Variations of Institutional Design for Empowered Deliberation

This paper lays out the practical and theoretical characteristics of formally empowered deliberation as a distinctive subset of deliberative processes. As part of a recent broad shift toward a more deliberative conception of democratic politics, participatory deliberative processes increasingly have been formally empowered as part of democratic governance. Governments have moved to delegate authority and deliberative responsibility from elite bodies to lay publics more quickly than scholars have been able to fully identify the implications of this institutionalization for the quality of both deliberation and democracy. This paper describes the emerging characteristics of formally empowered deliberation as a distinctive subset of deliberative processes, in which deliberation between members of the general public is given credible formal authority over policy development and decision making. We first develop a clearer conceptualization of empowered deliberation within the general trend toward participatory governance. We also review critical and supportive perspectives on empowered deliberation, making explicit tradeoffs inherent in the decision to develop an empowered deliberative process. Next, we identify four key dimensions of variation in the design of empowered deliberative institutions, in particular embeddedness in the social/ political context and the scope of authority of the deliberative decision. To illustrate these dimensions, we discuss key cases from around the world, noting which forms of empowered deliberation have seen less common innovation and documentation. Finally, we briefly consider how specific processes may become empowered or transform over time, as they transition from experimental or one-off pilot projects to recurring and institutionalized aspects of democratic governance.

In the past three decades, academics and practitioners alike have witnessed a shift toward a more deliberative conception of democratic politics (Dryzek, 2010;Nabatchi et al., 2012).Reason-giving and respectful exchange between people with different perspectives and understandings of the world now serve as common standards for public and private decision making.Despite concerns that deliberative norms may privilege some voices over others (Tracy, 2011), generate demobilizing ambivalence within the public (Mutz, 2006), or stifle more aggressive activism (Levine & Nierras, 2007), this deliberative turn has influenced those who seek to reform existing political systems.Political theorists have emphasized the importance of deliberation to building good citizens, an expressive public, and more normatively satisfying democracy (Gutmann & Thompson, 2004;Chambers, 2003;Barber, 1984).Meanwhile, policy makers, activists, and governance experts alike have rediscovered deliberation as a prospective tool for incorporating public demand for access to the policymaking process (Smith, 2005;Fagotto & Fung, 2006).
As the case for sound public decision making as a product of democratic deliberation gathers support, participatory deliberative processes increasingly are being empowered as part of democratic governance.Much of the research and theorizing on democratic deliberation has, however, focused on deliberation as an end in itself, rather as a part of the policy process with a definable impact.
Empowering public deliberation with formal authority in policy making raises additional questions.Some scholars worry that increasing the stakes of deliberation could undermine the quality of the process (Levine, Fung, & Gastil, 2005), or it could short-circuit essential political conflicts (Lee, McQuarrie, & Walker, 2015).Regardless, practical questions concern how best to institutionalize deliberative forums and processes within existing representative governments, as has been the focus of considerable research on "interactive policy making" (e.g., Edelenbos, Klok, & Tatenhove, 2009).
As governments delegate increasing authority and deliberative responsibility from elite bodies to lay publics, it is important to understand what this means both for deliberation and democracy.Formally committing to abide by the citizens' decisions represents a valuable move toward the goal of building a deliberative democratic system (Parkinson & Mansbridge, 2012).Taking a pragmatic approach to democratic renewal (Fung, 2012) leads us to recognize that building the institutions of a deliberative democratic system will not lead to a one-size-fitsall solution.A pragmatic approach seeks to understand the diversity of deliberative practice in the real world, exploring different potential solutions to multiple challenges of democratic life.
In this paper, we lay out the emerging characteristics of formally empowered deliberation as a distinctive subset of deliberative processes.We first develop a clearer conceptualization of empowered deliberation as part of a general movement toward participatory governance.We also review critical and supportive perspectives on empowered deliberation, making explicit tradeoffs inherent in the decision to develop an empowered deliberative process.Next, we identify four key dimensions of variation in the design of empowered deliberative institutions, in particular embeddedness in the social/political context and the scope of authority of the deliberative decision.To illustrate these dimensions, we discuss key cases from around the world, noting which forms of empowered deliberation have seen less common innovation and documentation.Finally, we briefly consider how specific processes may become empowered or transform over time, as they transition from experimental or one-off pilot projects to recurring and institutionalized aspects of democratic governance.

Theorizing Empowered Deliberation
Some conceptions of deliberation have isolated it from related concepts of power and participation.Much of the early theoretical work on deliberation was premised on a distinction between deliberation in civil society and negotiation within state decision-making, where decision making is imbued with the political power of the state (Habermas, 1989).Other theorists have drawn unnecessarily sharp lines between participatory and deliberative politics (Mutz, 2006;Pateman, 2012), as though they stood as mutually exclusive practices.
For the purposes of this analysis, we begin with a broad definition of deliberation.One of the approaches Black et al. (2010) describe for operationalizing deliberation focuses on the structural features of a decision making body or discussion format.To the extent that those structures are designed to foster deliberation, whether through formal parliamentary rules or through more openended discussion, the process meets a minimal definition for having a deliberative capacity.On top of that, we add the requirement that there must be a modicum of behavioral evidence of actual democratic deliberation, as recorded through direct observation, case studies, or other forms of research.Such discourse would be characterized by thoughtful analysis of the problem at hand within an egalitarian and respectful climate (Burkhalter, Gastil, & Kelshaw, 2002).
Using this broad definition, democratic deliberation can be seen in a variety of spheres, in ways that may overlap with other communicative practices such as dialogue or negotiation and bargaining (Nabatchi et al., 2012).Deliberation may be seen in legislative bodies deciding on policy, non-profit organizations deciding on priorities for the next year, or even among family members working out positions on private (or public) issues within the family.Much of the current innovation to revitalize the deliberative and democratic aspects of society focuses on increasing opportunities for deliberation among members of the general public, rather than within defined interest groups or political elites.This public deliberation can be more or less closely linked to the policy process and political outcomes.1At the more consequential end of the spectrum are instances of public deliberation that begin to fall into the category of participatory governance, "the devolution of decision making authority to state-sanctioned policy-making venues" that may operate in conjunction with traditional representative institutions and extend public voice into real political decision making beyond elections (Wampler, 2012: 669).
