Deliberative Democracy in Theory and Practice: Connecticut s Medicaid Managed Care Council

Transcription

1 Deliberative Democracy in Theory and Practice: Connecticut s Medicaid Managed Care Council Colleen M. Grogan, University of Chicago Michael K. Gusmano, Columbia University abstract Despite calls for greater deliberation among citizens on public policy, we have little information about how existing deliberation is structured or how well it works. We examine Connecticut s effort to put public deliberation to work in Medicaid policymaking. Findings from our participant-observation study and in-person interviews with 100 participants in this process suggest some important qualifications to literature on public deliberation. Greater inclusion of diverse social groups from the target population is important, but this should not replace the inclusion of professional advocates since the latter are often more willing than citizen representatives to challenge policy experts on technical issues. Incorporating public deliberation into the process at an early stage is ideal, but deliberation during the later stages of policymaking (including during implementation) can still produce useful results. Finally, the style and purpose of deliberation can shape the range of topics on the agenda, so it is important to understand how the structure of a deliberative forum can affect the style and purpose of deliberation. The essence of democracy itself is now widely taken to be deliberation, as opposed to voting, interest aggregation, constitutional rights, or even selfgovernment. (Dryzek 2000, 1). As Dryzek suggests, there is growing agreement among scholars and practitioners that citizen deliberation benefits the policymaking process, but much debate remains about how best to establish and structure such deliberations. Should they be open to the general public or only to carefully chosen invitees? If the latter, who should be invited to participate? What topics should be discussed and how should discussions be facilitated? These questions are significant because public deliberation can take many forms. Gutmann and Thompson (1996, 12 3) advocate taking advantage of this variation: State Politics and Policy Quarterly, Vol. 5, No. 2 (Summer 2005): pp by the Board of Trustees of the University of Illinois

2 summer 2005 / state politics and policy quarterly 127 The forums of deliberation... embrace virtually any setting in which citizens come together on a regular basis to reach collective decisions about public issues governmental as well as nongovernmental institutions. They include not only legislative sessions, court proceedings, and administrative hearings at all levels of government but also meetings of grass roots organizations, professional associations, shareholders meetings, and citizens committees in hospitals and other similar institutions. Because public policymaking deliberations can occur in so many different settings, decisions about how to structure a particular deliberation may be customized to fit the political context. For example, a deliberative structure that worked well in a legislative session may not be as effective for forums on the same issue held in administrative hearings or meetings of grassroots organizations. While there is growing empirical literature on deliberative democracy (Abelson et al. 2003; Hayward 2003; Ravenscroft, Curry, and Markwell 2002; Sunstein 2000), there are some political settings in which deliberation remains largely unexplored. State government is one such setting in which efforts to implement deliberative forums have not been studied. We address this gap in the literature by presenting a case study of one state s effort to put deliberative democracy to work. We study deliberation among stakeholder representatives in the design and implementation of health care policy for poor families in Connecticut in the mid-1990s. Concurrent with the trend toward increased deliberation in policymaking (Button and Mattson 1999; Bohman and Rehg 1999; Cohen 1998; Daniels 1999), there has been a general devolution of policy authority in the United States (Donahue 1999; Nathan 1996; Peterson 1995). As the states have been given more discretion in domestic public policy in areas such as social welfare, health, and environmental policy (Thompson and Dilulio 1998; Winston 2002; Beierle and Konisky 2000; Ravenscroft, Curry, and Markwell 2002), there has been growing pressure to set up more publicly deliberative processes to inform state policymaking (Thomas 1995; King 1998; Kathlene and Martin 1991). Given these trends, it is important to understand how deliberative forums are structured and how well such forums are working in the states. Does the theoretical literature offer a useful guide to state policymakers in pursuing deliberative democracy? Can the practice of deliberative democracy in the states help inform and test theories of deliberation? As a single-state case study, the generalizability of our conclusions may have limits (Nicholson-Crotty and Meier 2002), but our study makes two significant contributions. First, we provide a framework, based on theories of deliberative democracy, that can help state policymakers think systematically about implementing deliberative forums for stakeholder representatives.

