What did we learn about citizen involvement in the health policy process: lessons from Brazil

In this paper I argue that citizen involvement helped to promote a more equitable distribution of public health services in Brazil. This achievement involved a balance of contributions from social actors and health system managers in forging policy innovations and institutional arrangements that linked bottom up innovation with national policy leveraging and decentralized implementation. The paper briefly describes this cycle and its relation with the implementation of a national network of forums for citizen involvement in health policy, inquiring in more detail the conditions that favor the association between these forums and the policy making process. Our results do not corroborate the idea that deliberative arenas should be insulated from political passions; rather, they suggest that participation of mobilized social actors contributes to the effectiveness of these forums. This contribution happens both due to the knowledge that these actors bring about problems in the area and to their insertion in networks that connect forums to a wide set of social organizations and political, governmental, and health institutions, which in turn facilitate the dissemination and negotiation of the proposals and demands formulated by the forums. However, our results also suggest that the inclusion of these actors increases confrontation to the detriment of deliberation, which brings us to the discussion of the role that could be played by mediators who are well-equipped to construct deliberative processes.

What did we learn about citizen involvement in the health policy process: lessons from Brazil 1 Vera Schattan P. Coelho; CDRC/CEBRAP-NCD/CEM In this paper I argue that citizen involvement helped to promote a more equitable distribution of public health services in Brazil. This achievement involved a balance of contributions from social actors and health system managers in forging policy innovations and institutional arrangements that linked bottom up innovation with national policy leveraging and decentralized implementation. The paper briefly describes this cycle and its relation with the implementation of a national network of forums for citizen involvement in health policy, inquiring in more detail the conditions that favor the association between these forums and the policy making process. Our results do not corroborate the idea that deliberative arenas should be insulated from political passions; rather, they suggest that participation of mobilized social actors contributes to the effectiveness of these forums. This contribution happens both due to the knowledge that these actors bring about problems in the area and to their insertion in networks that connect forums to a wide set of social organizations and political, governmental, and health institutions, which in turn facilitate the dissemination and negotiation of the proposals and demands formulated by the forums. Despite these achievements the results also call attention to a slight increase in inequality in the distribution of basic services between the poorest regions.
In the beginning of the 1980s it became clear in Brazil, as well as in a number of other countries, that greater resources for public healthcare were allocated to wealthier regions and citizens. Over the next few years, the difficulties in overcoming this distributive pattern gained momentum in the public policy literature and by the end of this decade there were different pathways being described as possible avenues to overcoming distributive inequalities in the health sector. Of the several possibilities being discussed at this time, two gained the particular attention of governments, scholars and donors: citizen involvement and decentralization. The idea was that both could contribute by promoting innovation, accountability for the needs of the poorest citizens, and social control. In Brazil this agenda gained political relevance after the enactment of the "Citizen Constitution" in 1988. In addition to declaring health to be a universal right of citizens, the Constitution replaced the old public health system with the SUS, or Unified Health System. At the time of its creation, the SUS reflected the aspiration for a system that integrated the local, the state and the national levels as well as preventive and curative medicine. Recent data show that the SUS has, in its 20 years, been working for the poor, as efforts have successfully been made to achieve a more balanced distribution of resources between the worse and better off regions. Also, in addition to the reduction of inter and intra regional inequalities, there has also been a decline in health inequalities between the population as a whole and some of Brazil's most vulnerable groups (MS 2010;IPEA 2009;Siops 2008;Souza 2003). 2 Nevertheless, there was also a small increase in the inequality in the distribution of basic services within the poorest areas and groups, which I will refer to as horizontal inequalities 3 (Shankland 2010, Coelho et al 2010a.
In this paper I begin by calling attention to the debate on citizen involvement and policy change. I then present the results of a research program that, during the last ten years, has made systematic inquiries into the relationships between citizen involvement and changes in the distributive profile of public health services in the city of São Paulo. The paper concludes by discussing features of these processes that can be improved in order to strengthen the processes' accountability to the most vulnerable and least mobilized groups.