In practice, deliberative exchange features prominently in much of the work outlining the structure and anticipated benefits of many participatory governance reforms.Participatory governance gives the general public real control over government or policy decisions, and evaluations of many of the prominent models of innovations in participatory decision making put deliberative exchange between citizens at the center of their design (Newton & Geissel, 2012;Renn, Webler, & Wiedemann, 1995).The capacity for deliberative processes to generate well-reasoned and justified decisions makes deliberation an understandable choice for new venues of collective decision making.
The distinct set of cases that we call "empowered deliberation" consist of those public deliberation processes that have binding and pre-determined authority or influence over policy outcomes (either the set of possible outcomes or final decisions).Such deliberation includes processes that may elsewhere be described as part of a broader class of "direct authority" and "co-governance" (Smith, 2005;Fung, 2006).In this essay, we focus on those cases of public deliberation in which members of the lay public examine issues of public concern, rather than those forms of deliberation that occur within bodies made up of elite stakeholders.To be "empowered," these processes have to be known to be "consequential" for the legal or policy outcomes that follow them (Levine et al., 2005), with these consequences articulated at the outset of the process.Many deliberations not formally empowered may still carry many real consequences.As numerous scholars have described (Goodin & Dryzek, 2006;Gutmann & Thompson, 1996;Jacobs, Cook, & Delli Carpini, 2009), we can see a wide range of possibilities of how deliberation may have real consequences on both people and politics without being directly and formally empowered.Nevertheless, as discussed below, many others have suggested that formal, credible expectations of influence have unique implications (both positive and negative) that may influence the basic patterns of participation and communication and require additional attention to elements of process design.This category of empowered deliberation includes a wide diversity of deliberative arrangements.How deliberation is structured and authority operates vary considerably, even within this subset of deliberative occasions initially convened with expectations of binding authority.Recent well-known examples of deliberation in practice can help to illustrate this diversity.In the British Columbia Citizens' Assembly on Electoral Reform (BCCA) a selection of citizens were brought together to draft a public referendum on provincial electoral reform (Warren & Pearse, 2008).Participatory budgeting as developed in Porto Alegre, Brazil has brought together activists and everyday people from communities across a city in order to directly allocate the budget priorities for a significant portion of municipal funds through iterated community meetings and assemblies (Heller, 2001).Chicago's community policing took a different tack entirely, bringing together government authority and the public in the form of community based 'beat meetings' where government officials and members of the public come to joint agreements on police strategy and priorities (Fung & Wright, 2003).These three examples have become iconic cases in the developing world of empowered deliberation.
Though these three examples all share basic features of providing binding policy impact, they also demonstrate the diversity of participants and mechanisms available to deliberative innovators.A review of the arguments of both the supportive and critical perspectives on empowered deliberation will help to clarify the value choices and practical tradeoffs that are a crucial part of building a framework to inform future institutional design.

Critical Perspectives
Concerns with formally empowering deliberation generally focus either on system-wide issues of democratic legitimacy or on possible negative effects of empowerment on the quality of deliberation itself.Deliberation can produce decisions considered legitimate within the body that deliberated (Gutmann & Thompson, 1996), but these deliberative decisions are not necessarily legitimate as binding decisions over a wider democratic polity.As our political systems have evolved, democratic legitimacy has most often been understood as conveyed by formal electoral representation with accountability to a defined public.In this view, binding policy decisions arrived at by an unelected deliberative body, regardless of the quality of the deliberations or the descriptive representativeness of the group, should not be seen as legitimate (Parkinson, 2006;Fraser, 1992; see also discussion in Mansbridge, 2007).At worst, introducing a new decisionmaking public body, however deliberative it may be, may present risks of further undermining trust in existing institutions of democracy; while functioning democracies require critical citizens, they also require a degree of trust in the full institutional structure (Mackenzie & Warren, 2012).These concerns with legitimacy within a wider political system focus primarily on deliberative events empowered to produce final policy decisions.As we discuss further below, these concerns may be mitigated by including deliberation earlier in the policy process, at the point of agenda setting or drafting a set of policy options, rather than for final decision making.
The second set of major concerns centers on the quality of the deliberation itself.Giving formal power to deliberative processes could increase the stakes of participants in the outcome, strengthen "pre-deliberative commitments," or diminish the creativity and openness of participants to the process (Mansbridge, 2007: 262;Warren, 2007;Fraser, 1992;Fung, 2007;Parkinson, 2006).The expectation that a deliberative decision will have real consequences may bring people to the table who are primed to maximize their own benefit with fixed ideas about possible outcomes, making high quality, open-minded, reasonable exchange even more elusive than usual.In addition to these challenges specific to empowered deliberation, broader concerns regarding the democratic quality of deliberation more generally apply.These concerns include the risks of enclave decision making (Mendelberg & Oleske, 2000) and the persistence of exclusion and silences built on inequalities of power and modes of expression within even carefully managed interactions (Sanders, 1997;Young, 1996Young, , 2000;;Mouffe, 1996;Fraser, 1992).

Positive Perspectives
In response to the above criticisms, a practice-oriented perspective offers a number of compelling arguments on behalf of empowering deliberation.In a world where practical implementation consists of imperfect aspiration toward contested ideals, empowering deliberation can improve both the depth of public participation and deliberation's cumulative effect on participants.Empowerment can also result in more effective and wider reaching political impact of the outputs from deliberation.
By increasing the stakes participants have in the outcome of deliberation, empowerment better motivates people to take up deliberative opportunities and encourages participants to take the deliberation seriously (Smith, 2009;Walsh, 2007;Fung & Wright, 2003;Fung, 2007;Goodin & Dryzek, 1980).Additionally, empowered deliberation may result in significant and durable individual-level effects (Gastil et al., 2010;Jacobs et al., 2009;Button & Mattson, 1999), increasing the sense of political efficacy of both participants and observers and encouraging them to value reasoned exchange of views and participation in the future (Parkinson, 2007;Lowndes, Pratchett, & Stoker, 2001b;Pateman, 1971).People are more likely to care about and remember their experience of deliberative events when they know that it matters.