3 128 grogan and gusmano In particular, we discuss five central elements states should consider when setting up a deliberative process. Second, our study offers some preliminary insights to help guide states in structuring their deliberative forums. Following Button and Mattson (1999, 610), we seek to show what actual efforts at deliberation reveal to democratic theorists and, conversely, inform the practice of deliberation with the concerns of democratic theory. a framework for structuring state-level deliberative democracy At the heart of arguments in favor of deliberative democracy is the notion that public input into policymaking should be more than the simple aggregation of individual interests. Instead, deliberation encourages decisionmaking by discussion among free and equal citizens and, through such deliberation, opinions can be transformed through the active exchange of ideas (Habermas 1984, 1987; Young 2000; Gutmann and Thompson 1996; Macedo 1999). While this notion that individual ideas, values, and world views can change through discussion is a central idea underlying deliberative democracy, it does not have to be the only goal of deliberation. Indeed, deliberative theorists have suggested a range of potential benefits that can be summarized into five main outcomes (Bohman and Rehg 1999; Elster 1998, 11; Macedo 1999). First, regardless of whether deliberation changes individual preferences, it may create a greater understanding and tolerance for opposing views (Gutmann and Thompson 1996). Second, through discussion, deliberation can encourage a public-spirited way of thinking about social problems. By exchanging ideas and beliefs, individuals may begin to think about their community, their state, or their country, and not just their own self-interest (Gutmann and Thompson 1996; Macedo 1999). Third, deliberation may help participants clarify and refine their own positions on issues, and, as a result, it may offer new solutions to solve social problems (Disch 1995; Fearon 1998). Fourth, some theorists have suggested that deliberation can help democracies create better public policy (Young 2000; Gutmann and Thompson 1996). For example, Gutmann and Thompson (1996) argue that good deliberation can help societies correct past mistakes. Through deliberation, participants can learn about the consequences of implemented policies and the reasons for past failures and, thereby, can develop solutions that avoid and correct past mistakes. Fifth, regardless of the quality of policy outcomes, theorists argue that decisions emerging from an open and inclusive deliberative process should be considered more democratically legitimate than decisions

4 summer 2005 / state politics and policy quarterly 129 emerging from a representative government process where participation is systematically low among vulnerable populations (Gutmann and Thompson 1996; Elster 1998; Young 2000). But most deliberative theorists agree that without clear and accepted criteria and procedures to structure a public deliberation, few, if any, of these potential benefits will be realized. This is because the existing social and political power dynamics under which people are not treated equally can influence the public conversations causing certain groups to feel marginalized and to not participate fully in the discussion (Bohman and Rehg 1999; Elster 1998; Gutmann and Thompson 1996; Macedo 1999; Young 2000). Deliberative democracy institutions need to structure the process so that a buffer exists between the larger political environment and the process of deliberation thereby protecting the discussion from the influences that lead to inequality among deliberators (Figure 1). Of course, any structure can serve only as a partial buffer between external political factors and the deliberative forum. The participants are a product of the larger political and social environment and the deliberations emerge out of a political process in the external world (Bessette 1994; Button and Mattson 1999; Gargarella 1998; Habermas 1984, 1990; Johnson 1998; Mackie 1998; Przeworski 1998; Ross 1995; Simon 1999; Stokes 1998; Young 2000). And even if we could create an isolated deliberative forum, such a contrived process might suffer because participants may feel irrelevant to the larger political process and lack the motivation to struggle through difficult discussions. Nonetheless, a buffer is necessary to create a deliberative environment where marginalized or previously excluded groups feel comfortable voicing their concerns. Figure 1. The Dynamics of Deliberative Democracy Political Environment Criteria for the Structure of Deliberation Potential Benefits Political Context Social relationships Social inequities External Shocks Elections New legislation New information Equality Inclusion Openness Publicity Purpose and style Tolerance Public-spiritedness New solutions Just outcomes Policy legitamacy

5 130 grogan and gusmano Based on theories of deliberative democracy, we identify five criteria for successful deliberative structures: equality, inclusion, openness, publicity, and purpose and style (Figure 1). In this section, we not only present these criteria as the ideal types discussed in the theoretical literature on deliberative democracy, but we also discuss some of the practical difficulties involved in their implementation. Theoretically, if a deliberative structure meets these criteria, the potential benefits of deliberation discussed above may be realized. 1 Equality Participants in the deliberative process should have an equal right and an equal opportunity to voice their concerns (Elster 1998; Cohen 1998). 2 Attempts to equalize political resources are important to avoid what Stokes (1998, 136) describes as the deliberative pathology of powerful special interests manipulat[ing] common citizens notions of what they want the government to do. One of the central difficulties in creating an open and honest public discussion is that participants come from unequal power positions in society (Stokes 1998; Przeworski 1998). A particularly perplexing problem is how to set up a deliberative process in which people feel equally empowered to participate given that they feel unequal (and, in fact, are unequal in terms of wealth, education, and social status) in the greater society. Some deliberative theorists offer suggestions for how to equalize participation in deliberative processes (Cohen and Rogers 1995; Schmitter 1995). First, all participants should be given equal access to political and policy information (Stokes 1998). Second, all participants should be given an equal opportunity to set the topic agenda for discussion (Young 2000). Third, all participants should be given an equal opportunity and equal time to voice their views and concerns (Young 2000). Fourth, all forms of talk as long as it is not disrespectful of another participant should be allowed so that those from less powerful groups, who often have different styles of speech than those from more powerful groups, do not feel excluded (Dryzek 2000). Finally, the deliberative body should specifically discuss power differentials in the greater society to lay bare the distinctions that exist within the deliberative forum (Young 2000). Inclusion Participants from all groups directly affected by the policy under consideration should be included in the deliberative process (Young 2000; Cohen 1998; Gutmann and Thompson 1996). While proportional representation

