteding /lglimllural 1"'\'eed.r cluster members han agricultural affiliations). Each cluster was represented in the Colorado Basin Roundtable; among the members on the roundtable, the clusters had different ayerage perceptions of the members of the other clusters. l\Iost clusters' members reported more interaction and positiYe perceptions of those in their cluster than those outside their cluster. Finally, the different clusters have different perceptions of the success of the roundtable process. Examining the Propositions The proposmons presented in chapter three were intended as a beginning of exploring hO\v to bring together the literature on public participation, the importance of 192

PAGE 212

representation, and the need for a fair and competent process with the policy literature on the role of interests and beliefs in policy change. The fiye clusters are summarized in Table 6.1 within the context of these three propositions, but a more thorough discussion is needed. Table 6.1 (p.17) is based on the analyses presented and discussed in Chapter FiYe, and focuses on the three key issues conred in the propositions: (a) Alignment between demographics and belief clusters. The table focuses specifically on affiliations as this was the only areas where statistical significance was found. Additionally, the statute that created the water roundtable process required selection of members representing specific "interests" such as these affiliations. for example, the state requires the appointment of one person who "shall represent environmental interests and who shall be selected from nominees submitted by one or more regionally, state-\vide, or nationally recognized environmental conservation organizations that have operated in Colorado for at least fin years" (37-75-104(4)a(I\'), C.R.S.). (b) Le,,el of cooperation within the cluster and with other clusters. Kim and Roh's (2008) descriptions of different types of collaboration and cooperation are the basis of this portion of the table. They define substantial collaborators as haYing common beliefs and engaged in action together, likely as a result of haYing both trust and resources. Potential collaborators haYe common beliefs but are not engaged in action together, potentially due to a lack of trust or resources. Reciprocal collaborators do not share beliefs, but are engaged in shared action regardless, suggesting some trust. 193

PAGE 213

Table 6.1. Summary of the FiYe Clusters Affiliations (Proposition 1), Level of Cooperation within the Cluster (Proposition 2), and Engagement in the Participatory Process (Proposition 3)* (1) Affiliations (2) Len! of Cooperation within Cluster Protecting Potential Collaborators ConsumptiYe emironmental (within cluster) Needs I recreational -trust information sharing + agricultural Potential Collaborators Protecting ('vithin cluster and with Agricultural -Balancing Needs cluster) Needs environmental +trust I recrea tiona! information sharing Substantial Collaborators (within cluster) +trust + goYernment + information sharing Bal::tncing Needs (not significant) Reciprocal Collaborators (outside cluster) +trust + information sharing Protecting + Substantial Collaborators NonenYironmen tal (within cluster) Consumptive I recrea tiona! +trust Needs + information sharing Current + SYstem environmental liiJtt.flitimt mtmbmjor cma!yJis Broken I recreational *The symbols represent:"-" = lack of, negauve percepuon "+" = Presence of, positive perception (3) Engagement in Process Low Engagement information sharing outside cluster perceptions of success Medium Engagement -information sharing outside cluster perceptions of success High Engagement + information sharing outside cluster + perceptions of success Medium Engagement information sharing outside cluster perceptions of success Low Engagement perceptions of success "-" = Some cYidence of, neither positive nor negatiYe perception (c) Engagement in the process. l'sing two types of data, each cluster is given a low, medium, or high leYel of engagement in the process. First, their level of information sharing across all clusters, as this indicates interaction \Vith members of the 194

PAGE 214

roundtable \Vho have different pnonues for the roundtable process. Second, their perception of the success of the process, that is, whether or not it is succeeding at meeting their goals, the goals set out for it in legislation, and the goals associated with activities currentlv underwav at the Colorado Basin Roundtable. E.'plo1ing PropoJiliou One Participatory processes are composed of individuals who represent policy interests, also knmvn as advocacy coalitions, and these coalitions are defined by shared beliefs rather than demographic characteristics. Proposition 1 combines the concept of the Coalition Framework's (ACF) belief coalitions, which undertake shared action to influence policy decisions (Sabatier and Jenkins-Smith, 1993) with the context of a participatory process where representatives of demographics or interests are brought together to make shared decisions. ACF defines these coalitions as a function of coalition members' beliefs, rather than members' demographics. If traditional participatory processes usc demographics to define participation, as is found not only in the literature, but also in the statute that created the roundtable, than the distribution of beliefs may or may not match the intended distribution. Participants in the Colorado Basin Roundtable represented all fiye of the clusters of beliefs found among water stakeholders in the study. Although some demographics were distributed across the clusters in a manner that was statistically significant, the majority of demographics were not. In fact, the demographics emphasized in much of the public participation literature, such things as age and gender, have nothing to do with 195

PAGE 215

the interests represented by the stakeholders. EYen a demographic that is specific to this policy arena, water rights ownership, only helped define one of the clusters, Proteding J\'eed1, and then not significantly. The affiliations that were the basis of the legislation creating the roundtables also did not accurately reflect the interests at the table, as according to the legislation, only one of the appointed members represents em-ironmental/recreational interests in the Colorado Basin Roundtable; the study found that 16 members of the cluster most supportive of enYironmental and recreational interests, Pro/eding 1\'on-ConJI!Illplil'e NeedJ cluster, sit on the roundtable. The proposition appears to be relatinly accurate: the participatory process of the Colorado Basin Roundtable was composed of distinctly different clusters whose beliefs distinguished them and those clusters could not haYe been identified through demographics alone. If proposition one is accurate in the context of this thesis, it presents some challenges to the public participation literature. First, if a participatory process is intended to be representatiYe, the question needs to be asked as to what defines represcntatin. Second, if a participatory process is intended to be a fair process, this presents a significant problem. If faimeH is defined as a process that includes all "indi,iduals or groups whose interest or Yalues may be affected by the problem or proposed decision action" (Wehler, 1995, p. 52), how can those indi,iduals be identified and brought into the process if their interests cannot be easily identified through demographics or affiliations? 196

PAGE 216

E.-.:amining Propo.rition Two The coalitions represented in a policy process will engage in shared actions, as seen by their shared information and trust in each other. The members of the fin coalitions imoh-ed in the process were shown to have had Yarying leYels nf shared action. l'nlike the definition of an adn>eacy coalition, where the assumption is made that shared beliefs (core, policy, and secondary) lead to shared actions (Sabatier & Jenkins-Smith, 1997), the clusters at the Colorado Basin Roundtable nried, with some ha,ing higher leYels of information sharing within their cluster (Proleding j\'oi!-Coi!JIImptil'e I\'eedJ), some ha,ing very low Ienis of information sharing \Vithin their cluster and across clusters Col!.rlflllp!il'e !'\red.,. Proteding A,gliatl!!lral I\'eed.,), and one cluster sharing information across all of the clusters (Balcmdng Need.1). Similarly, when questions on trust were asked, almost the same pattern emerged. The conclusion to these findings is that shared interests at the Colorado Basin Roundtable have not led to coalitions of actors who primarily interact with each other. The exception is the Proleding Non-Com11mptil'i' NeedJ cluster, \vhose leYel of information sharing and trust with members of their mvn cluster is unique among the clusters. The fact that many cluster members in this study are not interacting with and trusting only their own cluster members, those who share their goals for the water roundtable process, is to the benefit of the participatory process, as it suggests the interest groups arc not operating at odds with each other, at least at the point in the roundtable process that was studied (18 months into their participatory process). The presence of interaction among 197

PAGE 217

Bala11(i11g 1\'eedJ members with all of the other cluster members also suggests that shared actions can come from something other than shared beliefs. Although most of the coalitions do not resemble advocacy coalitions, the Proledi11g J\'on-Con.w/J/ptil'e 1\'eedJ cluster is reflective of the Advocacy Coalition framework's assumption of shared beliefs leading to shared action. The stakeholders' interpretation of the cluster may be the best indicator as to why this cluster functions more like ACF would predict than any other cluster: their lack of legal mechanisms for meeting their needs has driven them to utilize collaboration to push their agenda (Hatmaker, 2008). This historical and legal precedent for how the belief cluster functions leads to the question of whether ACF, Kim and Roh (2008), or public participation theorists have neglected potential drivers of levels of shared action and interaction: historical precedent and legal necessity. E:,:amining PropoJilion Three In a participatory process, "representation of interests" may not equate to "representing interests" if the stakeholders are not equally engaged in the process. This proposition is the hardest to explore in this study. "Engagement" in the process can be seen through the extent to which cluster members are sharing information with roundtable members inside and outside their clusters, but also through the cluster members' perceptions of how successful the process is to date. Based on these two types of information, the Bala11ti11g 1\'eed.f and Proteding l\'on-Co/IJIIIJ/pli!Je J\'eedJ clusters seem the most engaged, though in different ways. The Balanting Need.!" cluster 198

PAGE 218

seems engaged across clusters, while the Proteding Xon-Co11Jt11Jlptille j\'eed.r cluster is the most engaged with its own members. The Proteding Conmmpti11e 1\1eed.r and Clfm'lll J_y.rlt'lll Broken cluster appear to be the least engaged of all the clusters. HO\v do differing levels of engagement haYe an impact on the representation of interests in a participatory process? Wehler (1995) deftnes aji1irprocess as one in which the level of participation across interests is equal, with participating including such things as bringing information into the process and participating in the discourse. The informacion sharing component appears to be unequal, at least among the indiYidual members of the Colorado Basin Roundtable process. Petts and Leach (2000) deftnc representation as needing the representati,-es to not represent interests, but rather be "representative of interests" (p. 21), something that occurs when participants bring interest-based agendas to the participatory process that reflect the interests of the groups they are charged to represent. The Yision Petts and Leach (2000) haYe for how representatives of interests behave in a participatory process is reflectiYe of how the Advocacy Coalition Framework believes policy stakeholders behave. They are both setting an expectation that participants will engage in shared actions that reflect the broader beliefs of their interest groups or belief coalitions. The Colorado Basin Roundtable members do represent the ftve interests that were found outside of the roundtable, which is a start at meeting the expectations of these authors. However, their differing levels of engagement may suggest the interests at the table are not equally engaged in bringing their agendas to the forefront. This may challenge an unspoken 199