In the next section, I review different theorethical and analytical approaches to the relationship between citizen involvement and policy change. In section three, the cycle of innovation that emerged during the 1990s within the SUS is described. Section four presents the methodology developed to allow for a more systematic analysis of performance and the capacity to impact policy of the formal participatory arenas. This methodology was tested in the city of São Paulo. In section five, the results of this work are presented, pointing to: 1) A reduction in the disparity between the supply and 2 See Annex 1, which presents data on Brazil showing a reduction in inter-regional inequalities and data for the Municipality of São Paulo showing a reduction in intra-regional inequalities. 3 I use the term horizontal inequalities to refer to inequalities that grow between groups that depart from similar socio-economic features. They result from a process where sub-groups leave behind other subgroups with whom they had shared membership of a broader, formerly excluded group ("peasants," "indigenous people," "the urban poor," etc.).
consumption of public health services across the areas with the best and worst indices of income, education and health; 2) A slight increase in inequality in the distribution of basic services between the poorest areas; 3) A positive association between citizen involvement, mobilization and these distributive results. Finally, in section six, a summary of findings is presented which moreover includes suggestions concerning features of participation that may contribute to tackling horizontal inequalities.

Linking citizens to the policy decision making process: a brief overview of the debate
In parallel with a public policy debate that argues for decentralization and participation as pathways to promoting government accountability to citizens' needs and preferences (UNDP 2002), there is an academic debate focused on improving the quality of democracy. Theorists of deliberative democracy, deepen democracy, and participatory governance have been discussing how procedural, institutional, and social features can improve the quality of citizens' political involvement.
Authors associated with deliberative democracy theory believe that the core idea of democratic deliberations is that decisions are made with more and better information and these decisions come to be accepted as legitimate and justified by participants. They call attention to the process of decision-making and offer nuanced criteria for assessing the quality of these processes (Habermas 1997;Dryzek 2001;Mansbridge 2003). Those associated with deepen democracy argue that citizenship should mean far more than just the enjoyment of legal rights and the election of representatives, highlighting citizens' potential to collectively mobilize in order to be directly involved in deliberation and decision-making on political and policy issues (Gaventa and Cornwall 2001;Heller 2001;Avritzer 2002). Finally, those associated with participatory governance are particularly interested in how to coordinate these new and more participatory political arenas with the congressional and executive governmental bodies. They inquire about the institutional framework in which these bodies interact with a view to developing and implementing policies that are more accountable to citizens' needs (Fung and Wright 2003;Melo and Baiochhi 2006).
Departing from these perspectives, in the 1990s and 2000s, there were an important number of studies on Southern and Northern empirical experiments that were concerned with the impacts of decentralization and public involvement.
In summarizing the findings of research on decentralization, Robinson (2007) points to the absence of cross-national and cross-sectoral studies on the impacts of decentralization on service delivery outcomes. In a partial review based on available case studies, he concluded that improved equity outcomes have generally not been achieved and that the quality of public service provided has not improved as a result of restoring power and resources to local governments. Regarding the Brazilian experience, Medici (2001), Ugá, Piola, Porto and Vianna (2003), and Marques and Arretche (2004) drew attention to the fact that improvements in access to services in all Brazilian regions reflected increased resources invested rather than a more equitable distributive profile. According to these authors, this profile remained skewed in favor of more prosperous regions. Souza (2003) and Melamed and Costa (2003), on the other hand, stated that decentralization measures had increased equality in the distribution of resources and access to services across Brazil's regions, states and municipalities. In short, these empirical studies confirm the relevance of the theoretical approaches described earlier and, at the same time, call attention to the fact that these perspectives have been studied separately. They also highlight that there is currently no well-established knowledge about the quality of the processes (inclusion, legitimacy, involvement, and transparency), their capacity to impact the policy process, or the quality of the outcomes that are being associated with participation (innovation, distribution).