6 summer 2005 / state politics and policy quarterly 131 of groups may not be possible or desirable, the full range of views should be represented in the deliberation to avoid bias and distortion (Gargarella 1998). This inclusion should reflect a concerted effort to involve groups that are typically marginalized by more formal political decisionmaking (Young 2000). While inclusion is important to developing a productive democratic deliberative process, there is no easy definition of a perfectly inclusive body (Young 2000; Gargarella 1998; Gutmann and Thompson 1996; Cohen 1998). There are always more perspectives that could be included in a process, but, at some point, a deliberative body becomes overly inclusive and, as a result, unwieldy and unmanageable (Young 2000). In addition to representatives from groups directly affected by the policy under discussion, Young (2000, 1990) argues that representatives from social groups (such as racial, ethnic, or gender groups) as distinct from economic interest groups, should be included in a deliberation because they reflect structural inequities in society. Others contend that simply including social groups in the deliberative process is not enough to satisfy concerns of inclusion, but rather participants should also explicitly discuss social group differences (McBride 1993, 330 1; Disch 1996, 22). Openness Achieving openness means that every stage of the decisionmaking process is deliberative, from the formation of policy options to the implementation of policy solutions. Without openness, scholars worry that deliberations will be used only after important decisions have been made as a token exercise to pacify dissent (Disch 1996; Young 2000). While this is certainly a legitimate concern, complete openness is difficult to achieve in practice; any deliberative process will emerge out of an already established political environment, so participants must respond to and accept previous policy decisions. Implicitly acknowledging this difficulty, some scholars advocate for deliberation as late in the policymaking process as the planning and implementation stages (Jennings 1994). For example, Daniels (1999) argues that Health Maintenance Organizations should use deliberative processes to aid in controversial medical decisionmaking even at the micro-level of implementation. Similarly, Gould (1996) contends that for medical allocation policy, while elected bodies should set the broad structure, specific trade-off decisions should be determined through a deliberative process at the local level. But when the openness criterion is not completely met, participants should explicitly acknowledge prior policy decisions under which certain groups were excluded (Young 2000).

7 132 grogan and gusmano Publicity Deliberative processes should be well publicized to insure democratic accountability (Gutmann and Thompson 1996; Rawls 1997). The publicity of deliberations both broadens participation and allows the public to hold representatives accountable for the reasoning and moral justifications behind their decisions. On the other hand, while most deliberative theorists agree with this evaluation, some argue that excess publicity can be damaging to the honest exchange of ideas in a deliberative process (Przeworski 1998). With too much publicity, participants may avoid exploring new ideas for fear they may say something extemporaneously that they will be held accountable for in the future. Thus, publicity is necessary for democratic accountability but the appropriate level of publicity is unclear. Purpose and Style Scholars and advocates have suggested various purposes and styles for public deliberation. The purpose and style of a deliberation sets the tone and criteria for evaluating that process. Button and Mattson (1999, 612 3) summarize four models for processes in the deliberative democracy literature as shown in the quote below (see also Jacobs, Cook, and Carpini 2000). Educative: The main purpose of this type of public deliberation is to provide participants with more information and knowledge as a means of encouraging political learning about an issue or problem. Consensual: This approach stresses procedures by which participants can come to a common agreement on an issue, values, or the direction of a future course of action. Activist/Instrumental: This approach to deliberation stresses direct political or legislative results as the purpose and end of public discussion. Conflictual: This approach emphasizes giving the widest possible space to the expression and development of individual points of view without being constrained by other demands on public talk. This conception stresses conflict and difference over resolution and agreement. The purpose and style criterion is the least well-specified in deliberative theory. It is unclear whether purpose and style should be articulated as organizing frames to help govern the deliberative forum or whether they should emerge from the deliberation itself. Nonetheless, Button and Mattson (1999, 613) find that the purpose and style of a deliberation are important because they reflect very different goals and, as a result, have a significant influence upon the form and substance of [the] deliberation.