PAGE 219

assumption of public participation, that indi,iduals are equally engaged. Does unequal engagement lead to a process where representation is unequal, not due to the numbers of interests represented, but due to the level of engagement of the representatives? Implications of the Findings To summanze the three propositions explored in this study, they collectively help to understand the context of the Colorado Basin Roundtable. They contribute to understanding the concept of policy beliefs in a participatory process. By doing so, they begin to establish the connection between participation literature that has heavily emphasized the "public" and policy literature that has largely ignored the manr participatory. processes for developing policies. One of the implications of the existence of the Ba/andng l\'eedJ cluster is the question of whether the Advocacy Coalition Framework is able to account for stakeholders who seek to achieve their policy goals through collaboration, rather than working most closely with others who share their policy goals. The Balancing Need.r cluster was not composed of indi,iduals lacking core beliefs or conflicting policy goals, but rather was composed of individuals prioritizing cooperation above their policy goals. This type of stakeholder could be critical for a participatory process an invaluable partner, as the stakeholders participating in this research noted -and perhaps should be explored in more depth in the context of both policy and participation literature. Similarly, an implication of the existence of the Proteding Non-Co!Wflllplit'e l\'eedJ cluster is the critical necessity for public participation literature to consider the affect of 200

PAGE 220

belief coalitions in participatory processes. The Pmteding Non-ConJ!fmptire i\'ml.r cluster exemplifies, more than any other cluster in this analysis, that the .\d,ocacy Coalition Framework's belief coalitions are real and will influence interaction in a participatory process. If a participatory process is analyzed for its fairne.r.r, from selection of members through to opportunities to deliberate, but the analysis fails to account for the belief coalitions that participants bring with them to the table and their influence on interactions, how can the researcher understand the participation process? Due to the case study approach, the in-depth exploration demonstrates the nlue of exploring participatory processes using belief coalitions. It also lea,es future researchers \Vith a methodological approach to identifying those belief coalitions in other settings, using cluster analysis with a set of questions defining different priorities for the policy outcomes of the participatory process. HoweYer, it is important to consider that the belief coalitions identified in this stud\haYe no relevance outside of Colorado water policy. These belief clusters are unique to the context of the water roundtables and Colorado water issues. The extensin stakeholder participation in designing the research resulted in all of the suney questions being highly tailored to the research setting and thus not only are the belief questions largely only releYant to Colorado, the same can be said of most other quesuons. For example, the demographics include specific questions about water rights, framed in the language of Colorado water law. Similarly, the success questions arc specific to the list of actiYities the law and water stakeholders identified as underway in 201

PAGE 221

the water roundtables. Though little of the research tools in this study can be easily applied to future research, the theoretical focus and onrall methodological approach can be further explored, as it did lead to interesting and statistically significant findings despite the small sample size. This theoretical approach had both practical applications and contributed to the more academic explorations of participation and policymaking. The propositions and findings from the studY lead to new questions on the complexity of connecting a participatory process to the broader concepts of policymaking, belief coalitions, and representation. Should future studies continue this line of exploration, one of the first steps may be to deYelop a more clearly defined framework that pulls the ideas together, builds on the areas where the findings reflected the propositions, and tests the framework's Yalidity within one or more participatory processes. Policy Implications for the Colorado Basin Roundtable and other Participatory Processes Colorado initiated the water roundtable process through House Bill 1177, which began with the following: Water rights protections. (1) It is the policy of the general assembly that the current system of allocating water within Colorado shall not be superseded abrogated or otherwise impaired by this article. Nothing in this article shall be interpreted to repeal or in any manner amend the existing \Vater rights adjudication system. The general assembly affirms the state constitution's recognition of water rights as a priYate usufructuary property right, and this article is not intended to restrict the ability of the holder of a water right to use or to dispose of that water right in any manner permitted under Colorado law (37-75-102, C.R.S.). 202

PAGE 222

In application, this component of the bill ensures that implementation of many, if not most, of the water roundtables decisions will be dependent on the agreement of the existing water rights holders in Colorado. The roundtable process does not haYc the authority to interfere with indiYidual water rights or the current legal structure. If representation and participation on the water roundtables is a function of interest groups, as captured in this study through the fiye belief clusters, will the Colorado Basin Roundtable be able to come to consensus around decisions that would then be implemented by those who currentlY haYe water rights? The Colorado Basin Roundtable's highest le,els of engagement are from the Balancing i\'eedJ cluster and the Pro/eding Non-Co!uttmptil'l' NeedJ cluster. The two clusters with the highest le,els of water rights ownership, the Proleding Consttmptile j\'eedJ cluster and Pmtedi1zg lglimlluml ;\1eedx cluster, are less engaged. Does this haYe consequences for the types of decisions made at the roundtable? Additionally, does it ha,-e implications for how decisions arc implemented after being made by the roundtable? In a context where stakeholders hold the power, not the participatory process or enn the policymakers, does this process have potential to succeed in addressing \Vater challenges without engagement of the two clusters with the highest \Vater rights ownership? While debriefing the study results with the members of the Colorado Basin Roundtable, one of the stakeholders noted that "if consumpti,-e uses do not take the roundtable process seriously, it won't amount to much." The roundtable members discussed the impact possible through the roundtable process and one consumptiYe usc 203

PAGE 223

member noted that, at best, the water roundtables haYe "some indirect influence on consumptlYe users." Another stakeholder suggested the process might be more successful if it manages to proYide "clear, unbiased information and educational outreach," all activities which will not require the same level of consensus building across interests as attempts to change policy (Colorado Basin Roundtable, 2008). The implication for the Colorado Basin Roundtable, and possibly the broader roundtable process, is that engagement is unequal, and between this observation and the lack of legal authority, the stakeholders uwolnd are articulating best-case scenario outcomes for the roundtable process that do not change policy, but it indirectly through education and outreach. Though the legislation calls for local roundtables like the Colorado Basin Roundtable to denlop "locally driven collaborative solutions to water supply challenges" (37-75-104(1)a, C.R.S.), without equal and full engagement perhaps tlus process will more realistically develop locally driven information and education that can inform decision-makers in other venues that han more authority to denlop and implement solutions to water supply challenges. The challenge for the Colorado Basin Roundtable is to assess whether information and education is enough, or whether they need to identify strategies for more fully engaging all stakeholders across all interests. Grounds for Further Research As an exploratory study bringing together the concepts of representation, policy beliefs and coalitions, and participatory processes, this dissertation creates more 204

PAGE 224

questions than It answers. Further research of particular interest to this researcher include: (a) Follow up on Proposition 1 to identify realistic ways for public participation processes to identify interests that need to be at the table through beliefs, rather than demographics or affiliations. This might be accomplished by expanding the exploration of beliefs in participatory processes through additional srudies that redo that component of this analysis, replacing the nine statements used here with statements appropriate to other processes, and comparing the resulting belief clusters to demographics. (b) Following up on Proposition 2 to assess the Ienis of interaction among those with shared beliefs and conflicting beliefs in other policy contexts. In particular, it may be interesting to explore interaction in a participatory process that has successfully achieYed a consensus decision for a policy change and in a participatory process that has been unable to reach consensus. Kim and Rob's (2008) concept of resources a\'ailable for coordination was not addressed in this study, and adding that to the inYestigation of shared action might re\eal additional information about how representati,,es interact within a participatory process. It may also be interesting to assess the historical context of the different interests and the policy tools aYailable to them. As \Ve saw with the members of the Protedi11g I\'on-Co!WtllJptif'e NeedJ cluster, past policy adYocacy strategies and current legal structures necessitated more coordination than other interest groups in the \Vater policy context required. Does this same pattern occur in other policy settings? 205

PAGE 225

How ha,e other historical necessities influenced the deYelopment and level of shared action in belief coalitions? (c) Expand the innstigation of Proposition 3 to better understand what it means to represent interests fully, including creating a fair process where interests equally share their information and debate their nlues in the process. Additionally, expand the inYestigation of Proposition 3 eYen further to begin to address the concept of competeJh'e, which goes beyond representation to assess whether individuals haYe the skills and knowledge to participate equally. It may be useful to explore the qualitatiYe information that can be gained from obsetTing a participatory process where the beliefs of participants han been mapped. Do participants whose belief clusters are engaged in shared action participate in the dialogue more or less? Is there action occurring between meetings and thus driving the work forward by some interests and not others? This study ignored the actual participatory process as it did not attempt to collect data that would describe how the participatory process itself is functioning, but limited its scope to the participants. Future studies may find the faimeJJ of the process can best be determined by obsening the process itself, not just exploring the actors and their selfreported connections to each other and the process. Another area for further research goes beyond the focus on interactions within a participatory process to ask hmv those interactions lead to changing outcomes. This study did not explore outcomes beyond asking about participants perceptions of the leYel of success of the roundtable in achieYing some basic process steps. HoweYer, 206

PAGE 226

ultimately, participatory processes are intended to achie\'e policy outcomes. To truly examine these participatory policymaking processes, researchers must pay attention to the outcomes. To apply the exploration of belief coalitions to the examination of outcomes from participatory processes, further research could explore whether the leYcl of engagement of belief coalitions participating in a process is related to achieYing policy wtns. i\lore broadly, how does the distribution of belief coalitions in a participatory process relate to the outcomes achieYed by the process? \V'hat distribution and interaction of belief coalitions leads to compromise outcomes Yersus '\vins" by one side or another? :\'lost of the additional areas for research \vould benefit from qualitati,,e data collection and analysis. Early study of new concepts benefits from qualitati,-e methods, as they allow for a depth often not a\'ailable in a quantitatiYe study. This case study benefited from a richness of data that allowed for a depth of study despite its quantitatin methods, but many of the unanswered questions would require more careful and thorough obserntion of the process of participation. 207