Implementing deliberation that has tangible, relevant public benefits may help to mitigate problems of exclusion of marginalized or less powerful groups.As Fung (2007) and Weatherford and McDonnell (2007) explain, the evident impact of empowered deliberation can draw out members of communities whose voices are less commonly heard in less consequential public forums.Elements of process design and framing may also help to build a more inclusive deliberation.Muhlberger (2007) reports evidence that introducing reminders of citizenship into discussion generates more 'pro-social' reasoning and choices.Formal empowerment of deliberation, framed as an act of responsible citizenship may help to minimize the risk of narrow enclave deliberation identified by Mendelberg and Oleske above.
Empowered deliberation can also affect the quality and relevance of the outputs of the event.Fung (2004Fung ( , 2007;;Fung & Wright, 2003) and Parkinson (2006) have both described how empowered deliberation produces more detailed, relevant, and problem-oriented outputs.Participants' recognition of the real-world impact of their deliberation can help to increase the practical effectiveness of deliberative outputs.
The possible impact of empowering deliberation is not limited to the isolated moment of decision making or specific policy it generates.Formal inclusion of deliberation in the policy process can also help ensure that deliberation has a formative role in the democratic system.Institutionalizing the formal decision making relevance of the outputs of citizens' deliberations is an important step in encouraging the state and other authorities to take deliberative instances seriously, rather than to treat deliberative processes as an exercise in co-option or deflection of public concern (Abers, 2003;Smith, 2005Smith, , 2009;;John, Smith, & Stoker, 2009;Fung & Wright, 2003;Lowndes, Pratchett, & Stoker, 2001a).Institutionalizing the political impact of deliberation may also be important to strengthening its relevance to broader political structures, building important links to a macro-level deliberative system (Parkinson & Mansbridge, 2012;Smith, 2009;Parkinson, 2006).

Normative and Empirical Theoretical Purposes
Finally, we wish to clarify the theoretical orientation of this essay.Our primary aim is to improve our conceptualization and understanding of empowered deliberation.We develop a taxonomy of such processes with an eye toward foregrounding gaps in both the practice and the study of deliberation, and with an eye toward furnishing a better model for what engenders and sustains empowered deliberation in existing governments.
This principally empirical purpose may run afoul of some normative critiques of deliberation, which warn against a deliberative turn that threatens to undermine more radical purposes (Lee et al., 2015;Levine & Nierras, 2007).The general criticism that empowering non-elected bodies compromises the legitimacy of democratic government can be countered by Mansbridge's reminder (2007: 263): No actual democracy, no matter how small and egalitarian, ever reaches fully equal power in aggregation.Thus no law, whether made by aggregation, by deliberation to consensus, or by a mixture of the two, is ever fully legitimate.Legitimacy is a spectrum, not a dichotomy.
Democracy is an ideal, never fully realized in practice (Dahl, 1989).If empowered deliberation can strengthen the voices of marginal groups or allow for greater creativity in public decision making, it may be worth the compromise of more complex patterns of accountability or the open acknowledgment that democratic legitimacy is never absolute.In any case, we hold that the normative desirability of empowered deliberation remains an empirical question, based on observation of its actual practice.

Ordering Empowered Deliberation
Aside from such theoretical questions, incorporating public deliberation as a decisive part of a democratic policy process raises questions for its design and institutionalization.The design of any new process must take into account the tensions between principles of legitimacy, inclusivity, deliberative quality, and effective policy impact.Perfectly legitimate empowered deliberation may not be attainable, but it provides us with ideal normative principles by which we can evaluate institutional innovations or build a framework to help imagine new approaches.
Reconciliation of the multiple constraints and priorities that shape the implementation of empowered deliberation results in multiple possible institutional forms.The literature describing participatory governance more generally provides a starting point for describing this variation, but does not provide adequate detail to describe empowered deliberation specifically.For example, Fung (2006Fung ( , 2012) ) developed a useful typology for participatory governance that describes processes by the mode of selection (who the participants are), the mode of communication (from receiving information to active deliberation), and the type of authority given to the process.In this typology, cases of empowered deliberation would fall at the deliberative and authoritative ends of the latter two dimensions, with only partial scope for variation in the first dimension (participant selection).While Fung's general framework effectively positions empowered deliberation within the larger field of participatory governance innovations, such broad frameworks do not adequately capture the important scope for variation within empowered deliberation itself.Even a basic examination of existing examples of empowered deliberation illustrates additional dimensions of variation that can be expected to have important implications for the normative evaluation of any new institutions.
Empowering public deliberation with formal authority introduces design choices that are practically and normatively important.The practical question of who has authority and how introduces at least two new dimensions of possible variation: 1) the extent to which the deliberative process is embedded in or insulated from the existing social/political power, and 2) whether any decision produced by public deliberation has final or provisional authority.In addition, as with other public processes, empowered deliberation can vary in terms of the geographic scale and repetition of events over time.Variation along each of these dimensions may have consequences for the normative evaluation of such processes in terms of values such as accessibility, equitable inclusion, or democratic legitimacy.Table 1 outlines these dimensions, highlighting key, well-documented, real-world examples that illustrate the several different approaches to empowered deliberation in practice.
It should be noted that these four dimensions are not necessarily the only possible dimensions along which empowered deliberation could vary.They represent important patterns in the design and distribution of examples of empowered deliberation that have been recorded in academic literature and practice-oriented resources, and reflect clusters of innovation and diffusion of popular forms such as citizens' juries or participatory budgeting.2It is possible that innovation could introduce additional variation not yet captured in this descriptive framework.

Embedded versus Autonomous Authority
A key dimension of variation is the extent to which the deliberative event(s) are either insulated from or embedded in existing structures of social and political organization.A deliberative event such as a citizens' jury or citizens' assembly particular organizations: the former at The Jefferson Center (www.jefferson-center.org) and the latter at The Center for Deliberative Democracy (cdd.stanford.edu/polls).