To deal with these questions, we developed a methodology that helps to assess how far public involvement and social mobilization have come in promoting inclusion, deliberation, and innovation in health policymaking. We also inquired about how these features may be related to changes in the distribution of public health services.
To investigate the extent to which public involvement promotes inclusion, deliberation and innovation, we followed the approach of a group of researchers (House and Howe 2000; Rowe and Frewer 2004;Abelson andGauvin 2006, Ansell andGash, 2007) who have highlighted the need to construct models that make it possible to analyze and compare arenas for citizen engagement. At the core, for these authors, is the possibility of identifying procedures and incentives that favor the expression of demands by those who have fewer resources. These authors also recognize the importance of investigating the location of these arenas within governance structures.
To deal with the impact of different types and trajectories of mobilization in the forums' performance, we followed another group of authors (whose ideas are published in Cornwall and Coelho eds. 2007) who suggest the importance of social mobilization processes in guaranteeing conditions so that actors that have fewer resources are able to participate. In parallel, authors who deal with the broad notion of social capital maintain that desirable levels of participation and deliberation, along with positive outcomes, can only be achieved in social environments that have at least some record of civic engagement and political mobilization (Putnam 1993;Verba et al., 1995;Costa 1997).
To identify changes in the distributive profile of public health services, we developed a geographically-based methodology that uses data from the Health Ministry and the Demographic Census to monitor distribution of basic appointments and hospital admissions (Coelho and Silva).
In this paper I present a preliminary set of findings that resulted from the application of this model, with the expectation that they will contribute to a better understanding of the conditions and mechanism that link participation to the policy decision making process. In the next section, before entering into this analysis, I briefly present the cycle of innovation that guaranteed the institutionalization of a robust framework for citizen involvement in Brazil. In doing so, I hope to make the nature of the mobilization processes to which I am referring clearer, as well as the type of health governance structure currently in effect today in Brazil.

Bringing the citizenry back in: describing a cycle of innovation
The cycle of innovation that made it possible for the SUS, the Brazilian Public Health System, to successfully tackle entrenched inequalities over the last twenty years is clearly related to the development of a new governance structure and can be summarized as follows: (i) In the late 1980s and in the 1990s, there was a bottom-up process with decentralized (civil society and municipal or state level) programs being created and successfully tested before becoming National Programs. At this stage, civil society associations and social movements played an important role in disseminating the notion of health as a citizen's right and engaged in a number of local initiatives such as, for example, creating local health councils and HIV-AIDS initiatives. One of the most successful Brazilian programs dealt with HIV-AIDS, wherein organized civil society pressed for a more comprehensive approach than the one recommended at the time by the World Bank and the government; (ii) Contracts began to be signed between the Health Ministry and the Municipalities for the full management of health activities. The contracts define responsibilities and transparent financing rules for the implementation of the national health policy. In this division of responsibilities the Ministry provides funding and sets policy but does not directly deliver services. At this stage, the "Brazilian Health Movement" played a crucial role, advocating for the effective institutionalization of the health conferences, a national health council, and also health councils in all twenty-six states and in nearly all of the 5,561 municipalities. These forums play a decisive role in regularly engaging civil society in a way that allows for challenges to be posed to the Health Authorities on policy rather than civil society merely having a participative role in implementation. As a result, there is increased citizen involvement, transparency and accountability; (iii) At the local level, the proportion of municipalities taking decentralized responsibility for aggressive programs of primary care increased from 23.4 percent to 88.7 percent. As an example, in 1995, 1 million users were registered with the Family Health Program (PSF); today 97 million out of 140 million Brazilian SUS users are registered.