PAGE 227

1) 2) 3) 4) 5) 6) 7) 8) 9) APPENDIX A FIN.\L COLORADO BASIN ROL'NDTABLE SURVEY Outcomes and Expectations of the Roundtable Process 1. What outcomes are you pursuing in water policy and management issues? 2. What outcomes do rou desire from the Colorado Basin Roundtable? 3. What outcomes do you desire from the HB 1177 process? .... 0 G..l c:: G..l <:.I eo 11..1 eo -< <:: G..l >-G..l How strongly do you agree that the 1177 'M .... G..l G..l C.l eo c:: G..l ...c eo 0 11..1 .... <:: process should support: .... 'Q) <:: cr. .... CD cr. a ... -< z;..s (./'J \Vater transfers to high growth areas/ sectors 1 2 3 4 Protecting the agricultural economy and way of life 1 2 3 4 Protecting the recreational economy and its water 1 2 3 4 needs Protecting ecosystems and non-human species (or 1 2 3 4 just the environment) Increasing cooperation among water basins 2 3 4 Balancing the water demands between consumptiYe 1 2 3 4 uses and non-consumptiYe uses Balancing water supplies and demands 1 2 3 4 Protecting existing indiYidual water rights 2 3 4 Allocation and management of water resources 1 2 3 4 through the market 208 <:.I <:.I ..... OJ:; <:: "' :.a >, 'M c:: 0 !: (./'J 5 5 5 5 5 5 5 5 5

PAGE 228

2 c2 2 'f. 'f. "f. 'f. 'L 2 -r. "' "' "' u u 'L u u u 'F. u ::l '-1 ::l f, 'f. u 2 'f. u t; ""' Recognizing that the 1177 process ;:i r-"f. "' ] '-1 '-1 ;... c;: is only 18 months old: i:;' u (.; ..... tJ ::::; c;: > v; z z 1 Oj How successful has your involYemcnt been with the 2 3 4 5 Colorado Basin Roundtable in terms of achieYing you or your organization's desired outcomes, as you listed in question 1? 1lj How successful has the Colorado Basin Roundtable 1 2 3 4 5 been at achieYing outcomes you belieYe are important for the Colorado Basin, as you listed in question 2? 2 'F. 'L -r. 'F. 'F. 2 'F. tJ tJ tJ u u ru u r-;:i qJ 'f 'f, u 2 'F. i:: u ""' ;:i 'F. .... 'f. 'F. ] qJ .... Recognizing that the 1177 process tJ > c;: c u .... u ..... c is only 18 months old: tJ ::l c;: ._, > (/", z z 12) How successful has the Colorado Basin Roundtable 2 3 4 5 been at achieving outcomes you belieYe are important for the statewide 1177 process, as you listed in question 3? 13) How successful has the Colorado Basin Roundtable 2 3 4 5 been at prioritizing key water policy and management issues to address? 14) How successful has the Colorado Basin Roundtable 2 3 4 5 been at developing a grants process to manage the Water Supply Resene Account applications? 15) How successful has the Colorado Basin Roundtable 2 3 4 5 been at educating its members on key water issues? 16) How successful has the Colorado Basin Roundtable ..., 3 4 5 .... been at developing a needs assessment plan? 209

PAGE 229

17) How successful has the Colorado Basin Roundtable 2 3 4 5 been at fostering dialogue and collaboration within the Colorado Ba.,in? 18) How successful has the Colorado Basin Roundtable 2 3 4 5 been at fostering dialogue and collaboration acroJJ other bmi11J? 19) How successful has the Colorado Basin Roundtable 1 2 3 4 5 been at outreach to stakeholders outside of the Roundtable? 20) Is there anything else your roundtable has accomplished at this point? 210

PAGE 230

NETWORK QUESTIONS: The Roundtable and Other Water Stakeholders Pleme but an.rwerfor all other memben and attendeeJ at the Colorado Ba.,in Roundtable. 20) On a scale of 1 -5 (daily to nenr), please indicate how often in the last year you have exchanged information related to water policy and management issues with other members of the roundtable. 2' ..... -E ... .... c: :::: c Roundtable E c Roundtable ... c: Members c u ;...._ -E ... ::::;;: .... .... c: u "' :;. ;:: u 0 'J) u ,..... "< z Members c ..... u .-. -E -E ..... ::::;;: c: :.; "' ca :.; c r. Q :2 ...... 1 (names of 1 2 3 4 5 27 1 2 3 4 2 roundtable 1 2 3 4 5 28 1 2 3 4 3 members are 1 2 3 4 5 29 1 ') 3 4 4 listed here) 1 2 3 4 5 30 1 ') 3 4 5 1 2 3 4 5 31 1 2 3 4 6 1 2 3 4 5 32 1 2 3 4 7 1 2 3 4 5 33 1 ') 3 4 8 1 2 3 4 5 34 1 ') 3 4 9 1 2 3 4 5 35 1 2 3 4 10 1 2 3 4 5 36 1 2 3 4 l1 1 2 3 4 5 37 1 2 3 4 12 1 2 3 4 5 38 1 2 3 4 13 1 2 3 4 5 39 1 2 3 4 14 1 2 3 4 5 40 1 2 3 4 15 1 2 3 4 5 41 1 2 3 4 16 1 2 3 4 5 42 1 ') 3 4 17 1 2 3 4 5 43 1 2 3 4 18 1 2 3 4 5 44 1 2 3 4 19 1 2 3 4 5 45 1 ') 3 4 20 1 2 3 4 5 46 1 2 3 4 21 1 2 3 4 5 47 1 2 .., 4 .) 22 1 ') 3 4 5 48 1 2 3 4 23 1 2 3 4 5 49 1 2 3 4 24 1 2 3 4 5 50 1 ') 3 4 25 1 2 3 4 5 26 1 2 3 4 5 211 ..... u :.-:.; z 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

PAGE 231

21) Please identify the people NOT on the Colorado Basin Roundtable, but involved in water policy and management issues anywhere within the state, with whom you most frequently exchange information related to water policy and management issues. On a scale of 1 -4 (daily to less then monthly), how often ha,e you exchanged information with them in the last year? Names ofiNDIVIDUALS ;..-, ::::2 (Please also list the organizations of each individual you list) IU Q,l Q 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 212 ...2:" ...c ..... c 0 3 3 3 3 3 3 3 3 c Q,l ;..... ,... v; ::: v; 0 -E 4 4 4 4 4 4 4 4

PAGE 232

1 .., ... 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 22) On a scale of 1 -5 (nry important to not at all important), please indicate the extent to which the information you recein from each roundtable member is important, that is, helps you achie,'e your goals for water policy and management issues. If you do not exchange information with a roundtable member, please mark "Not applicable" (x). .... -t:: .... -5 c: t:: .... 5 0 .... .... "' .... t:: ;:::-0 c.. Roundtable 5 .S c. "' ... '""; Members 0 .... .... .!::! c. c: i::' c.. 5 ..c '?. ... .... c. c 0:.: 0:.: b c. E .... 0 0 (.J E 0 0 > './} z z z t:: -c: t: 5 ..... c ... ,... c. 0 Roundtable c. Members -c.. c: ..... i-: 5 .... ... c <:,) :,.. t:; c. 0 <:,) E > ...... z (names of 1 2 3 4 5 X 27 1 2 3 4 roundtable 1 2 3 4 5 X 28 1 2 3 4 members are 1 2 3 4 5 X 29 1 2 3 4 listed here) 1 2 3 4 5 X 30 1 .., 3 4 1 2 3 4 5 X 31 1 2 3 4 1 2 3 4 5 X 32 1 2 3 4 1 2 3 4 5 X 33 1 2 3 4 1 2 3 4 5 X 34 1 .., 3 4 1 2 3 4 5 X 35 1 .., 3 4 1 2 3 4 5 X 36 1 .., 3 4 1 .., 3 4 5 X 37 1 2 3 4 1 2 3 4 5 X 38 1 .., 3 4 1 2 3 4 5 X 39 1 .., 3 4 1 2 3 4 5 X 40 1 .., 3 4 1 2 3 4 5 X 41 1 2 3 4 1 2 3 4 5 X 42 1 2 3 4 1 .., 3 4 5 X 43 1 2 3 4 1 2 3 4 5 X 44 1 .., 3 4 1 2 3 4 5 X 45 1 .., 3 4 1 2 3 4 5 X 46 1 .., 3 4 1 2 3 4 5 X 47 1 2 3 4 1 .., 3 4 5 X 48 1 .., 3 4 1 .., 3 4 5 x ... 49 1 2 3 4 1 2 3 4 5 X 50 1 2 3 4 1 2 3 4 5 X 1 2 3 4 5 X 213 ..... 5 ;:: c.. '?. -.... z 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 :, 3 .!::! c.. c... c 2 X X X X X X X X X X X X X X X X X X X X X X X X

PAGE 233

13) Please identify the people NOT on the Colorado Basin Roundtable, but involved in water policy and management issues within the state, who provide you with information that is important, that is, that helps you achie,e your goals for water policy and management. These may be the same people or different people from questions 25. On a scale of 1 -4 (very important to not very important), how important is the information you exchange with each person? ... c 5 .... .... 0 c c. 5 .... 0 ... .... Names of INDIVIDUALS c. c ...: ...: ...c ... ::: .... 0 QJ (Please also list the organizations of each individual you list) c 0.. E Q.) E 0 > -VJ 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 214 .... c 5 .... 0 c. i::;' '-1 .... 0 z 4 4 4 4 4 4 4 4