3 Cases included in this table are drawn from a range of sources, including academic literature on deliberation and/or participatory governance, online resources (particularly participedia.net),and the authors' knowledge of the field.References and further information for each case are available in the appendix.Though we did not conduct a census of all cases of empowered public deliberation, we did make every effort to populate empty cells in this table, including extensive review of secondary literature and examination of all cases posted to Participedia that were classified as having a purpose of "co-governance" or "making public decisions."may be designed to maximize independence from existing organized interests or authority, or it may be designed to incorporate or even amplify the structure and relationships of the local civic environment.This dimension represents the tension between efforts to nurture protected spaces of representative public (i.e., nonelite) deliberation, by minimizing influence from existing powerful actors, and efforts to include all relevant stakeholders and authorities whose cooperation and expertise may be critical to implementation of decisions.Deliberation along the former model is referred to here as 'autonomous,' whereas deliberation carefully integrated and drawing on existing networks of community interests, actors, and authority is referred to as 'embedded.' Fagotto and Fung advocate for a particular kind of embedded deliberation, where deliberative practices are "incorporated into…the communicative and decision making routines of organizations, institutions, and the communities of which they are part " (2006: 6).In this sense, embeddedness occurs when deliberative events occur regularly and have an impact in communities to address a range of public problems.The strength of embedded deliberation here comes from building it into existing sources of authority, developing a sense of ownership across the spectrum of powerful and relevant actors.In various discussions of interactive governance in Europe, Edelenbos and collaborators have identified similar patterns of embeddedness (specifically the active involvement of relevant actors in governance innovations) as necessary to the effective incorporation of new institutions into existing processes (Edelenbos et al., 2009).They distinguish between professional, political, executive, and policy embedding (Edelenbos, van Schie, & Gerrits, 2010), emphasizing the importance of incorporating a wide range of powerful actors, including politicians, relevant bureaucrats, and professional experts.In each approach, dense and comprehensive links to existing networks of power and implementation are necessary for new forms of governance to be effective.In contrast to a theoretical ideal of disinterested, autonomous deliberation, embedding starts from the assumption that empowered deliberation requires broad engagement by existing actors and authorities in any new processes.This engagement is particularly dependent on building a process in which the range of relevant actors, from state officials to private interests to community groups, feel they have a stake in the success of the project (Hendricks, 2006;Abers, 2003).
Embedded deliberation can be found in diverse contexts, often but not exclusively operating at the municipal level.The experiences of Chicago Community Policing, or the deliberative governance practiced in Hampton, Virginia (Potapchuk, Carlson, & Kennedy, 2005), and ongoing experiences of participatory budgeting in the US offer classic examples of embedded deliberation.Participatory budgeting, as initially developed in Porto Alegre, Brazil and in many of its worldwide reinterpretations (Sintomer, Herzberg, & Röcke, 2008), emphasizes the importance of social and organizational embedding of community deliberation within and alongside the official and bureaucratic interfaces of the state.Participatory budgeting often gives greater weight to the mobilization and incorporation of community activists and organized interests to facilitate effective public engagement and accurate representation in a multi-stage process of decision making (Baiocchi, 2005).In all these cases, inclusive and effective public deliberation is produced through careful integration of deliberative processes into the existing institutions and interests of the community.
Empowered deliberation may often be embedded, especially when instituted at the local level where it is most feasible to bring all stakeholders together.It may, alternatively, be carefully insulated from the existing hierarchies of political, bureaucratic, and social interests, in a distinctive representative form often described as a 'minipublic' (Davies, Blackstock, & Rauschmayer, 2005;Fung, 2003;Gastil & Richards, 2013).In an effort to preserve the unique virtues of deliberation as a mode of communication, even as it is formally empowered, a number of scholars and practitioners have emphasized the importance of insulation from articulated interests rather than contextual embedding.Goodin and Dryzek (2006), for example, consider the possibilities for wider impact from deliberation in representative minipublics which need not be embedded in existing institutions.Though most of Goodin and Dryzek's examples of deliberation cannot be described as formally empowered, they also highlight the canonical case of the Citizens' Assembly on Electoral Reform in British Columbia (BCCA).
Tasked with creating a proposal for a new voting system for the province, this mostly-randomly selected group of 160 citizens from communities across the province were brought together to deliberate on important and often technical issues of electoral reform, with a collective decision to be approved or rejected in a later public referendum.
Not only were members of the Assembly invited and selected in order to produce an Assembly that was descriptively representative of the province as a whole, the Assembly was explicitly isolated from existing political and civic actors.While information was included from experts of electoral systems, political parties and established political elites were not even allowed a voice in the information gathering phase of the deliberations (Ratner, 2008: 163).Fewer examples of autonomous empowered deliberation are currently available, but the British Columbia experiment has been picked up and replicated in Ontario and the Netherlands (Smith, 2009: 108) and related assemblies and forums have been used at the national level as part of Iceland's 2010 constitutional reform (Landemore, 2014) and in the Citizens' Panel in the city of Canada Bay in Australia (Thompson, 2012).
The decision whether to embed deliberation into existing systems or instead to build new autonomous institutions from the ground up carries several important value choices.Embedded deliberation may require time and social resources to generate buy-in from the full range of relevant communities, but if established it promises effective and durable deliberative policy outcomes, as governance and deliberation unfold together.Alternatively, maintaining autonomy for formally empowered deliberative events promises to preserve the high quality of deliberation even in a context of a high-stakes outcome.Designing new institutions with formal authority requires careful consideration of protections of transparency and democratic legitimacy.The two approaches also prioritize inclusivity differently.Embedded deliberation takes a self-selecting stakeholder approach, which has often succeeded in drawing out a diverse range of participants through a focus on community networks.Autonomous deliberation tends to emphasize formal descriptive representativeness, typically using a variation on random selection with control over the demographic profile of the deliberative body.Alternatives to random selection may not be able to preserve the autonomy of the group, as they may be dependent on existing community networks to recruit participants.
It is important to note that complex governance processes may successfully incorporate both approaches, using a combination of embedded and autonomous deliberative spaces in the development of a final policy decision.The Geraldton 2029 planning process exemplifies such a combination.While on the whole, this multi-year process is best characterized as embedded in the community, it also includes an autonomous (randomly selected) deliberative poll and survey.