8 summer 2005 / state politics and policy quarterly 133 While this brief summary sheds light on these criteria for a successful public deliberation, not one of them is defined with great clarity in the theoretical literature. Below, we explore how well these criteria were met in one state-level deliberative forum to help us understand their relative importance and usefulness. deliberative democracy and medicaid reform in connecticut Our case study of deliberative democracy examines the deliberations on how to design and implement a reform of Connecticut s Medicaid program. In the early 1990s, Connecticut s Department of Social Services (DSS), the agency responsible for administering the state s Medicaid program, began to plan in earnest for Medicaid Managed Care (MMC). This new program was intended to reduce Medicaid costs to the state, while increasing the quality of medical care Medicaid recipients received and improving their access to health care (Hurley and McCue 2000). DSS believed that Managed Care Organizations (MCOs) could be used to encourage physicians to participate in the Medicaid program (Holahan et al. 1998). For years, state officials had been frustrated by physicians low participation rates, particularly among pediatricians and obstetricians. They believed that MCOs could expand physician participation by, in theory, making participation in their private insurance networks contingent on the willingness of physicians to accept the plan s Medicaid patients. Because of his strong belief that the state should soon implement MMC, Connecticut s Medicaid Director, David Parella, created a Medicaid managed care advisory committee in 1993, which was primarily comprised of representatives from the managed care industry. The primary purpose of the advisory committee was to provide DSS with feedback about the design of the new program and, in so doing, to generate among MCOs interest and commitment to participate in the program. The makeup of this advisory committee was intentionally limited to representatives of MCOs and Community Health Centers (CHC) 3 and included only one Medicaid recipient advocate. 4 DSS did not want to involve a broader set of actors until they had developed the basic design of the program. DSS officials believed that opposition to managed care from Medicaid recipient advocates would have discouraged MCOs from participating in MMC. Clearly, by design, this advisory committee did not represent an open or inclusive deliberative process. Excluded groups criticized this closed process, asserting that it was biased toward the medical industry and insufficiently took into account the views

9 134 grogan and gusmano of other stakeholders in Medicaid. Toward the end of 1993, DDS responded to this criticism by dissolving the advisory committee and embarking on a public relations campaign. As part of their campaign, DDS held a conference on MMC, which was attended by over 200 people, and they met with more than 300 organizations, including other state agencies, provider groups, and advocacy organizations, to answer questions about the reform proposal. The purpose of this public relations campaign was to persuade, with no effort to involve these groups in a discussion about the policy s design. Not surprisingly, Medicaid advocates and Medicaid providers continued to feel dissatisfied with the decisionmaking process, and they made their views known to key persons in the state legislature. In response to these concerns, the legislature gave DSS authority to submit a waiver to the federal government to implement an MMC program in 1995, but the legislature also prohibited that implementation for one year, so that a newly created 26 member Medicaid Managed Care Council could deliberate on its design and implementation. 5 DSS was required to provide the MMC Council with monthly reports on its plans for implementing MMC, and the Council was charged with discussing these proposals and advising DSS. The MMC Council met monthly; Council members included Medicaid advocates, provider representatives, health plan executives, Medicaid consumers, state legislators, and state agency representatives. The Council also developed four subcommittees with much wider participation among the same types of individuals focusing on different issue areas: Access, Quality Assurance, Public Health, and Behavioral (Mental/Substance Abuse) Health. These subcommittees also held monthly meetings. Case Study Methods Our primary methods of investigation were close participant observation of the Connecticut experience with deliberative democracy in Medicaid reform and in-person interviews with participants in that process. Our participant observation occurred over an 18-month period, from December 1995 to May 1996, during which time we attended 67 public meetings. Of the 103 participants in the process, we interviewed 100 for a response rate of 97 percent. 6 When it established its deliberation process, Connecticut s legislature did not explicitly define the structure of its deliberative forum. Only a definition of inclusion was specified in the legislation, which mandated the stakeholder groups that should be represented in the deliberations. This lack of mandated structure allows our case study to be more generalizable to similar deliberative forums set up by other state and local governments,

10 summer 2005 / state politics and policy quarterly 135 where the criteria that deliberative theorists suggest to be important are not typically adopted either. This lack of mandated structure allows us to explore the following questions: What elements of deliberative structure develop in deliberative processes intuitively? What elements are difficult to implement, and how would prior planning and structure for them help deliberation? As Button and Mattson (1999, 613) argue, developing an understanding of how citizens [or, in this case, stakeholder representatives] actually approach political deliberation and decision-making within the structural features of a representative system requires turning to those experiments where these relationships are being explored. In this spirit, we turn to the history and activities of those on and around the MMC Council to document and assess how well it met the five criteria discussed herein. findings Equality As the history of the MMC Council indicates, Connecticut s deliberative process for Medicaid reform was created specifically because of the inequality in decisionmaking power and the exclusion of key stakeholders in the previous policymaking process. One of the central difficulties in creating an effective deliberative process is that its participants come from the inequality of the preexisting power structure embedded in society. As a result, a particularly perplexing problem is setting up a process in which people feel equally empowered to participate even though they may be unequal in political, social, or economic terms in the greater society. Connecticut s MMC Council appeared to give all participants equal access to talk and an equal chance to influence the agenda. For example, persons indicated their willingness to talk by raising their hand or signaling to the Council chair in some other way. Participants who signaled were always given the opportunity to speak. Indeed, in the meetings we observed, each participant talked at some point. Participants, at least in part, also generated the topic agenda for the meetings. Any participant could simply ask the chair to put an item on the agenda, and it was always done. While these procedures made equal access the overt norm, nothing was done explicitly to identify the existing power differences between group participants or to reduce the potential effects of these differences. Since this deliberative process emerged out of a fight for representation, some groups had been excluded from the decisionmaking process and wanted a greater voice. Our interviews revealed that certain group representatives felt unequal