PAGE 234

24) On a scale of 2 to -2 (strongly agree to strongly disagree), please indicate the to which you belie,e the following roundtable members agree with or disagree with your goals for the roundtable process. If you do not know, please mark 'Tnknown" (x). Roundtable Members 1 (names of 2 1 2 roundtable 2 1 3 members are 2 1 4 listed here) 2 1 5 2 1 6 2 1 7 2 1 8 2 1 0 -1 -2 X 0 -1 -2 X 0 -1 -2 X 0 -1 -2 X 0 -1 -2 X 0 -1 -2 X 0 -1 -2 X 0 -1 -2 X 9 2 1 0 -1 -2 X 10 2 1 0 -1 -2 X 11 2 1 0 -1 -2 X 12 2 1 0 -1 -2 X 13 2 1 0 -1 -2 X 14 2 1 0 -1 -2 X 15 2 1 0 -1 -2 X 16 2 1 0 -1 -2 X 17 2 1 0 -1 -2 X 18 2 1 0 -1 -2 X 19 2 1 0 -1 -2 X 20 2 1 0 -1 -2 X 21 2 1 0 -1 -2 X 22 2 1 0 -1 -2 X 23 2 1 0 -1 -2 X 24 2 1 0 -1 -2 X 25 2 1 0 -1 -2 X 26 2 1 0 -1 -2 X Roundtable Members 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 215 (.) (.) '"' b.C :, (.) E ... b.C '../] < 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 .., 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 c ..... '"' (.) b.C (.) '"' 'f. b.C (j ;... (.) <:J i (j '"' ... b.C tJ c;: c:; c;: "' c f. ;::, !::! -'../] 0 -1 -2 () -1 .., 0 -1 .., 0 -1 -2 0 -1 -2 0 -1 .., 0 -1 .., 0 -1 .., 0 -1 .., 0 -1 -2 0 -1 -2 0 -1 2 0 -1 2 0 -1 .., 0 -1 2 0 -1 -2 0 -1 2 0 -1 -2 0 -1 .., 0 -1 .., 0 -1 -2 0 -1 -2 0 -1 2 0 -1 2 c ;:: c ..... ....:;: ..... :.., X X X X X X X X X X X X X X X X X X X X X X X X

PAGE 235

25) Please identify the people NOT on the Colorado Basin Roundtable, but involnd in water policy and management issues within the state, who you belieYe most agree with your goals for the water roundtable process. These may be the same people or different people from questions 25 and 27. To what extent on a scale of 1 -2 (strongly agree to agree) do you belien they agree with your goals for the roundtable process? Names of INDIVIDUALS (Please also list the organizations of each individual you list) c:: QJ 0 QJ ... ... .... /;)[; (,/'; 2 2 2 2 2 2 2 2 216 QJ QJ ... OJ) < 1 1 1 1 1 1 1 1

PAGE 236

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 26) On a scale of 1 -5 (completely to neYer), please indicate the extent to which you would trust the following roundtable members to keep your interests in mind on \Vater policy and management issues, regardless of whether he/she shares your mterests. If you do not know, please mark 'Tnknown" (x). '/. = 0 ';:) '{:. c:; c:; 2 Roundtable Vi ;;..... "'0 Members = .... (OJ a 0.. ..... ..;::-. ..... II) ,.. E u :.:l II) ;:. ...::;: ...::: ..... c c:; (OJ = u ,...._ c:: p::: z v -Roundtable 2 ;;; ..... "'0 Members t:l tl c.. ..... ,.. c::; t:l ...::: ... c:; '-.: 0 c:: ..... (names of 1 ') 3 4 5 X 27 1 ') 3 4 roundtable 1 2 3 4 5 X 28 1 2 3 4 members are 1 2 3 4 5 X 29 1 2 3 4 listed here) 1 ') 3 4 5 X .... 30 1 2 3 4 1 2 3 4 5 X 31 1 2 3 4 1 2 3 4 5 X 32 1 ') 3 4 1 2 3 4 5 X 33 1 2 3 4 1 ') 3 4 5 X 34 1 2 3 4 1 2 3 4 5 X 35 1 ') 3 4 1 2 3 4 5 X 36 1 ') 3 4 1 2 3 4 5 X 37 1 2 3 4 1 2 3 4 5 X 38 1 ') 3 4 1 2 3 4 5 X 39 1 2 3 4 1 2 3 4 5 X 40 1 2 3 4 1 2 3 4 5 X 41 1 ') 3 4 1 2 3 4 5 X 42 1 2 3 4 1 ') 3 4 5 X 43 1 ') 3 4 1 2 3 4 5 X 44 1 2 3 4 1 2 "' 4 5 .) X 45 1 ') 3 4 1 2 3 4 5 X 46 1 2 3 4 1 2 3 4 5 X 47 1 2 3 4 1 2 3 4 5 X 48 1 ') 3 4 1 2 3 4 5 X 49 1 ') 3 4 1 2 3 4 5 X so 1 ') 3 4 1 ') 3 4 5 X 1 2 3 4 5 X 217 ..... > ..... 6 ,.. (OJ ...::;: ;.. (OJ = z :,; 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X 5 X

PAGE 237

'27) Please identify the people NOT on the Colorado Basin Roundtable, but involved in water policy and management issues within the state, who you trust the most to keep your interests in mind on water policy and management issues, regardless of whether he/ she shares your interests. These may be the same people or different people from questions 25, 27, and 29. To what extent, on a scale of 1 -4 (completely to rarely), do you trust them to keep rour interests in mind? ;;.., 1l "' Names ofiNDIVIDUALS '1i ... ... c cu -0 0. c '!:] (Please also list the organizations of each individual you list) E cu 0:: ..::: -;:l 0 c u 0 ...... CI) 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 218 ,.-, '1j .... 0:: p::: 4 4 4 4 4 4 4 4

PAGE 238

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 28) On a scale of 1 -5 (nry confident to not at all confident), please indicate the extent to which you are confident each of the following roundtable members will follow through on a commitment. If you do not knO\v, please mark 'Tnknown" (x). -... ..... ... ... a:; c c "0 (.) '"' u:::: "0 "0 ... c u:::: u:::: c Roundtable "' c c E --..!<: <:.1 c 0 0 0 c > u rJ; z z ::. ..... ... ... i:i c ..... "0 (.J (.) u:::: "0 -::; c ;:: u:::: u:::: Roundtable QJ u "0 u u Members u:::: ... ... "' c c ...!: (:;' "?. c t.l ::: u "0 :.-"' i:; ..... ... ... t.l ,q c c > '-' z z '-' v:. (names of 1 2 3 4 5 X 27 1 2 3 4 5 roundtable 1 2 3 4 5 X 28 1 2 3 4 5 members are 1 2 3 4 5 X 29 1 ') 3 4 5 listed here) 1 2 3 4 5 X 30 1 2 3 4 5 1 2 3 4 5 X 31 1 ') 3 4 5 1 2 3 4 5 X 32 1 ') 3 4 5 1 2 3 4 5 X 33 1 2 3 4 5 1 2 3 4 5 X I 34 1 2 3 4 5 1 2 3 4 5 X 35 1 2 3 4 5 1 2 3 4 5 X 36 1 2 3 4 5 1 2 3 4 5 X 37 1 ') 3 4 5 1 2 3 4 5 X 38 1 2 3 4 5 1 2 3 4 5 X 39 1 ') 3 4 5 1 2 3 4 5 X 40 1 2 3 4 5 1 2 3 4 5 X 41 1 2 3 4 5 1 2 3 4 5 X 42 1 ') 3 4 5 1 ') 3 4 5 X 43 1 2 3 4 5 1 2 3 4 5 X 44 1 2 3 4 5 1 2 3 4 5 X 45 1 2 3 4 5 1 ') 3 4 5 X 46 1 2 3 4 5 1 2 3 4 5 X 47 1 2 3 4 5 1 2 3 4 5 X 48 1 2 3 4 5 1 2 3 4 5 X 49 1 ') 3 4 5 1 2 3 4 5 X 50 1 2 3 4 5 1 2 3 4 5 X 1 2 3 4 5 X 219 ..... :::: c c ..!<: ..... X X X X X X X X X X X X X X X X X X X X X X X X

PAGE 239

29) Please identify the people NOT on the Colorado Basin Roundtable, but imolYed in water policy and management issues within the state, who you are most confident \Vill follow through on a commitment. These may be the same people or different people from questions 25, 27, 29, and 31. To what extent, on a scale of 1 -4 (very confident to not very confident), are you confident they will follow through on a commitment? -c v "'0 u:::: .... c c c CJ "'0 u u:::: -.... Names of INDIVIDUALS c c c.: ...!: 0 v u "'0 u:::: v (Please also list the organizations of each individual you list) i:;' c E "" 0 0 > u '..ll 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 220 .... c "" "'0 u:::: .... c u ;-:. .... (iJ > 0 z 4 4 4 4 4 4 4 4

PAGE 240

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 30) On a scale of 1 5 (nry influential to not at all influential), please indicate the extent to which you consider each person listed below to be influential in water policy and management issues in the STATE OF COLORADO; that is, the people who seem to haYe pull, weight, or clout with others on these issues. If you do not know, please mark "l'nknown" (x). ] "?. ] .o (j t: t: ] ::; IIJ tJ ::; ::; Roundtable .5 :: .s Members (IJ ] .... .S ::; "" ...!: c "?. .5 :: ::: tJ .... tJ <:J > "" c :::; E 0 c ...::;: <:J 0 :: > :: z z -v; ] ::; :: ::; ] ::; <:J ::; ::; Roundtable :: <:J .... :: Members ::; "" t: ...!: "?. .5 ::: <:J .... tJ ;.. "" i::' c .... (IJ > :: z z (names of 1 2 3 4 5 X 27 1 ') 3 4 5 roundtable 1 2 3 4 5 X 28 1 ') 3 4 5 members are 1 2 3 4 5 X 29 1 2 3 4 5 listed here) 1 2 3 4 5 X 30 1 ') 3 4 5 1 2 3 4 5 X 31 1 ') 3 4 5 1 2 3 4 5 X 32 1 ') 3 4 5 1 2 3 4 5 X 33 1 2 3 4 5 1 2 3 4 5 X 34 1 2 3 4 5 1 2 3 4 5 X 35 1 2 3 4 5 1 2 3 4 5 X 36 1 ') 3 4 5 1 2 3 4 5 X 37 1 ') 3 4 5 1 2 3 4 5 X 38 1 ') 3 4 5 1 2 3 4 5 X 39 1 2 3 4 5 1 ') 3 4 5 X 40 1 ') 3 4 5 1 2 3 4 5 X 41 1 2 -3 4 5 1 2 3 4 5 X 42 1 2 3 4 5 1 ') 3 4 5 X 43 1 ? 3 4 5 1 2 3 4 5 X 44 1 ') 3 4 5 1 2 3 4 5 X 45 1 ? 3 4 5 1 2 3 4 5 X 46 1 ') 3 4 5 1 2 3 4 5 X 47 1 ') 3 4 5 1 2 3 4 5 X 48 1 ') 3 4 5 1 2 3 4 5 X 49 1 2 3 4 5 1 2 3 4 5 X 50 1 2 3 4 5 1 2 3 4 5 X 1 ') 3 4 5 X 221 :: :> 6 ...::;: :: X X X X X X X X X X X X X X X X X X X X X X X X