Sustained cases of empowered deliberation may develop into sequences of multiple events with distinct design features.Attention to the implications of these design choices may help to identify the most effective combinations for new governance challenges in the future.The open design process for the UK-based NHS Citizen project to improve the public involvement in governance of the country's National Health Service (still in development at time of submission) illustrates the evolving considerations of trade-offs between embedded and autonomous processes.The emerging final design is likely to incorporate autonomous elements within a broadly embedded citizen structure in an effort to maximize benefits from both approaches.

Provisional versus Final Authority
The second major dimension of variation specific to the design of empowered deliberation is the extent of authority held by the deliberating group.The outcome of a deliberative process may be directly implemented into policy, or it may be subject to a further ratification or approval.
Returning to one of our canonical cases of empowered deliberation, Chicago Community Policing illustrates the final authority end of the spectrum.Local residents and police officers in Chicago came together in neighborhood 'beat meetings' where their discussions directly determined priorities and policing strategies for each neighborhood.The decisions made in the decentralized deliberations in Chicago thus generated final decisions regarding local policing priorities and strategies.
Another, more recent, example of final authority can be seen in the California Citizens Redistricting Commission, which was established by initiative and first put into practice in 2010. 4Though drawing district lines might seem a trivial administrative exercise, it has profound implications for the election results that follow.This fourteen-member commission exercises its power in a state of exceptional size and power (more than 38 million people who generate more than two trillion dollars in gross state product), and its decisions have implications for federal Congressional elections.Its complicated selection process includes voluntary applications, a review of citizen applicants by the state auditor, and a random draw from the sixty finalists.
On the other end of the spectrum, the BCCA, discussed above, provides a clear example of a design with provisional authority.The final proposal for electoral reform from the BCCA proceedings, despite emerging from hours of deliberation and a formal vote in the assembly, required subsequent approval by a supermajority of voters in a public referendum in order to become law (Ferejohn, 2008).Voters in British Columbia ultimately rejected the Assembly's provisional decision when the referendum failed to pass the supermajority threshold.Other examples of provisionally empowered deliberation, such as the Oregon Citizens' Initiative Review or Ontario's assembly for electoral reform, have also been subject to final approval by the general public.
Not all provisional authority is dependent on a final public vote.Other instances of empowered deliberation are subject to confirmation or reconsideration by elected or appointed officials.For example, the model of participatory budgeting implemented in Vallejo, California in 2012 allows the Vallejo City Council a final vote on implementation of the projects that made it through the public deliberation and voting phases. 5Similarly, the recommendations of the 2012 Canada Bay Citizens' Panel in Australia (Thompson, 2012), tasked with establishing priorities for the city over the next several years, were also subject to approval by the city council.In the face of the sweeping task of rewriting the country's constitution, the Icelandic National Forum in 2010 yielded output that served as the starting point of the deliberations of the elected Constitutional Council, whose decision would in turn be subject to public referendum and legislative approval (Landemore, 2014).
Whether authority was provisional or final, participants deliberated with a realistic expectation that they would have a predictable and binding impact on the policy process.Deliberative decisions can thus be effectively empowered at a variety of points in the policy process as a whole, within the wider system of governance and legitimate decision making that may in fact offer multiple overlapping sites of deliberation and decision making (Smith, 2009;Warren, 2007;Parkinson, 2006).Thinking carefully about which decision points are appropriate to each potential instance of deliberation may help to produce an empowered deliberative process that can best balance the sometimes conflicting priorities in the development of new deliberative processes.
Careful sequencing in a policy process can be instituted to deal explicitly with concerns of legitimacy and appropriate authority.Thus, the BCCA can be seen to have resolved the ambiguities of legitimacy raised by a non-elected deliberative body deciding on public policy, through careful sequencing with a secondary full public deliberation over the referendum (Ferejohn, 2008;Smith, 2009;Warren & Pearse, 2008).Weatherford and McDonnell (2007) similarly propose implementation of deliberation in a multi-stage process that moves through different levels of the community.The international variations in the design of participatory budgeting processes demonstrate a wide range of interpretations of provisional versus final authority.For example, the Porto Alegre model demonstrates the development of a complex, iterated sequence of deliberative events that combines open community forums with deliberation within specially elected bodies.This multi-level structure helps PB to obtain legitimate agreement across broader polities than could meet face to face (Baiocchi, 2005: 75).
Similarly, recent implementations of PB in North America combine multiple instances of public deliberation to generate preferred outcomes which are finally determined in a non-deliberative public vote.
A design based on provisional authority allows for greater creativity about institutional combinations, as single innovations need not be considered in isolation, but can rather be imagined as part of a whole governance system.Integration into complex processes can create opportunities for carefully targeted negotiation between the values at stake in design choices.On the other hand, limiting a deliberative process to only provisional authority carries with it the possibility of obscuring or diluting the outcome of the deliberation, increasing challenges to the transparency of the decision process and possibly weakening some of the possible positive impacts of empowering deliberation discussed above.As with the choice to embed or isolate deliberation, the choice of the extent of final authority carries with it real trade-offs for the different democratic values that may be sought by democratic reforms.

Deliberation across Space and Time
Two other important aspects of design include the geographic scale and repetition over time.These characteristics may not always be open to manipulation by those implementing a process, but they carry important implications for the values at stake in the wider design.
Geographic scale, in particular, has played a role in the evolution of empowered deliberation.Deliberation, empowered or otherwise, has often been conceived of as something that happens face to face, within groups small enough in both numbers and viewpoints to be able to talk to and understand one another (Burkhalter et al., 2002;Chambers, 2003: 309).Many of the best known experiments in empowered deliberation have occurred at the neighborhood or city level, in contexts in which stakeholders can be more easily identified, problems are often of immediate and tangible importance to members of the public, and fewer logistical barriers to participation exist (Fagotto & Fung, 2006;Gastil & Levine, 2005;Nabatchi et al., 2012;Weatherford & McDonnell, 2007).Archon Fung has been a proponent of the idea that not only can citizens "be the shock troops of democracy," whose "local knowledge, wisdom, commitment, authority, even rectitude can address wicked failures of legitimacy, justice, and effectiveness" (2006: 74), but that this dynamic is most effective at the very local level.Fung and early collaborators proposed a form of 'empowered participatory governance' (EPG), one of the principal design features of which is the "devolution of public decision authority to empowered local units" (Fung & Wright, 2003: 15).