11 136 grogan and gusmano as they entered the deliberative body. This feeling of a lack in power was mentioned primarily by participants who felt that their groups had been excluded from the decisionmaking process originally. Thus, although the deliberative process formally allowed for equal access to the floor and the agenda, it did not attempt to compensate for these feelings of exclusion through a public acknowledgement of them or some other mechanism. Our interview data suggest that, as a result, some participants consciously and strategically decided not to give voice to certain concerns. This was not because they lacked formal power relative to other participants in the deliberative process, but because they strategically decided that these concerns were not palatable in the broader political environment. On the topics participants chose to discuss publicly, we observed a fairly equal exchange of ideas and concerns. Both the presence of a state senator as the chair of the Council and bureaucrats from DSS as potential topic initiators might have been expected to create a significant differential and deference among participants in terms of political power and technical expertise (Button and Mattson 1996). Yet, even on technical issues, such as how the state should set payment levels for MCOs, Medicaid advocates and providers were active and challenged participants in the deliberation. On the topics that participants reported being strategically important to them, they fought hard and debated on an equal footing. Even professionals acting on behalf of Medicaid recipients were capable participants, who, unlike the citizens in other studies (Stokes 1998; Button and Mattson 1996), did not defer to DSS or legislative members, 7 and the public discussion rarely consisted of a substantively vacuous question-and-answer session. This suggests that attempts to compensate for societal inequalities among deliberation participants in ways other than giving everyone formal procedural equality, would not be a worthwhile investment for state actors setting up a deliberative forum. 8 Inclusion Despite the long-run advantages of including many stakeholders and considering many points of view in a deliberative process, the level of inclusiveness desirable or practical is always a matter of degree (Young 2000; Cohen 1998; Gutmann and Thompson 1996; Gargarella 1998). There are always more groups and individuals with a stake in the discussion of a state-level public policy than can be included practically in any given deliberation. Continual expansion of inclusivity in a deliberative body will, at some point, result in an unwieldy and unmanageable process. Despite the inherent tension on this criterion, Connecticut s MMC

12 summer 2005 / state politics and policy quarterly 137 deliberation was reasonably inclusive, perhaps because excluded groups had fought for their inclusion. Participants in the MMC Council included the range of groups affected by, and interested in, Medicaid policy, including provider groups (physicians, hospitals, community health centers, mental health and substance abuse services providers, and family-planning groups), the target population (Medicaid consumers, as they were called), advocates representing the target population, state legislators with a particular interest in Medicaid policy, representatives of state agencies administering programs related to Medicaid, and MCO representatives. A variety of conflicting interests existed among these participants, many cutting across ideological and economic lines. Yet it was inevitable that, given the inherent limits of inclusiveness, there were areas where inclusivity fell short. For example, the direct representation of Medicaid recipients was limited. At many meetings, only one Medicaid recipient attended, and this was the Caucasian mother of a disabled child. While her perspective was valuable and contributed to the deliberation, given the demographics of the recipient population, the input of African-American and Hispanic mothers on cash assistance would have been useful, particularly in discussions of whether recipients from various ethnic and racial groups were making informed health care choices. Another aspect of inclusiveness is whether the stakeholder representatives in the deliberation reflected accurately the views of the groups they were meant to represent (Young 2000; Gutmann and Thompson 1996). This became a significant issue in the Connecticut case, as some participants questioned both publicly and in our private interviews how well the participating Medicaid advocates represented the interests of the poor. Medicaid advocates primarily came from the social service agencies that provided services to Medicaid recipients. Some participants believed that these advocates were primarily concerned with their own self-interest, fighting for more services (or increased payment levels or policies to improve quality) to increase the financial well-being of their agencies rather than the medical well-being of Medicaid recipients. This highlights the importance not only of the breadth of inclusion but also of the quality of this inclusion. Can representative members of a target population ever effectively replace advocates for that population in a deliberative policymaking process? Members of most target populations probably are able to present their concerns effectively, except perhaps when nuanced, technical issues are under consideration. Citizens typically defer to policymakers and bureaucrats on technical issues in such deliberations (Stokes 1998; Button and Mattson 1996). Given the multiple demands on peoples time, it may be unrealistic to expect lay