PAGE 241

31) Please identify the people NOT on the Colorado Basin Roundtable, but imol\'ed in water policy and management issues within the state, who you consider the most influential in \Vater policy and management issues in the state of Colorado; that is, the people \vho seem to han pull, weight, or clout on these issues. These may be the same people or different people from questions 25, 27, 29, 31, and 33. Please indicate on a scale of 1 -4 (very influential to not very influential) how influential each person is in water policy and management issues. -.:= o c: (.J :" ;:J t::::: E .5 (.J -.:= ..... Names ofiNDIVIDUALS ;:l co:: o ..c .5 c:
  • c: ....... r.r; 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 222 -.:= o .... CJ ;:J co:: .E !::;' (.J :.-..... 0 z 4 4 4 4 4 4 4 4

    PAGE 242

    DEMOGRAPHIC QUESTIONS 32) What is your name? --------------33) If you are affiliated with an organization, what is your organization's name? 34) Total years residing in Colorado D Less than 5 years D 5-10 years D 11-20 years D Over 20 years 36) l\lonths actively invohed in the Rott11dtable 0 0-3 months 0 4-6 months 0 7-9 months D 10-12 months 0 More then one year 38) Gender 0 Male D Female 35) Years acti,ely imolved in water issues D Less then one year D 1-4 years D 5-1 0 \'ears D 11-20 \'ears D Over 20 vears 37) Age D 18-3o D 31-4o D 41-5o D s1-6o D 61-7o D Over 70 years old 39) In which Colorado county do you primarily live? 40) Which sub-basins do you work in and with (e.g. Upper or Lower basin of a specific basin or a tributary sub-basin)? If your organization is statewide in focus, please write "statewide." 223

    PAGE 243

    41) Do you as an indiYidual own absolute water rights? __ Yes No If yes, to the best of your recollection: a) What is year of your most senior water right? b. If the rights you personally own c. If the rights you personally own (including through shares in a ditch company) include flow rights, combined, how many cubic feet per second (CFS) of water flow do they allow? (including through shares in a ditch company) include storage rights, combined, how many acre-feet of water storage do they allow? 0 Less than 5 acre feet 0 6 50 acre feet 0 51 150 acre feet 0 1511000 acre feet 0 1001 + acre feet 0 Less than 1 CFS D 1-10 CJ:.s D 11 -100 CFS 0 Greater than 100 CFS 42) Does your organization own absolute water rights? (Please answer for the organization you represent, if you represent one on the Colorado Basin Roundtable) Yes No If yes, to the best of your recollection: a) What is year of your organization's most senior water right? b. If your organization's rights include c. If your organization's rights include storage rights, combined, how flow rights, combined, how many many acre-feet of water storage do cubic feet per second (CFS) of water they allow? flow do theY allow? D Less than 5 acre feet D Less than 1 CFS D 6 50 acre feet D 1-10 CFS D 51 150 acre feet D 11-lOOCFS D 151 -1 000 acre feet D Greater than 100 CFS D 1 001 + acre feet 224

    PAGE 244

    43) If you participate on one or more roundtables or the IBCC, please check each one on which you participate and your le\cl of participation. If your participation lcYcl is "other", please explain. -----. = b.() a c: c.. ::: ... .. .. :--. 0 c: :.; c:..c >..c 0 .. u E c!: E Cll 'J:. ; Cll ..c .... 0 u ..c :.; .... z:; ;,:j 0 0-S D D Arkansas D D D D D D Colorado D D D D D D Gunnison D D D D D D North Platte D D D D D D 1\letro D D D D Dolores/San Juan/San D D Miguel D D 0 D D D Yampa/White/ Green D D D D D D South Platte D D D D D D Rio Grande D D D D D D IBCC D D D D 44) Please check any formal/ elected roles you haYe on the Colorado Basin Roundtable: D Chair D Elected Representative to the IBCC D Vice-Chair D Other Fonnal Role (Please Explain) D Secretary D No formal role 225

    PAGE 245

    45) If you are a member of the Colorado Basin Roundtable, please check the interest that you were appointed to represent on the roundtable: D Appointed by a County D Appointed by the Municipalities D D D Appointed by the Board of Directors of a Water Conservancy District Appointed by the Board of Directors of a Water Conservation District Appointed by the Legislature D At large member representing D An environmental conservation organization D Agriculture interests D Recreation interests D Local domestic '.Vater provider interests D Industrial interests D Other at large member: 0 Owner of adjudicated water rights D Contract with the Federal Bureau of Reclamation D Other:--------------------0 Non-voting member from outside the basin D Colorado \Xiater Conservation Board representative D Liaison from state agency D Liaison from federal agency D Other appointed position: -----------------0 Not appointed 46) :\rc you \Vorking between meetings with members of the Colorado Basin Roundtable on specific issues for the Colorado Basin Roundtable (including as part of subcommittees), such as needs assessments, other studies, grant processes, etc.? l f yes, please explain. 226

    PAGE 246

    47) i\'lark a "1" next to the affiliation/interest that best describes your role in the water policy and management community. If more then one affiliation/interest describes your role, please usc a "2" and a "3" to identify up to three affiliations/interests total. Elected Offices Municipal Countv State Non-Elected Gmernment Municipal Countv State Consultant Water resources Agriculture Water quality Environmental/ consenation Governmental relations Recreation Other consultant: ----Special Districts Water Consenancv District \X'ater Consenation District Irrigation District Rural Water District Mutual Irrigation District Other special district: Other PriYate Enterprises Public l' tilitY Research Institute/ L'ni,ersity Engineering Firm/Consultant Legal Firm/Consultant Recreation/Tourism Company Agriculture (rained, dryland) Agriculture (irrigated) l\Iedia Other private enterprise: __ N on-Governmental/NonProfit Watershed Group Land Trust Consen-ation Group Environmental Group Recreation Group Other-Non Profit: ____ Other Affiliations 227 Colorado \Xfater Consenation Board lnterbasin Compact Committee Any other affiliation: ___

    PAGE 247

    APPENDIX B NORl\L\L DISTRIBl'TIONS Table B.1: Significance leYel for normal distributions by cluster on each network Yariablc using a Shapiro-Wilk test. Significance of < .05 indicates assumption of normality is Yiolated. Reporting on ConsumpAgNon-Question Balancing ConsumpWhich Cluster? tive ricultural tive ConsumptiYe .635 .064 .058 .160 Share vour .. Ag_ricul tural .. 081 .850 .048 .597 goals Balancing .996 .904 .376 .181 Non -ConsumptiYe .. 301 .732 .086 .369 Information Consumptive .412 .. 291 .. 683 .007 SharingAgricultural .230 .594 .149 .029 Balancing .. 713 .682 .462 .036 Frequency N on-Consum_Q_tiYe .313 .105 .. 619 .296 Information ConsumptiYe .035 .522 .. 838 .272 Agricultural .876 .. 907 .012 .071 SharingBalancing .178 .393 .746 .758 Value Non-Consumptive .002 .555 .402 .545 Trust to Consumptive .825 .. 322 .. 649 .014 keep Agricultural .030 .. 817 .547 .018 interests in Balancing .513 .. 672 .009 .035 mind Non-Consumptive .772 .045 .. 299 .330 Trust to Consumptin .369 .. 378 .. 381 .145 follow:\.gricultural .953 .431 .. 987 .070 through on Balancing .. 954 .553 .124 .794 commitment Non-Consumptive .033 .093 .. 831 .215 ConsumptiYe .176 .898 .404 .219 Influence Agricultural .041 .039 .400 .135 statewide Balancing .085 .. 939 .088 .557 N on-ConsumptiYe .808 .350 .. 938 .. 059 0 o of questions \Vith a normal 83% 93% 88-'o 75/o distribution for each cluster 228

    PAGE 248

    REFERENCES Abelson, J., Forest, P. G., Eyles, J., Smith, P., l\lartin, E., & Gamin, 1.-.P. (2003). Deliberations about deliberative methods: Issues in the design and cnluation of public participation processes. Soda/ Stienl"l' and Medidm, 57(2), 239-251. Aldenderfer, M. S., & Blashfield, R. K. (1984). Cl11ster Benrly Hills, C.\: Sage UniversitY Press. Ambruster, A. (2008). Co//aborati11e Perms tedmotTalit" making: Jtaleu,ide JJJtl!rr plan. Sacramento, CA: Center for Collaborative Policy. Anderson, G. L., Herr, K., & Nihlen, A. S. (1994). Stllqying)'OIIr OJVJI .rd1ool: ln ed11cator'J guide to qttalitalifle pmditioner reseanb. Thousand Oaks, CA.: Sage. Arnstein, S. R. (1969). A ladder of citizen participation. Journal Ammmn lnJii/11/e Planners, 35(4), 216-225. Barber, B. R. (2004). Strong democra':)': Pm1itipatOI)' politics jor a 11e1v t{ge, tJvenlietb-annitet:rct!J' edition. Berkeley, CA: L'niversity of California Press. Barnett, T. P., & Pierce, D. W. (2009). 511Jiainable water delil'en"e.rjivm the Colorado Riter in a climate (\' ol. 1 06). Washington, DC: National Academy of Sciences. Beierle, T. C. (1999). l'sing social goals to enluate public participation in em'ironmcntal decisions. PolioStudieJ Reriew, 16(3/4), 75-103. Beierle, T. C., & Cay ford, J. (2002). DenJOtnll)' in practice: p11blit" in emirommnlal decisiollJ. \X'ashington, D.C.: Resources for the Future. Bickerstaff, K., & Walker, G. (2005). Shared visions, unholy alliances: Power, gmernance, and deliberative processes in local transport planning. Urban St11dit.r, .J2(12), 2123-2144. Bishop, P., & Davis, G. (2002). Mapping public participation in policy choices. 1/f.llllllian ]o11ma/ 1dminiJ'Iration, 61 (1 ), 14-29. Borgatti, S. P., Everett, i\1. G., & Freeman, L. C. (1992). UCINET 1\': Net\vork analysis soft\vare. Connedion..-, XI /(1-2), 12-15. 229