Nevertheless, as deliberation comes to be seen as important to democratizing whole polities rather than insular communities, the question of scale has come increasingly into play.Criticisms of a localized approach to building a deliberative democratic society come in several forms.Extreme localism may have greater potential to damage minority rights or the inclusiveness of the process (Parvin, 2009;Young, 2000).Alternatively, highly local developments may arguably only tinker at the edges of policy without the means to deliberate on or challenge the decisions, values, or agenda of the larger political reality (Gutmann & Thompson, 2004;Parkinson, 2007;Parkinson & Mansbridge, 2012;Smith, 2009).By this argument, without the clear potential for impact upon polity-wide realities of power and distribution of resources, local community deliberation cannot be understood as effectively empowered.This ineffectiveness may be recognized by participants, which implies subsequent limitations to the projected positive effects of empowered deliberation.
As the cases highlighted in this paper demonstrate, empowered deliberation has been effectively implemented at both the local and 'translocal' levels (i.e., broader social scales that extend beyond the local level).These examples also illustrate how different choices along this dimension may play out in practice.The smallscale ideal is exemplified by Chicago Community Policing.As Fung (2004) describes, neighborhood residents know the streets and the problems and are equipped with the expert knowledge to imagine creative strategies for dealing with the joint problems of drinking and drug use in their immediate area.The proximity of citizens to the problems and the tangible benefits at stake ensures broad participation in the process, appealing especially to marginalized and traditionally disempowered populations.At the other end of the spectrum, the BCCA is a demonstration of a well-crafted attempt to institutionalize empowered deliberation across a wider public.With the necessity of developing high-level electoral policy that would impact the province as a whole, this descriptively representative group of citizens from across the entire province became empowered representatives of a broader polity, albeit with only provisional authority.
These cases also illustrate how the spatial extent of deliberation affects tradeoffs between fundamental values such as legitimacy, inclusivity, deliberative quality, and empowerment.For example, limiting the translocal BCCA to only provisional authority was an important part of establishing the legitimacy of the outcome for province as a whole, where political legitimacy is generally understood to derive from the will of the voters.On the other hand, the very local decisions made in community beat meetings through cooperation between city officials and the public operated at a small enough scale not to compete with other institutions for democratic legitimacy.Large-scale deliberation faces further challenges of inclusivity; in a polity where all relevant or interested voices could not be heard individually, random selection seeks to produce a microcosm of the polity as a whole.In the small-scale context of the community beat meetings, inclusivity was prioritized through maximizing the local relevance of meetings and keeping them open to the communities.As Fung (2004Fung ( , 2007) ) reports, these meetings had unusually high success at bringing out usually marginalized members of the community.
Geographic scale is constrained by the scope of the problem.Deliberation on polity-wide reform in the case of the BCCA required construction of polity-wide deliberation; local-level deliberations may not have legitimacy to determine policy for the whole province.Similarly, intractable local problems such as policing in the community may call for localized intensive deliberation.However, this local process is limited in scale; while empowered locally, it is not empowered to change broader socio-economic realities which may be at the root of the problems.Issue relevance and tractability of local problems are traded for large-scale impact and wider empowerment.6 Finally, examples of empowered deliberation can vary between deliberative processes that are one-off events and those that have been repeated.Whereas event repetition is not precisely a design element, it is worth considering explicitly.The transition from isolated or pilot project to a recurring event carries with it particular implications for accessibility, accountability, and transparency.As illustrated in Table 1 above, it is distinctly challenging to find any modern examples of deliberative processes with final authority that were implemented as isolated events.
Of our examples of recurring processes, most began as a pilot or single event; a choice was made by the community or government to repeat it.Whether a process is recurring may have an impact on the values at stake in the design process, displaying an evolutionary dynamic that may make an important context for evaluating institutional choices.Repeated experiences of deliberation can reinforce the norms and perceptions of legitimacy of the deliberative process.
Repetition can increase the quality of deliberation by repeated learning and encouraging previously disinterested members of the community to buy in to the process as they see deliberative outcomes have an effect.This iterative process is seen to be so important that it has in fact been identified as a critical component of both embedded deliberation and accountable autonomy (Fagotto & Fung, 2006: 25-27; also suggested in Parkinson, 2006).

Making Sense of Variation
While the cases presented in Table 1 and preceding sections do not come from a systematic sample of the universe of public deliberation, and cannot be understood as representing the accurate distribution of cases of empowered deliberation in the world, it is worth sharing some of the observations we made as we worked to identify documented cases to exemplify all possible cells in the table .Recognizing where we were unable to identify examples suggests some interesting patterns across existing experiments in empowered deliberation.
Embedded deliberation seems to be more commonly recurring, reflecting the impact of generating buy-in from stakeholders and relevant authorities.As Fagotto and Fung (2006) observe, there is a kind of virtuous cycle in an embedded process whereby repeated implementation of a process and buy-in from stakeholders feedback on each other.The greater number of recurring embedded processes may also reflect the dominance of the EPG framework during the development of experiments in local empowered deliberation or simply reflect the practical constraints of working within dense, problem-oriented networks at the local level.In addition, while we found fewer examples of such processes in sources such as Participedia, National League of Cities online resources, and scholarly literature, it is reasonable to expect that there may be more single embedded events in practice.Such processes are likely implemented as a solution to an isolated governance challenge, without being documented in the wider community around deliberation and democratic reforms.7 Overall, provisional empowerment appeared to generate a wider range of institutional forms, perhaps reflecting the greater flexibility provisional empowerment offers for integrating with existing institutions and resolving legitimacy.We did, however, have difficulty finding translocal single events with provisional or final authority.Cases with final authority are generally embedded at the local level rather than translocal, perhaps reflecting the greater challenge of generating legitimate decisions while preserving deliberative quality over a larger area.The California Redistricting Commission discussed above is an innovative experiment that moves beyond the local level, while remaining embedded in existing networks of authority or established interest, with its system of application and nomination prior to random selection.This carefully designed selection process likely helps to preserve the legitimacy of the body's decisions.