13 138 grogan and gusmano citizens, even from an affected population, to acquire technical knowledge about the mechanics of public policy. Professional advocates for the target population have more time and resources to invest and, therefore, may be better equipped to discuss more technical issues. For example, the Medicaid advocates in our study were particularly articulate in discussing the technical issues regarding Medicaid access and payments to MCOs; such discussion might have been more difficult for a lay person or Medicaid recipient. These advocates appreciated the importance of policy details and empirical data, and they often used the deliberative forum as an opportunity to better understand certain aspects of the program. Yet, they were not just passive learners. After receiving information, they would often shift into a more conflictual mode to make a point. Despite the complexity of an issue, Medicaid advocates were unrelenting, continuing to delve into the meaning of data presented in a self-confident manner that non-advocate, rank-and-file Medicaid recipients might not possess, due to a lack of technical training and the existing social power structure. An excerpt from the deliberation illustrates this point: Provider Representative: I m not confident that rate is right. DSS Consultant: I feel very confident that this rate is right. Provider Representative: We are going to expand access based on this rate, and I m not convinced the plans don t have something to be concerned about. Medicaid Advocate #1: Someone suggested rates would go up, not down. MCOs were counting on an increase in the rates. Help me understand why these differences exist. Help the layperson understand. DSS Consultant: It is very complicated; there are so many factors... Medicaid Advocate #1: Okay, I know, it s very complicated, and that it is beyond me, but help at the policy level. [For example,] how is Fairfield County $123 and Tolland County $179? What does this tell us? DSS Consultant: We can quantify the numbers and show that these numbers are persisting year after year. You should not infer poorer access where the rates are lower. A lot of research still cannot answer why such variations occur. Medicaid Advocate #2: But this does reflect differences in access. Do we want to pay based on regional variations that currently exist? Do we want to perpetuate those variations? The above representative exchange demonstrates that these Medicaid advocates were not afraid to challenge technical experts even when they acknowledged their lack of full understanding and information. Contrary to findings of deliberation among the general public (Button and Mattson 1999), advocates for the poor in our study did not defer to technical experts.

14 summer 2005 / state politics and policy quarterly 139 Openness A deliberative process is open when deliberation is incorporated into every stage of policymaking, from the development of policy options to its adoption and implementation (Disch 1996; Young 2000). Connecticut s MMC deliberation process clearly was not open at every stage. As discussed above, the initial period of policy option formulation was significantly closed, with a clear bias toward industry participation. By the time the broader deliberative structure was created, clear boundaries had been placed around the range of policy options for discussion. When the MMC Council was established, the state had already decided that Medicaid managed care would be the favored approach and that private insurers would be key participants. There was little public deliberation about the adoption of MMC in the state legislature. The enabling bill was passed late in the 1995 legislative session under broader pressure to balance the budget. Indeed, this lack of legislative deliberation was interpreted by the vast majority of our interviewees as the result of a political environment that was not open to a variety of points of view. Our interviews revealed that, as a result of this closed initial phase of the process, most MMC Council members who were opposed to, or at least concerned with, MMC came to the deliberation skeptical that the state administration genuinely cared about Medicaid service provision. These participants interpreted the bill s lack of legislative debate and its link to budget politics as signals that MMC was primarily a mechanism to control costs and not a policy designed to improve access to, or the quality of, medical care for Medicaid recipients. Did this skepticism about the state government s intent lead these participants to view the deliberative process as a token gesture giving them little hope for an impact on Medicaid policy? In fact, despite this initial skepticism, the majority of the participants we interviewed came to believe that the deliberation process was legitimate and useful. While a few thought the process was a waste of time, most believed that the design and implementation of MMC had been improved by the deliberative process. In evaluating our interview responses, we must acknowledge the likely bias that those participating in the deliberation believed that deliberation after policy adoption was a worthwhile activity. That is, those who thought deliberation at this stage was a token exercise were likely not to have been asked to participate on the MMC Council and, if asked, they were more likely to have declined the invitation. Additionally, if a skeptical person did make it onto the Council, he or she was probably less likely to participate on the subcommittees. Although some vocal opponents of MMC were asked

15 140 grogan and gusmano to serve on the MMC Council, those participants who agreed to do so likely still believed they could make a difference. Reflecting this bias, we found that the majority of active participants believed their participation could lead to change, whereas less active participants were much less optimistic about their effectiveness. An important question, then, is whether reduced openness reduces the participation and, ultimately, the exclusion of more dissenting voices in a deliberation. While we did not find significant differences in opinion among active and inactive participants, it is important to acknowledge that a lack of openness may cause a selection bias and this may be common. Nonetheless, while opening the policymaking process to deliberation early is desirable, we found that deliberation can be beneficial even when it begins late in the process. This is important because we suspect that the pattern of introducing deliberation late in the process, as seen in Connecticut, is not unusual. Indeed, many deliberative processes emerge out of political struggle where participants must respond to, and accept, previous policy decisions (Young 2000). As in the Connecticut case, legislation is often passed with vague language to meet political contingencies, and the details are worked out later perhaps at least partially through a deliberative process. Given the likelihood that deliberative processes may often begin in the middle of a policymaking process, theorists should examine more closely how deliberations should respond to previous policy decisions. Publicity While deliberative theorists contend that the policy arguments and moral positions taken by participants in a deliberative process should be made public, the ideal level of such publicity is unclear (Gutmann and Thompson 1996; Disch 1996). Public announcement of these arguments and positions facilitates trust and open deliberation. However, too much publicity can be damaging to the deliberative process if participants limit their exploration of new ideas for fear that they may be held accountable for an offhanded comment or preliminary idea. Connecticut s MMC Council was very public. Its monthly meetings were designed so that only Council members and invited guests spoke, but a public audience of about 50 people, on average, listened to the deliberations. Sometimes the audience included members of the press. Furthermore, anyone could participate in the meetings of the four subcommittees. Subcommittee meetings were held in relatively small rooms and all participants sat around a conference table. While the press often covered various Council meetings and reports, the subcommittee deliberations were more removed