    PAGE 249

    Box, R. (1998). Citi::;_e11 goJenJaJJt'e. Thousand Oaks, CA: Sage Publications. Bryson, I., & Crosby, B. (1992). Leader"Jhip or the t'Oillll/011 good. San Francisco, CA: Jossey Bass. Bureau of Reclamation. (2005). W)'ater 2025: Prel'mting COI!flid and emu in the JVeJ'I. RetrieYed from the l1nited States Department of the Interior's website: http:// permanent.access.gpo.goy /lps36032/Water2025.pdf. Burnett, M., & DaYis, C. (2002). Getting out the cut: Politics and national forest timber harYests, 1960-1995. AdminiJ!ration and Sotie()', 3.f.(2), 202-228. Bussemaker, J ., & \' oet, R. (1998). Citizenship and gender: Theoretical approaches and historical legacies. Cn'timl S odal Poliq, 18(3), 277-307. Caponera, D. A. (1992). P1intipleJ of water law and adminiJtration: National and intemational. Rotterdam, Netherlands: Balkema Publishers. Carley, K. 1\.1. (2003). Dynamic network analysis. In R. Breiger, K. Carley, & P. Pattison, (Eds.), Dynamic Social Network Modeling and Analysis: Workshop Summary and Papers (pp. 133-145). Washington, D.C.: The National Academic Press. CDM. (2004). Statewide Jvaler JIIPP!J' i11itiati1'e repo11. Denyer, CO: Colorado Department of Natural Resources. CDI\.1. (2006). W'ater JIIPP/J' a11d needJ repon for the Colorado Basi11. Denver, CO: Colorado Department of Natural Resources. Chess, C., & Purcell, K. (1999). Public participation and the environment: Do we know what works? Emiro11mental Sa'mce & Tedmolo!J, 33(16), 2685-2692. Coleman, J. S. (1988). Social capital in the creation of human capital. Amni((JII ]omwal So[iology, 9.f., 95-120. Colorado Institute of Public Policy. (2006). Liti11g i11 the Roc.-9' A1otmtain IVest, u;)'ater 2025: and I /al11e.r a.1 a Mea11.1 for Cooperatio11. Fort Collins, CO: Colorado State UniversitY. Colorado Water Conset:Yation Board. (2007). Colorado Bcui11 Roundtable prumtation to the lntedJaJin Compad Committee. DenYer, CO: Interbasin Compact Committee. Colorado Water Consen-ation Board (n.d.). lnter.rtate water allomtion. Denver, CO: Colorado Department of Natural Resources. 230

    PAGE 250

    Connick, S., & Innes, J. (2001). Outcomu Collaborafil'e lr'ater Poliq _lpp(Jing Comple."\ity TbeOIJ' to El!a!ttation (Working paper 2001-08). Berkeley, CA.: Institute of LTrban and Regional Dc,elopment, Uninrsity of California. Corbridge, J. N. (1998). Historical water usc and the protection of ,ested rights: A challenge for Colorado water law. lrnil'mi(J Re1iew. 69, 503-531. Cornwall, A., & Jewkes, R. (1995). What is parttctpatory research? Sotial Stiena c> Afedidne, .J 1 (12), 166 7-1676. Creighton, J. L. (1999). How to design a public pmticipation progralll (El\I-22). Washington, D.C.: Office of Intergovernmental and Public .-\ccountability, L1nited States Department of Energy. Crosby, N., Kelley, J., & Schaefer, P. (1986). Citizen panels: A new approach to citizen participation. Pub lit" AdminiJfmtion Rmiew, .J-6(2), 170-1'19. Cross, R., & Parker, A. (2004). The hidden power Jotial networks: Undenlandi,zg bow work. real(y getx done in or;gaui::;:_ationJ. Boston, l\L\: Harvard Business School Press. Cunningham, F. (2002). The01iu of democrao: oitiml intmd11diou. New York, NY: Routledge. Curry, N. (2001). Community paruopation in rural policy: Representatinness in the development of millennium greens. Journal En/Jimnmmtal Planning and Managemmt, .J.J( 4), 561-5 76. Curtain, R. (2003). What role for citizens in denloping and implementing public policy? Canberra Bulletin P11blic AdminiJtration, 109, 61-65. Dahl, R. A. (1989). DemoOl1()' and itJ oifit".J". New HaYen, CT: Yale University. Dahl, R.A. (1994). A democratic dilemma: System effectiveness Yersus cltlzen participation. Po/iti"al Stience Qttm1er!J 1 09(1), 23-34. Davis, C. (2009). Advocacy and change in national forest wilderness policies. Jouma/ Ruoun't'J Poliq Re.1earcb, 1(4), 293-306. Davis, C., & Davis, S. (1988). Analyzing change in public lands policymaking: From subsystems to advocacy coalition. Studie.r]o11mal. 17(3), 3-24. Davis, S. K. (2001). The politics of water scarcity in the Western states. Tbe Sotial Scimce ]ouma/, 38( 4), 527-503. 231

    PAGE 251

    Del .eon, P. (1990). Participatory policy analysis: Prescriptions and precautions -1Jimt P11blit". -1dmini.lraliou, 12(2), 29-54. DeLeon, P. (1992). The democratization of the policy sc1ences. P11blit" Rm"ew, 52(2), 125-129. DeLeon, P. (1997). and tbe polio JdenaJ. Albany, NY: State University of New York Press. Depoe, S. P., Delicath, J. W., & Elsenbeer, l\1. F. (Eds.). (2006). Conummication and pub/it" pmticipaliou iu elll'ironmen!al ded.rion-making. Albany, NY: State llniYersity of New York Press. Deutsch, l\L (1985). Di.sflilmtile jtt.slice:. -1 pe!"Jpedit'e. New Haven, CT: Yale LT niYersitv Press. Dietz, G., & Den Hartog, D. N. (2006). Measuring trust inside organizations. Pmonnel Retiew, 35(5), 557-588. Dowding, K.. (1995). l\Iodel or metaphor? A critical renew of the policy network approach. Political StndieJ, +3(1), 136-158. Dryzek, J. S. (1990). Di.amite democmo: Po/itiCJ, poliq, and Jtimce. London, England: Cambridge Llninrsity Press. Dryzek, J. S. (1996). DemotTat)' in t" time.r: ldeal.s, /imit.s, and Jtmgglu. New York, NY: Oxford l'niYersitY Press. Dunn, W. N. (1993). Policy reforms as arguments. In F. Fischer &J. Forester (Eds.), Tbe argllllleJtlatil;e 111m in polio anafy.riJ and planning (pp. 254-290). Durham, NC: Duke L1niYersitY Press. Ellison, B. A. (1998). The AdYocacy Coalition framework and implementation of the Endangered Species Act: A case study in western water politics. Poliry StudieJ ]o11ma/, 26(1), 11-30. Fischer, C. S. (1981). The public and private worlds of city life. Tbe Public aud Ptil'a/e lVodd.s Life, -16(3), 306-316. Fischer, F. (1990). Tedmocrao and tbe politic"J Newbun Park, C.\: Sage Publications. Fischer, r-. (1993). Citizen part1c1pation and the democratization of policy expertise: From theoretical inquiry to practical cases. S,ietuu, 26(3), 165-187. 232

    PAGE 252

    Fischer, F. (2003). Rejit1ming public polit)': Dixmni/J(' polilitJ and deliberati1'e pmpedii'I'J. Oxford: Oxford llniYersitY Press. Flyvbjerg, B. (2001 ). 1\laking xotia/ Jcience mallet: .lOti a/ faiiJ and boJP it mn Jllt"t"rwl agaiu. Cambridge, ;\lA: Cambridge UniYersity Press. Foster-Fishman, P. G., Salem, D. A., Allen, N .. -\., & Fahrbach, K. (2001). Facilitating interorganizational collaboration: 1l1e contributions of intcrorganizational alliances. Ametimn Joumal q/ComJJ!IIIIi(y PD'cbologJ', 29(6), 87 5-905. Frohardt, P. D. (2003). Citi::;_en'J guide to Colorado Jllaler quali(y protedion. Dennr, CO: Colorado Foundation for Water Education. Garson, G. D. (2006). ChtJter anab'JiJ. Retrie,ed August 20, 2007, from http:// garson/PA 765/ cluster Gastil, J., & Keith, W. ;\I. (2005). A nation that (sometimes) likes to talk: "\ brief history of public deliberation in the United States. In J. Gastil & P. Le,ine (Eds.), Tbe Deliberalil'e lfandbook: Stmtegie.rfor Ef!i'dil'e CiPic Engage111mt in tbe Ti/)('1!()'rlni Cmtlll)' (pp. 1-19). San Francisco, C\: Jossey-Bass. Grant, A. (n.d.). C/wter ana(yJiJ. Retriend August 27, 2007, from http:/ / -e130/2b 7ycluster.htm Griffin, C. B. (1999). Watershed councils: An emerging form of public participation in natural resource management. ]o1tma/ Of The Amelimn lf"ater ReJOIII
    PAGE 253