As we can see, public involvement played a decisive role in three moments of this cycle and involved different strategies: citizen education, mobilization of social movements, and formal engagement in health councils and conferences. This brief overview of innovation cycles in the Brazilian Public Health System (SUS) and the role played by citizen participation suggests that the distributive achievements described earlier were dependent on both health system managers and social actors as well as on specific institutional arrangements that ensured regular debate between policymakers, health professionals and service user representatives. In the next section I focus on the methodology used to research the health councils which was implemented through this cycle, investigating the conditions under which they can effectively contribute to the inclusion of citizens in the policy process.

Researching the features that contribute to deliberation and policy change
Health councils (HCs) address core issues of priority-setting and accountability.
They also approve annual plans and health budgets. If the plans and accounts are not approved, the city does not receive funding from the Health Ministry. It is important to note that although their legal powers reside mainly in the technical and administrative spheres, the councils are especially significant for their role in policy discussion (Mercadante 2000). Their substantive contribution is the expansion of public spaces with the possibility of open discussion and deliberation on health policy.
The authors who have analyzed these councils have reached ambivalent conclusions about their capacity to impact policy; while a number of cases presented relatively little achievement, there were also a number of successful cases. In these analyses, success was sometimes recognized as the capacity to include marginalized citizens, and at other times, as the ability to work as schools of citizenship or as the capacity to present innovative proposals. At still other times, success was recognized as the capacity to promote distributive gains in favor of the poor. As previously noted, these successes were interpreted as being the result of good design, or of the organization of civil society, or of the involvement by committed public managers.
Nevertheless, these conclusions came through a collection of case studies, while there were not, in fact, methodological instruments to move towards a systematic comparison of these experiences. To fill this gap, we began to work on a model of analysis that would allow for evaluation and analytical integration of the drivers of change mentioned above.
The first step was to develop a model that differentiates between inclusion, connections and participation in the HCs. We also typified the debates and decisions that took place in the HCs. In doing so we worked out indicators that characterize features related to two big questions in the field: who participates and how.
Concerning inclusion, from amongst the many possible criteria, we considered as 'more inclusive' those indicators that reflected participants associational and political plurality, a demographic profile that mirrors that of the population, and a socio-educational profile with significant presence of the poor and less educated. In short, we considered socio-economic, demographic, political and associative characteristics 4 .
Concerning connections, we assumed as 'more connected' those forums that presented a high level of references -in interviews with councillors and minutes of meetings -to links with the executive and legislative branches at the municipal, state, and national levels. We also refer to the connections with other participatory forums, with other institutions in the health system and with other public and private organizations.
Concerning dynamics, we looked for the features that can countervail power asymmetries between participants, and promote accountability of the participants to their constituencies, as well as the councillors' own satisfaction with the process.
In order to investigate the impact of social mobilization on the performance of forums and on the ability to influence the policy decision making process, we selected cases (in areas with similar Human Development Indices) that had a significant history of social mobilization regarding health demands, and compared them with the results obtained for councils located in areas where there had been fewer of these mobilizations 5 .
Finally, we classified the debates held as: health issues, including discussions about health policies and programs and problems with service delivery; participation issues, dealing with procedures for elections and meetings; and local problems, such as water supply, infrastructure or security.
Inquiring about these features helps to describe the forums and produce data that can be used to test hypotheses related to the role of design, as well as of social and state actors in defining the performance of the forums. To understand how the decisions made by the HCs entered into the policy decision making process we interviewed public officials.
In parallel with the effort to qualify citizen involvement, we monitored the distribution of public health services in order to find a plausible relationship between participation, mobilization, and distribution.
Research took place in São Paulo, which has a population of more than 11 million and which is conspicuous for its sharp social inequality and unequal access to  (Coelho, 2006), as well as on research concerning the LHC performance (Coelho, Ferraz, Fanti and Ribeiro, 2010) and the distribution of health services in the Municipality of São Paulo, carried out between 2001 and 2008 (Coelho, Dias and Fanti 2010a). 8 The MHDI is constructed for each sub-municipality from the following variables: the per capita household income, the household head's average number of schooling years, the illiteracy rate of those aged 15 years and over and the population's life expectancy .