16 summer 2005 / state politics and policy quarterly 141 from public view, but primarily due to a lack of press and public interest. The press covered subcommittee reports that were presented to the Council, but press did not typically attend subcommittee meetings, even though they were open to the public. The fact that some meetings were more widely publicized than others allowed us to compare the discussions that emerged in deliberations with different levels of publicity. In general, we found that an instrumental, decision-oriented deliberative environment emerged in the subcommittee meetings, while an educative orientation emerged in the more publicized Council meetings. Under the public s eye, Council participants proceeded as if on a fact-finding mission with the agreed goal of improving MMC. They invited guests to provide new reports to the Council and enhanced the educative style with publicized question-and-answer sessions. On the other hand, subcommittee participants stressed the creation of policy solutions as the goal of their deliberations. For example, the Quality Subcommittee focused its efforts on creating specific measures and indicators for Medicaid s quality assurance program. The Access Subcommittee created a common administrative form for MCOs to use to encourage more dentists to participate in the program. The lack of publicity and the more intimate setting of these meetings encouraged the solution-focused orientation. The Connecticut case shows that the publicity dilemma is a real concern because different levels of publicity can lead to different types of deliberation. It also illustrates that deliberations can be structured purposefully to include different, but complimentary, types of discussions by establishing settings with different levels of publicity. Purpose and Style Four main models of deliberative style are discussed in the empirical literature: educative, consensual, activist and instrumental, and conflictual (Button and Mattson 1999; Jacobs, Cook, and Carpini 2000). The Connecticut MMC deliberation was not explicitly designed to fit into any one of these orientations. As a result, its style developed largely based on the level of publicity given to the deliberative mechanisms, as discussed above. Wide publicity in the Council meetings created an educative style, while the lack of publicity in the subcommittee meetings encouraged a more instrumental approach. The level of openness also seems to have affected the style and purpose of Connecticut s MMC deliberations. Because the MMC Council was not created until after the legislature had already decided to adopt and implement MMC, a strong consensual norm developed in the deliberation, with the prevailing attitude being that everyone must work toward improving a

17 142 grogan and gusmano program that was going to be implemented regardless of what the Council decided. Our interviewees felt that, at times, this norm quashed criticism of the program because critical participants were labeled as contemptuous of the reform or obstructionist. Indeed, participants tended to view the main purpose of the deliberation process as one of solving immediate MMC implementation problems. The vast majority of participants deliberated with the very instrumental purpose of working on incremental changes to improve a program about which they had concerns. As a result, the conflictual style that Button and Mattson (1999, 612 3) found to provide the widest possible space to the expression and development of individual points of view without being constrained by other demands on public talk never emerged. conclusion We identified five criteria for evaluating deliberative structures based on theories of deliberative democracy: equality, inclusion, openness, publicity, and purpose and style. Here, we summarize the lessons learned from our case study of Medicaid reform in Connecticut for theory and practice for each criterion. First, while most deliberative theorists argue that attention to equality is necessary to compensate for societal inequalities among deliberation participants, they are unclear about what specific mechanisms are necessary for adequate compensation. The mechanisms will likely need to vary with the setting and purpose of the deliberative forum. Our case study suggests that overt, formal equality procedures may be sufficient to compensate (at least to a degree) for societal inequities. Our interviewees suggested that the public acknowledgement of societal inequalities would not have changed their strategic decisions about which topics to discuss. On the topics participants reported to be strategically important to them, they fought hard and debated on an equal footing. Second, we found that the inclusion in the deliberation of representatives of all target groups, especially socially disadvantaged groups, was crucial to achieving a legitimate deliberative forum. In particular, specific questions were often directed to representatives of the recipient target population for opinion and clarification. Indeed, our interviews and observations in Connecticut suggest that greater inclusion of the diverse social groups within this target population would have benefited the deliberation. But we also found that representatives of the target population should not replace professional advocates for them, because the former tend to defer to experts on technical issues whereas the latter do not.