    Hajer, l\1. (1993). Discourse coalitions and the institutionalization of practice: l11e case of acid rain in Britain. In F. Fischer & J. Forester (Eds.), The Atl',IJIJ/elltatiw Tum in Poliq. lna!y.ri1 and Planning (pp. 43-75). Durham, NC: Duke l1niversity Press. Hajer, l\1., & Wagenaar, H. (Eds.). (2003). Deliberatifle polit]' ana!Jsi.r: llnder.Jiandinggo/lemant"e in the network .rotie(y. Cambridge, l\L\: Cambridge UniYersity Press. Hamill, J. (1993). Tmn.roipt to the Intractable Conjlict/Con.rtrttdif!e Conjivntation Prqjed (\'\forking paper 93-12). Boulder, CO: Conflict Research Consortium. Hanel, J. (2008, January 6). Big oil casts big shadow over Colorado's water future. D11rango Herald, Retriend from http:/ / aspbin/ article _genera tion.asp?article_type= news&article_path =/news/ news0801 06 4.htm Hanneman, R. A., & Riddle, l\l. (2005). lntrod11dion to .rocial network metbod.r. Retriend from http:// faculty nettext/ index.html Healey, P. (1993). Planning through debate: The communicative turn in planning theory. In F. Fischer & J. Forester (Eds.), The All',IIIJJelltatille Tum in Poliry Anajy.i.r and Planning (pp. 233-253). Durham, NC: Duke Cniversity Press. Hecox, E. (2006). lnterba.rin t"ompad fad .rheet. Dennr, CO: Office of lnterbasin Compact Negotiations. Hendriks, C. I\1. (2006). When the forum meets interest politics: strategic uses of public deliberation. Sode(y, 3..,1(4), 571-602. Hobbs, G. J. (2004). Citi::;_m'.r guide to Colorado water law. Dennr, CO: Colorado Foundation for Water Education. Hope, K. W., & Waterman, H. A. (2003). Praiseworthy pragmatism? Validity and action research. ]o11mal J\'111:ring, ..,1..,1(2), 120-127. Hoppner, C. (2009). Trust-A monolithic panacea in land use planning? Land Ure 26(4), 1046-1054. Hoyt, L. H. (2005). Colorado's history of emironmental and land use laws, and their impact on Colorado's water. In K. A. Brown (Ed.), Citi::;_en :r g11ide to Colorado '.r 1!11/liron!lll'JJtal em (pp. 24-27). DenYer, CO: Colorado Foundation for Water Education. 234

    PAGE 254

    Innes, J. E., & Booher, D. E. (1999). Consensus building and complex adaptiYc systems: A framework for e,aluating collaboratiYe planning. Joumal -lmetiaw AJJodation, 65( 4), 412-423. Innes, J. E., & Booher, D. E. (2004). Reframing public participation: Strategies for the 21st century. Planning Theon & Pradice, 5(4), 419-436. Interbasin Compact Committee. (2008). Colomdo Ro11ndtable /Jomepuge. RctricYed August 14, 2008, from http:/ / International Association of Public Participation. (2000). P11blic pmtiapatioll JflednmJ. Retrieved August 12, 2006, from http:/ / Inin, R. A., & Stansbury, J. (2004). Citizen participation in decision making: Is it worth the effort? Publi( AdminiJ!ratio11 Rerie211, 6-1(1 ), 55-65. Jackson, J. S., & Shade, W. L. (1973). Citizen participation, democratic representation, and suney research. Urban 9(1), 57-89. Jain, A. K., Murty, J\1. N., & Flynn, P. J. (1999). Data clusters: A reYie,v -1CM S11meys, 31 (3), 264-323. Jenkins-Smith, H., & St. Clair, G. (1993). The politics of offshore energy: Empirically testing the AdYocacy Coalition Framework. In P. Sabatier & H. Jenkins-Smith (Eds.), Poliq Cbange and Leami11g (pp. 1 05-128). Boulder, CO: WestYiew Press. Jenkins-Smith, H., St. Clair, G., & Woods, B. (1991). Explaining change in policy subsystems: Analysis of coalition stability and defection oYer time. _/J/eti.mJ Joumal Science, 35(4), 851-872. Jenkins-Smith, H. C. (1991). Alternati,e theories of the policy process: Reflections on research strategy for the study of nuclear waste policy. PS: Political Sdena a11d PolititJ, 2.J(2), 157-166. Kathlene, L., Lynn, J., Greenwade, A., Sulli,an, W., & Lung, Q. (201 0). Colorudo ntim;: manage111ettt and land uxe planning intl'gmtion. Denver, CO: Center for Systems Integration on behalf of the Colorado Water Conservation Board, Colorado Department of Natural Resources. Kathlene, L., & i\Iartin, J. (1991 ). Enhancing c11:1zen par1:1c1pation: Panel designs, perspectiyes, and policy formation. ]o11mal Polio nabJiJ and lllanage111ent, 10(1 ), 46-63. 235

    PAGE 255

    Kenny, D. S. (2000) .. lrguing abo111 t"OIIJ"eJWtJ: E: .... :amining the ctlJe againJI JVeJiem walerJbed initialil'l'.f and olbt>r t"o!laboratite groupJ adin in nal11ral l"f!J0/111-eJ management. Boulder, CO: Natural Resources Law Center, UniYersity of Colorado School of Law. Kim, Y., & Roh, C. (2008). Beyond the AdYocacy Coalition Framework in policy process. lntemational]oumal qj"Public Admini.rlration, 31 (6), 668-689. King, C. S., Feltey, K. l\1., & Susel, B. 0. (1998). The question of participation: Toward authentic public participation in public administration. P11blic AdminiJtralion Rel!iew, 58(4), 317. Koontz, T. l\1., & Johnson, E. M. (2004). One size does not fit all: Matching breadth of stakeholder participation to watershed group accomplishments. Poliry Stiel!tU, 3 7(2), 185-204. Krimsky, S. (1979). Citizen participation in scientific and technological decision-making. In S. Langton (Ed.), Partzdpation Per.rpedil!e.r (pp. 181-187). Medford, MA: Lincoln Filene Center for Citizenship and Public Affairs. Lachapelle, P. R., & McCool, S. F. (2005). Exploring the concept of "ownership" 1n natural resource planning. 5 odery i..--Na111ral ReJo1mu, 18(3), 279-285. Leach, W. D. (2006). CollaboratiYe public management and democracy: Evidence from western watershed partnerships. P11blic AdminiJiration Re11iew, 66(S 1 ), 100-110. Leach, W. D., & Pelkey, N. W. (2001). Making watershed partnerships work: A review of the empirical literature. ]o11rnal qflF'ater Re.rom
    PAGE 256

    Lubell, l\1. (2007). Familiarity breeds trust: Collectin action 1n a policy domain. The Joumal q/ PoliticJ, 69(1 ), 23 7-250. Lynn, F. l\1., & Busenberg, G. J. (1995). Citizen aLkisory committees and enYironmental policy: What we know, what's left to disco,er. RiJk .na!J'.i.r, I 5(2), 147-162. Majone, G. (1977). Technology assessment and policy analysis. Polit)' Stientn, 8, 173-176. Mansbridge, J. J. (1983). Bqond AdPer.rao Chicago, IL: UniHrsity of Chicago Press. rvlarshall, B. K., & Jones, R. E. (2005). Citizen participation in natural resource management: Docs representatiYeness matter? Sotiologica/ SpedJ71111, 25(6), 715737. Marti, J., & Villasante, T. R. (2009). Quality in action research: Reflections for second order inquiry. 5J'.rtemic Pradit"e and Action Rmarcb. 22(5), 383-396. rvlartin, B., & Carson, L. (1999). Random .reledion in politicJ. Westport, CT: Praegcr Publishers. i\lcCool, S. F., & Guthrie, K. (2001 ). Mapping the dimensions of successful public participation in messy natural resources management situations. Sotie(y and Nat11ral Re.rolfl1'fJ, J.l.( 4), 309-323. McPherson, M., Smith-Lovin, L., & Cook, J. (2001). Birds of a feather: Homophily in social networks -lwma/ RetJiew ofSotiology, 27(1), 415-444. Meyers, C. (2008, March 9). Water rights, pheasant hunters clash. Demt'r Po .-t. Retrie\'Cd from http: I I www .denverpost.coml extremes I ci_ 8505 79 2 Milward, H. B., & Prmran, K. G. (1998). l\Ieasuring network structure. PH/J/it' Admini.rtration. 7 6(2), 387-407. Monge, P. R., & Contractor, N. S. (2003). TbeOJie.r t'OIJJ!Jllfllimtion JJetworkJ. New York, NY: Oxford Uninrsity Press. Moore, E. A., & Koontz, T. l\1. (2003). A typology of collaboratin watershed groups: Citizen-based, agency-based, and mixed partnerships.!J and 1\'atllral Rt.romn .-. 16(5), 451-460. l\Ioote, l\1. A., & 1\lcClaran, l\L P. (1997). Viewpoint: Implications of participatory democracy for public land planning. ]oumal q/Runge Management. 50(5), 473-481. 237

    PAGE 257

    I\Iootc, M. A., McClaran, I\I. P., & Chickering, D. K. (1997). Theory in practice: Applying participatory democracy theory to public land planning. Emlironmmtal Managemml. 21(6), 877-889. I\Iorgan, G. A., Leech, N. L., Gloeckner, G. \VI., & Barrett, K. C. (2004). SPSS jor Jtali.llit'.r: [ .:.fe and inle1pretation. Mahwah, N J: Lawrence Erlbaum Associates. I\lostcrt, E., ,an Beck, E., Bouman, N. W. M., Hey, E., Savenije, H. H. G., & Thissen, W. A. H. (1999). Ri1er bmin management and planning. Keynote paper for International Workshop on River Basin Management, October 27-29, The Hague, www .ct. rba/Keynote.html Narayan, D. (2002). EmpoweJ7JJenl and potJer(y redmtion: A Jomre book. Washington, D.C.: World Bank. Nature Consernncy. (2007). Homepage. Retrieved ;\ugust 14, 2008, from http:// O'Brien, R. (2001). An OYetTiew of the methodological approach of action research. In R. Richardson (Ed.), Theon and Pradice Adion Rman-b. Joao Pessoa, Brazil: L' niversidade Federal da Paraiba. O'Fallon, L. R., & Dearry, A. (2002). Community-based participatory research as a tool to advance environmental health sciences. Environmmtal Healtb PerJpedil'e-' S11pplemml.1 110(S2), 155-159. OwensViani, L. (2001 ). The immion of weJ1em walerJ I!)' non-nalit'e Tbreals to !be we.rl. Paper presented at the Western Regional Panel on Aquatic Nuisance Species, Oakland, CA. Parkins, J. R., & l\litchell, R. E. (2005). Public paructpation as public debate: A deliberative turn in natural resource management. Sode(y & J\'a111ral ReJotmu: An lntemalional Joumal, 18(6), 529-540. Pctts, J. (2001). Evaluating the cffectinness of deliberative processes: Waste management case-studies. ]o11mal Enl'ironmmtal Planning and A1anagemml, .J.J(2), 207-226. Petts, .J., & Leach, B. (2000). Et,alualing me/bods for pttblit' pmtiiipalion: Utera/ure l"t'ZJiezv (No. R&D Technical Report: E135). Bristol, Great Britain: Centre for Environmental Research & Training, The L'niversity of Birmingham. 238