The striking exception to the general absence of autonomous processes with final authority can be seen in criminal and civil juries.Juries exercise considerable authority whether called for municipal, state, or federal cases, and juries have become integral to the larger legal systems of which they are a part (Dzur, 2012;Gastil et al., 2010).In filling this otherwise empty quadrant, juries also notably operate at both local and translocal levels.It is even conceivable that such juries could become transnational, if established in bodies such as the World Court (Gastil, Lingle, & Deess, 2010).As one of the longest-established forms of authoritative deliberation we have, juries present a suggestive story about evolutionary dynamics by which familiarity and history lend legitimacy to more unusual institutional forms.
The distribution of cases between single and recurring is itself informative, particularly in the absence of any one-off cases with final authority.Repetition and familiarity of a process matters, and may support the evolution of increasingly empowered deliberation.Many of these cases illustrated here may have developed out of public processes that were initially advisory at best, but which served as effective 'proof of concept.' Thinking about evolution in this sense returns us to some of the conceptual fuzziness in the concept of empowerment.While formal empowerment is easy to identify, in cases of fully embedded deliberation as Fagotto and Fung (2006) describe it, where deliberative processes are institutionalized, engaged in and taken seriously by multiple actors in a political system, it becomes appropriate to consider this deliberation as which was largely adopted as the relevant city ordinance.It is debatable whether this was truly 'public' deliberation as opposed to deliberation by elite stakeholders, but it is probable that other undocumented one-off commissions or planning processes included a form of public deliberation as a key part of the decision process, especially at the local level.
effectively, if informally, empowered.As we discuss further below, some processes may effectively transition to de facto empowered deliberation.
Clearer understanding of the varied dynamics of empowered deliberation emphasizes the consequences of that transition, when an advisory or consultative process becomes informally empowered.How is this processes effectively empowered?At what point in the policy process does its impact take effect?Are there ways this impact could or should be formalized or made explicit?Who takes part in the deliberation?Are participants adequately representative of the scope of communities impacted by this deliberation?Considering the wide variation in process design we have illustrated here, these questions are important considerations when public deliberation is empowered, whether at the initial inception of a new process or as it evolves to be credibly consequential.

Conclusions and Implications for Future Practice
Empowered public deliberation combines the two tendencies towards increasing public authority in governance and expanding the available space for deliberative discussion and decision making.Genuine attempts to wed public authority to public deliberation constitute prominent examples of contemporary democratic innovation.As practitioners and reform-minded officials increasingly gravitate to empowered public deliberation as a step toward more effective and legitimate democracies, however, it has become clear that there is scope for better theoretical and practical understanding of variations in such practices.
Under ideal circumstances, empowered deliberation can address problems of difficult political decision making, disengaged communities, or challenges of democratic legitimacy.In political reality, by contrast, one must consider the fit of a given deliberative design to the particular time and place where it is implemented.An embedded deliberation requires outreach and extensive preliminary work to identify and bring in the relevant community, while an autonomous deliberation will require a clear understanding of what it means to be 'representative' of the community as well as a transparent and wellcommunicated process to establish legitimacy in the public eye. 8Similarly, choices of final versus provisional decision making and even geographic scale require distinct investments in transparency and stakeholder buy-in.
This diversity of models also points to a wide possibility for further innovation and creative solutions to these challenges.As illustrated in this paper, it is easier to find examples of certain models of empowered public deliberation than others, suggesting the emergence of clustering around certain models.This may be indicative of a set of evolved solutions to joint challenges of legitimacy and feasibility, but at the same time a full understanding of the possible design space may also point to areas where future innovation is possible.Participatory budgeting (in the narrow sense in which the term is used today) was invented in Brazil twenty-five years ago, and has diffused and been reinterpreted remarkably around the world since then (Baiocchi & Ganuza, 2012).Though the diffusion of this successfully compelling solution has established a density of cases around that highly sequenced and directly empowered model, hopefully the explication of other underdeveloped models of empowered deliberative participation can remind scholars and practitioners alike to continue to think creatively around solutions to the ongoing challenges to sustainable democratic practice.This creative exploration of different models of empowered deliberation may also help to expand our understanding of deliberation itself, illustrating different kinds of reason-giving and deliberative exchange that may emerge across such diverse settings.
An important line of future research can build on frameworks such as this to more fully categorize the diversity of practices in existence, or in recent history.After all, the cases presented herein were meant to illustrate different combinations of key features, without any intent to estimate the relative frequencies of these combinations in the world.Though we made note of the difficulty of finding any cases of a few particular types, some may exist in those empty cells (in Table 1).More importantly, it would be useful to get a better estimate of the prevalence of cases in those cells that we were able to populate.The best effort toward creating a database of this kind may be the Participedia database (Fung & Warren, 2011), from which we drew some case information for this essay.Though it still lacks a systematic method for identifying cases, and for characterizing them with a high degree of validity and reliability, the Participedia project represents the best current effort of this kind.
From a practical standpoint, the question remains whether a type for which we could not find an example could be brought into existence.We are inclined to trust in the wisdom of previous practice, such that the absence of cases in particular cells of Table 1 reflects the lived experience of democratic reformers, who saw good reasons to eschew such deliberative projects or who tried and failed to bring them into existence.That said, many democratic innovations have come into existence despite skepticism that they might ever arrive.Nearly every one of the processes we showcase, from participatory budgeting to the British Columbia Citizens' Assembly, was an anomaly at the time of its inception.
Our final reflections concern the dynamic character of deliberative processes in relation to the categories in our framework.Here, the modern jury provides an example of how a process can evolve-over centuries and across oceans-to take on a distinctive shape and function.Though taken for granted today as an autonomous body, this status was achieved through a legal appeal at a discrete point in its development in England, before which its function was to confirm the presumed verdict of the presiding judge (Dwyer, 2002).
A more modern instance comes from processes such as the Oregon Citizens' Initiative Review, which began as a pilot project fully autonomous from government in 2008 (Knobloch et al., 2013).The next year, it had become a government-sponsored one-year-trial, and two years later it was made a permanent institution with a commission within government.One might be tempted to code the Review as "embedded" if the state began providing the funds for its operation and if state actors played a more direct role in designing and overseeing its deliberations.In the meantime, however, it operates as a largely autonomous body, even as an ever-larger proportion of the public comes to know of the Review and chooses to use its findings when making voting choices (Gastil et al., 2014).