Municipal Human Development Index, 2000
To analyze the intra-municipal distribution of health services, we ranked the city's 31 submunicipalities according to their MHDI. For each submunicipality, the percentages of SUS users were calculated 9 followed by the consumption rates for primary appointments and for hospital admissions in the 31 submunicipalities. We followed this distribution from 2000 to 2008. In order to facilitate description, the submunicipalities were grouped into four quartiles according to the same index (MHDI).
The 6 LHCs selected for this study are located in poor regions of the city. Three we applied a questionnaire to eighty-five councillors -a mixture of service users and health managers 11 ; of the sixty-one who were service users, twenty-seven represented an association, while thirty-four did not. Third, we carried out participant observation of meetings. Fourth, we interviewed public officials working at the Technical Supervision.
Finally, we analyzed the distribution of health services in these areas from 2001 to2008.
We expected that in areas with stronger histories of social mobilization we would find LHCs with: a) better performance (broader inclusion; broader connections; broader 9 The calculation of the SUS population is a statistical inference based on data from the National Household Sample Survey -PNAD 2003 and the 2000 Demographic Census, both from the IBGE -Brazilian Institute of Geography and Statistics. This is done by obtaining, from the PNAD, the percentage of the population that does not have health insurance (SUS users) by family income stratum in the Metropolitan Region of Sao Paulo; using this information, the SUS population of each sub-municipality is calculated using the product of this percentage and the distribution of family income in each of these locations according to the census. and c) a higher increment in offer of services.
In the next sections I describe the research findings.

Changes in the distribution of healthcare services in São Paulo
Analyzing available data on the supply and consumption of public healthcare services in the city of Sao Paulo for 2001, Coelho and Pedroso (2002)  In 2000, 6 sub-municipalities had a rate of up to 499 hospitalizations per 10,000 users and 7 had a rate greater than 1,000. In 2008, only 1 sub-municipality had a rate of under 499 (Parelheiros) and only 1 was over 1,000 (Sé). Table 4 shows the admission rates for the municipal quartiles. 13 The Authorization for Hospitalization (AIH) is the means through which healthcare service providers in Brazil are reimbursed. Each of the procedures carried out at a center are reimbursed according to a payment chart. The number of AIHs has been used to oversee the distribution of the SUS's supply of hospital beds. AIH records indicate an address for those who used the SUS service, which allows for mapping of the consumption of hospitalizations in the 31 sub-municipalities. From the absolute number, provided by the Health Secretariat, we built a rate of hospitalizations, which is the ratio between these authorizations and the population that uses the SUS (per 10,000 users). In short, the data collected here clearly shows that the distribution of public healthcare services in Sao Paulo became more equitable. There was significant expansion in the supply as well in the consumption of services in the regions that present the worst socio-economic and health indicators. The analysis of the distribution of these resources showed that we now have a more equitable distribution pattern of public health services between locations with a reduction in the geographic inequalities hindering access to the public health system. What is not in line with this scenario are the increased differences (reflected in the rise in standard deviation) in the rates of basic appointments recorded in the sub-municipalities that have the worst socio-economic and health indicators.
Looking specifically at the areas under study and exploring the possible distributive impacts of participatory dynamics that are taking place in the submunicipalities, we found that of the eight sub-municipalities on the city's outskirts that presented low MDHIs, only two have municipal hospitals that were built in the period, namely Cidade Tiradentes and M´Boi Mirim, which have active councils. When we compared the number of Outpatient Health Units (AMAs) operating in these submunicipalities, we found that there were 16 units in the three sub-municipalities that have the most active councils vis-à-vis 10 in the areas where the councils are less active.
In this case, the second group should have had 15 units if distribution had merely followed population distribution criteria.