18 summer 2005 / state politics and policy quarterly 143 Third, advocates of openness argue that deliberative forums always benefit by beginning as early in the policymaking process as possible. Nonetheless, we found that opening policymaking to deliberation even late in the process can produce benefits. Thus, close attention to the other four criteria can lead to a legitimate and meaningful deliberation even if the discussion is constrained by prior political decisions. Fourth, while publicity might be important in holding participants accountable to the broader public, we found that the level of publicity to which a deliberation is exposed affects the style and purpose of that deliberation. In our study, high publicity led to a more educative and consensual style, whereas low publicity led to a more instrumental style. Because variation in deliberative style can foster a wider range of discussion, various levels of publicity should be explicitly structured into the deliberative forum. Fifth, the style and purpose of deliberation is important because it affects how and what topics are discussed and what the outcomes of a deliberation will be. We found that without a mandated style and purpose, one will emerge intrinsically from the structure of the deliberation. As discussed, publicity can determine style and purpose, but reduced openness also seems to have an impact. Thus, our study suggests that if a certain style and purpose is desired, the structure of deliberation needs to be developed carefully. In particular, at least for state-level deliberations among representatives, the conflictual style is unlikely to emerge unless particular action is taken to make that happen. endnotes We would like to thank the participants in the Health Policy Seminar at Yale University and three anonymous reviewers for SPPQ for helpful comments and suggestions on this project. 1. Even under an ideal deliberative structure, theorists admit that we should not assume that we can achieve these outcomes in any absolute sense, but we should be able to move closer to realizing such outcomes (Young 2000). 2. Cohen (1998, 186), for example, defines deliberation as free public reasoning among equals who are governed by the decisions. 3. Federal waivers at that time required the inclusion of federally qualified health centers in any type of Medicaid managed care reform. 4. This was Judith Soloman, the Litigation Director at Legal Aid Society of Hartford, Connecticut. 5. These provisions were contained in Connecticut Public Act 94 5, passed in May Due to the time constraints of four interviewees, we were unable to complete the entire interview protocol for them. Hence, we have a total of 96 fully completed interviews. The completed interview response rate is 93 percent.

19 144 grogan and gusmano 7. There is one exception to this generalization that we highlight in discussing the inclusion criterion. 8. On the other hand, this may be worthwhile if citizens are included in the process. Our finding is based on the participation of group representatives who may be less reticent to participate than disempowered citizens. references Abelson, Julia, Pierre-Gerlier Forest, John Eyles, Patricia Smith, Elisabeth Martin, and Francois-Pierre Gauvin Deliberations about Deliberative Methods: Issues in the Design and Evaluation of Public Participation Processes. Social Science and Medicine 57: Bessette, Joseph M The Mild Voice of Reason: Deliberative Democracy and American National Government. Chicago: University of Chicago Press. Beierle, Thomas C., and David M. Konisky Values, Conflict, and Trust in Participatory Environmental Planning. Journal of Policy Analysis and Management 19: Bohman, James, and William Rehg Deliberative Democracy: Essays on Reason and Politics. Cambridge, MA: MIT Press. Button, Mark, and Kevin Mattson Deliberative Democracy in Practice: Challenges and Prospects for Civic Deliberation within a Representative System. Polity 31: Cohen, Joshua Democracy and Liberty. In Deliberative Democracy, ed. Jon Elster. New York: Cambridge University Press. Cohen, Joshua, and Joel Rogers Secondary Associations and Democratic Governance. In Associations and Democracy, eds. Joshua Cohen, Joel Rogers, and Erik Olin Wright. London: Verso. Donahue, John D Hazardous Crosscurrents: Confronting Inequality in an Era of Devolution. New York: Century Foundation Press. Daniels, Norman Enabling Democratic Deliberation: How Managed Care Organizations Ought to Make Decisions about Coverage for New Technologies. In Deliberative Politics: Essays on Democracy and Disagreement, ed. Stephen Macedo. New York: Oxford University Press. Disch, Lisa Publicity-Stunt Participation and Sound Bite Polemics: The Health Care Debate Journal of Health Politics, Policy, and Law 21:3 34. Dryzek, John S Deliberative Democracy and Beyond: Liberals, Critics, Contestations. New York: Oxford University Press. Elster, Jon Introduction. In Deliberative Democracy, ed. Jon Elster. New York: Cambridge University Press. Fearon, James D Deliberation as Discussion. In Deliberative Democracy, ed. Jon Elster. New York: Cambridge University Press. Gargarella, Roberto Full Representation, Deliberation, and Impartiality. In Deliberative Democracy, ed. Jon Elster. New York: Cambridge University Press. Gutmann, Amy, and Dennis Thompson Democracy and Disagreement. Cambridge, MA: Belknap Press of Harvard University Press.