This dynamic character may even extend to the formal empowerment of deliberation.Some deliberative processes may have an impact on policy after the fact without the prior authority necessary to be understood as empowered.For example, the case of New Mexico's Strong Starts for Children demonstrated a sophisticated process of dialogue and deliberation that appears to have had a clear impact on policy, without having been formally empowered.Similarly, the broad public deliberation of Brazil's National Policy Conferences has been described as influencing national policy, even without a binding promise or formal authority granted to the conferences' recommendations (Pogrebinschi & Samuels, 2014).By our definition, these cases would not be empowered deliberation, as participants are not certain of whether or how their participation would have a policy impact.However, public deliberation that effectively informs policy may eventually be incorporated as an official part of the policy process.If such processes for developing policy recommendations were to be formally institutionalized, considering the placement of such deliberation on our observed dimensions of variation could help to clarify the possible democratic implications of its formalization.
These cases and others may cause even more fundamental changes in the political landscape.Advances in deliberative design may further refine the expectations of the public and public officials for what is possible.Here, the very idea of deliberative democracy serves as an example.Just as participatory conceptions of democracy have displaced the more elite models that preceded them, a deliberative conception of democracy may make possible forms of empowered deliberation that now seem improbable.These may include more creative models of trans-local embedded deliberation, even more diverse forms of autonomous deliberation with final authority, or models yet to be imagined.

Appendix: Brief Descriptions of Cases
Below are brief descriptions, with helpful references, for the processes described in Table 1 and at other points in this article.They are presented in alphabetical order.Additional information and further resources on all cases, except where noted, can also be found on Participedia.All webpages last accessed on 2/23/2015.
#changesCGGcommunity is an approach to participatory budgeting and participatory planning in Geraldton, Western Australia, in which two panels of up to 40 randomly selected members of the public deliberated separately on prioritization of spending on capital works and service delivery for the city.(This is a recent process not yet on Participedia; additional information can be found at the city's website, www.cgg.wa.gov.au/major-projects/changes-cgg-community/.) BC and Ontario Citizen's Assemblies were representative random samples of the population in each province tasked with the challenge of learning about different electoral systems and deciding on a best system to present to the public in a referendum (Warren & Pearse, 2008;Fournier et al., 2011).
California Citizens Redistricting Commission is a quasi-randomly selected body established in 2010 to oversee the drawing of district lines in that state (Korgan & McGhee, 2012).
Canada Bay Citizen's Panel was a random sample of 36 citizens in the city of Canada Bay in New South Wales, Australia, tasked with establishing service and budget priorities for the city, subject to approval by the city council (Thompson, 2014; see also the city's description at www.canadabay.nsw.gov.au/citizens-panel-pg.html).
Chicago Community Policing through the CAPS program was designed to bring local residents and local police together in neighborhood "beat meetings" identifying problems and priorities for policing in the local area (Fung, 2004).
Deliberative Polls in Wenling City in Zeguo Province in China use a random sample of the public to discover the informed priorities of members of the public to help determine the budget (Fishkin et al., 2010).
Geraldton 2029 and Beyond, in Western Australia, used a range of engagement activities including a random sample from the community to identify community assets and planning priorities for an extensive revision of the city's general plan.(See the city's own documentation on the multi-year project at www.cgg.wa.gov.au/majorprojects/changes-cgg-community/about-2029-and-beyond-project.) Oregon Citizens' Initiative Review was established in Oregon in 2009 to convene a small body of randomly-selected citizens, who analyze a ballot measure and write a onepage statement that appears in the official Voters' Pamphlet (Knobloch et al., 2013).
Participatory Budgeting in Brazil not only provides the origin story of municipal participatory budgeting, with repeated cycles of neighborhood assemblies, elected delegates deciding on priorities, and public exploration and deliberation over major capital investment.It also provides a number of examples of PB scaling up to large numbers of people, from PB in very large cities like São Paulo (with over 10 million residents) or the whole state of Rio Grande do Sul (Wampler 2008;Goldfrank & Schneider, 2006;Heller, 2001).
Participatory Budgeting in North America builds on the practice of municipal-level participatory budgeting developed in Porto Alegre, Brazil, using an annual cycle of open facilitated neighborhood assemblies, longer-term delegate meetings for project prioritization, and a public vote designed to maximize broad opportunities for participation (Lerner & Secondo, 2012; see also www.participatorybudgeting.org).
People's Planning Campaign (Kerala) was a system for decentralized participatory decision making in the state of Kerala in India, with a multi-stage process of open public meetings and delegated smaller committees (Isaac & Franke, 2002;Heller, 2001).
Strong Starts for Children (New Mexico) invited the public to dialogue circles to identify solutions to challenges of early childhood development and education, which then fed into action circles and a statewide policy forum resulting in a report of recommendations on the problem.This process was not formally empowered (and thus should not be considered as empowered deliberation as defined here), but has been linked to subsequent legislative action at the state level.(See an overview at Everyday Democracy, everyday-democracy.org/strong-starts-children-background,or in the Participedia entry.) Tuscan Law no.69/2007 institutionalized deliberative participation in the Italian region of Tuscany, establishing a right of the public to demand deliberation on major regional development projects as well as to request support for deliberative participation at the local level, the results of which must be implemented or responded to by the relevant government.Numerous case descriptions of local processes emerging from the law can be found on Participedia (Carson & Lewanski, 2008;Lewanski, 2008).

Table 1 : Varieties of Empowered Deliberation 3 Provisional Final
. First established in 2008 as a pilot project by Healthy Democracy in Oregon, this process became established in state law in 2009 and 2011 as a semi-autonomous commission funded by private donors but staffed by a mix of government appointees and Review alumni.During each statewide initiative election, the Review convenes two-dozen citizens to deliberate for a full week on a ballot measure, based loosely on the citizens' jury model.What distinguishes it from other processes is the fact that the Review panel's findings appear as a one-page document in the official Voter's Pamphlet, which the Secretary of State distributes to every registered voter in the state.Thus, the Review has tremendous potential influence on the wider electorate, which it induces to engage in a kind of "vicarious deliberation"