    PAGE 258

    Pierce, D. W., Barnett, T. P., Hidalgo, H. G., Das, T., Bonftls, C., Sat1ter, B. D., ... Nozawa, T. (2008). Attribution of declining western U.S. snowpack to human effects. ]o11ma/ q{Ciimale, 21 (23), 6425-6444. Pontius, D. S., Inc. (1997). Colomdo n/Jer ba.rin .rl11cj)'Tuscan, .-\Z: Western Water Policy Review Ad,,ison Commission. Portman, i\L (2009). lmolving the public in the impact assessment of offshore renewable energy facilities. Mmim Poliq, 33(2), 332-338. Provan, K., & 1\Iilward, H. B. (1995). A preliminary theory of interorganizational net\vork effecti,eness: A comparative study of four community mental health systems ldmini.rlmlil'e Stiel?t'e "Qumtnfy. -10(1 ), 1-33. Reed, M. S. (2008). Stakeholder participation for environmental management: .\ literature review. Biological Con.remalion, 1-11 (1 0), 2417-2431. Renn, 0., \'V'ebler, T., & Weidemann, P. (1995). EtimeJJ and compdwce in a11::;_en pmticipation: etl(i/1/alillg model.r jor emlironmental di.rt'ol! Boston, 1\L\: Kluwer Academic Publishers. Rocldoff, S. F., & Moore, S. A. (2006). Assessing representation at different scales of decision making: Rethinking local is better. Polio Studie.r)ot!mal, 3-1(4), 649-670. Rosener, J. B. (1978). Citizen participation: Can we measure its effectinness? P11blic AdiJiini.rtralion Rniew, 38(5), 457-463. Rowe, G., & Frewer, L. J. (2000). Public partlctpation methods: A framework for e''aluation. Sdmce Tedmofo_g) H11man 1 'al11e.r, 25(1), 3-29. Sabatier, P. (Ed.). (2007). Tbe01ie.r ql tbe policy proce.rJ (2nd Ed.). Boulder, CO: Wcsnicw Press. Sabatier, P., & Brasher, A. l\1. (1993). From ngue consensus to clearly differentiated coalitions: environmental policy at Lake Tahoe, 1964-1985. In P. Sabatier & H. Jenkins-Smith (Eds.), Polio' Change and Leaming (pp. 177 -208). Boulder, CO: Westview Press. Sabatier, P., Hunter, S., & 1\IcLaughl.i.n, S. (1987). The devil shift: Perceptions and misperceptions of opponents. IF'e.rlem Politimi"Qmnterfy. -10(3), 449-476. Sabatier, P., & Pelkey, N. (1990). Lmd de11elopmenl a/ Lake Tahoe, 1960-8-1: 'Jilt' qj' emironmental conlrol.r and emnomic conditioi!J on hou.ring con.rlmdion. Ft. Lauderdale, FL: FAU/fll1 Joint Center for Environmental and l'rban Policy. 239

    PAGE 259

    Sabatier, P., & Weible, C. M. (2007). The AdYocacy Coalition Framework: lnnoYations and clarifications. In P. Sabatier (Ed.), Tbe01ieJ the Poliq ProceJJ (2nd Ed.). Boulder, CO: West'.iew Press. Saba tier, P. "\. (1999). Tbeo1ieJ poliry protUJ (1st Ed.) .. Boulder, CO: Westview Press. Sabatier, P. A., & Jenkins-Smith, H. C. (Eds.). (1993). Polit)' t'bange and learning: An ad!'ocatJ! t'oalition approadJ. Boulder, CO: West'.iew Press Sanders, L. D., Walsh, R. G., & Loomis, J. B. (1990). Toward empirical estimation of the total nlue of protecting rinrs. IJ7ater Re..-oltn'eJ Rmarcb 26(7), 1345-1357. Santos, S. L., & Chess, C. (2003). Evaluating citizen advisory boards: The importance of theory and participant-based criteria and practical implications. RiJk Ana(yJiJ, 23(2), 269-279. Schneider, A., & Ingram, H. (1993). Social construction of target populations: Implications for politics and policy. Amelimn Politiml Sdwce Re1Jiew, 87(2), 334347. Schneider, A. L., & Ingram, H. (1997). deJignfor Lawrence, KS: University Press of Kansas. Schneider, 1\1., Scholz, .J ., Lubell, l\1., Mindruta, D., & Edwardsen, l\1. (2003). Building consensual institutions: Nenvorks and the national estuary program. }o11mal Stimt'e, ..,17(1), 143-158. Scott,J. (1991). Sotial netu,ork ana(piJ: A handbook. London, England: Sage Publications. Smith, P. D., & i\lcDonough, l\1. H. (2001). Beyond public participation: Fairness m natural resource decision-making. Sode(J' and l\'atural ReJoumJ, 1..,1(3), 239-249. Susskind, L., & Cruickshank, J. (1987). Breaking the impaJJe. ConJellJIIal approat'beJ to!ing p11blic di.rfJIItt'J. New York, NY: Basic Books. Susskind, L., I\IcKearnan, S., & Thomas-Larmer, J. (Eds.). (1999). Tbe t'OilJenJIIJ b11ilding handbook: t'o/Jiprebt'JIJ'il'e guide to /'Cat-bing agreement. Thousand Oaks, CA: Sage Publications. Truman, D. B. (1960). Tbr gor'mlllJelltal proceJJ polilic'lll intere..-tJ and public opinio11. New York, NY: Alfred A. Knopf, Inc. 240

    PAGE 260

    Tuler, S., & Wehler, T. (1999). Voices from the forest: What participants expect of a public participation process. 5 ocif(J' and j\'atural ReJomn:x: An fntrmational }olfnwl. 12(5), 437-453. Valelly, R. IVI. (1993). Public policy for reconnected citizenship. In H. Ingram & S. R. Smith (Eds.), P11blic Polia for \'V'ashington, D.C.: The Brookings Institution. Viswanathan, l\L, Ammerman, A., Eng, E., Garttehner, G., Lohr, K. N., & Griffith D. (2004). Collll111flli(J-btmd pa11idpatoo n.,-ecm-h: l.uu.,ing tbe Rochille, l\ID: Agency for Healthcare Research and Quality. Vogenbeck, D. M. (2005). Social network analysis for policy design: CollaboratiYe discourse between nonprofit/ gmernment organizations and the resulting effect on community leYel social capital. Digital Di.rxe11atiou.r, 66(07). RetrieYed September 7, 2010 from ProQuest Digital Dissertations database. (Publication No. AAT 3184196) Wagle, U. (2000). The policy science of democracy: The 1ssues of methodology and citizen participation. Polia Scinhu, 33(2), 207-123. Walters, L., Aydelotte, J., & Miller, J. (2000). Putting more public in policy analysis. P11blit" Admi11i.rtratiou Ret,ieJP, 60( 4), 349-359. Wasserman, S., & Faust, K. (1994). Social mtwork anajpiJ: Metbodr a11d applicationx. New York, NY: Cambridge University Press. Water Colorado (2008). BI!J', rent, and Jell water 1igbtJ in Colorado. Retrie,ed September 7, 2008, from http:/ / \Veber, E. P. (1000). A. New Yanguard for the emironment: grass-roots ecosystem management as a new environmental moYement. Sode!_y a11d 1\:a!11ml Re.rolfrte.f. 13(3), 237-259. Wehler, T. (1995). "Right" discourse in citizen participation: An enluatiYe yardstick. In 0. Renn, T. \Vebler, & P. Wiedemann (Eds.), l-"0irmJJ a11d Co111petfnce in Pmtidpatioll: E1Jal11aling ModeiJjor Emiro11mental Di.1comJe (pp. 35-86). Norwell, l\L\: Kluwer Academic Publishers. Wehler, T., & Renn, 0. (1995). "-\ brief primer on participation: philosophy and practice. In 0. Renn, T. Wehler, & P. Wiedemann (Eds.), l-'-aimeJJ and Colllpetnw in Citi::;_m Pmtiripation: E!'al11aling Mode/.r .for Emironmental DiJt"Oltr.J"e (pp. 1 7 -33). Norwell, l\L\: K.luwer Academic Publishers. 241

    PAGE 261

    Wehler, T., & Tuler, S. (2000). fairness and competence in cmzen partlctpation: Theoretical reflections from a case study ldmini.rlration and Sode(y, 32(5), 566595. Weible, C. Jvl., Sabatier, P., & Lubell, l\1. (2004). A comparison of a collaborative and top-down approach to the use of science in policy: Establishing marine protected areas in California. StlldieJ}om"'wl, 32(2), 187-207. \X.'eible, C. l\1., & Sabatier, P. A. (2005). Comparing policy networks: Marine protected areas in California. Polio StudieJjollma/, 33(2), 181-201. Weiss, J. (2000). Ci,,ic pa11idpation and Jmart growth: Transjomting Jprawl into a broadl'r citi::;_eiiJbip. Retrieved from http:/ / 4_ Civic_Participation_&_SG. pdf Welsh, i\1. (2004). Fast-forward to a participatory norm: .\gency response to public mobilization over oil and gas leasing in Pennsylvania. State and Local GoJ,enmJeJJt Rm.ew, 36(3), 186-197. White, C. (1997). Prospects for ci,ic deliberation in the information age. Sotial St11dieJ, 88(1 ), 23-28. Wittfogel, K. A. (1957). 01imtal de.rpotiJm: A comparatil'e .itucfy New Hann, CT: Yale Universin Press. Zafonte, l\1., & Sabatier, P. (1998). Shared beliefs and imposed interdependencies as determinants of ally nenvorks in overlapping subsystems. Journal Tbeorelical PoliticJ, 10(4), 473-505. Zafonte, l\1., & Sabatier, P. (2004). Short-term versus long-term coalitions in the policy process: Automotive pollution control, 1963-1989. Poliq StlldieJjoHmal, 32(1), 75107. 242