Comparing citizen involvement
The local health council (LHC) consists of 24 effective and 24 substitute councilors. In the LHCs studied, the councilors that represent civil society selfidentified themselves as representatives of: popular health movements; health units; religious associations; neighborhood associations; Unions; civil rights groups; participatory forums; homelessness movements; landless peasants movements; community or philanthropic groups; disabled persons associations; or as non-affiliated representatives (Coelho, 2006).
The tables and graphs presenting the indicators calculated for the three dimensions under study -inclusion, connections and dynamics -as well as the analysis of the debates held are presented in Annex II. The presence of a background of social mobilization proved itself to be an important factor in promoting more vibrant LHCs as well as in increasing the participation of the most vulnerable as shown by the greater inclusion of councillors with less education and women and non-whites on the councils located in the sub-prefeituras which had stronger backgrounds of mobilization. Also, the way debates are carried out is very different in areas with a greater or lesser history of mobilization. The results pointed out that in the sub-municipalities with a stronger history of mobilization, the LHC discussions were marked by more conflict and confrontation, but had better outcomes in the variables related to monitoring healthcare services and innovative proposals. As an example, in regards to the reduction of absenteeism, one suggestion made by the LHC that was implemented was to contact patients advising them of the date of the appointment. Also, monitoring of the construction of the two municipal hospitals built in the period helped in speeding the process. The organizations and the councils for these areas also present a greater number of connections with socio-political and institutional actors and have links to segments of bureaucracy, service providers, politicians and the civil society (Coelho, Ferraz, Fanti and Ribeiro 2010).
We also noted that in more mobilized areas this dynamic has contributed to promoting greater integration between the councils and their respective Technical Health Supervision Units (Coelho, Dias and Fanti 2010). In a situation of heated disputes over resources between sub-municipalities, this integration with councils has been welcomed by supervisors of the Technical Units. After all, those with the support and endorsement of civil society will be in a better position to negotiate their demands with the Municipal Secretariat of Health. The gains from this strategy are reflected in the increased ability to raise funds as shown by the three sub-municipalities which have more active councils. For example, as noted previously, the only two municipal hospitals opened in the period were built in Cidade Tiradentes and M´Boi Mirim.

Another example is given by the greater number of recently inaugurated Outpatient
Health Units in these sub-municipalities. These results help to explain a distributive tendency reported earlier: a reduction in inequalities in the supply of services among areas that have the best and the worst socio-economic and health indicators, as well as a slight increase in inequalities in distribution of basic services within areas with the worst indicators (Coelho, Dias and Fanti 2010).
The results also have shown that in both areas, with greater and lesser mobilization, 'vivid' participation is limited, inasmuch as few councillors raised issues and sustained discussions about them. In this sense, the creation of these LHCs, which are testaments to an impressive institutional process of building participatory forums, was not accompanied by innovation in the day-to-day operation of these spaces. In many cases they contributed, in the more mobilized areas, to simply reproducing the positions of health movements. Curiously enough, in the group with weaker backgrounds of mobilization, aspects associated with procedures -design and election themes -appeared more frequently, suggesting that these LHCs are looking for changes in their dynamics.
These results are summarized in connections -and the indicators that we have chosen to represent them run in different directions, highlighting the complexities of citizen involvement in the policy process. In the councils located in the sub-municipalities with a stronger history of mobilization, we found greater socio-economic inclusion, but less political and associative plurality.
Also, the discussions were less deliberative, marked by more conflict and confrontation and more resistance was offered to change in the procedures used to select representatives and organize the meetings. Yet, at the same time, better outcomes were presented in monitoring healthcare services and raising funds. On the other hand, the LHCs located in areas with less history of mobilization are the ones that worked out propositions to change procedures, which may favor new and more deliberative dynamics. They are searching for new ways to select the councillors and run meetings ).
On a final note, it is worth mentioning that in the last fifteen years growing resources have been expended in public health. However, the probability of continuing expansion is slight 14 . At the same time, other issues related to service quality, rising costs and an aging population are coming to the fore. In this scenario, it remains to be seen if the mechanism described that links mobilization, participation and distribution will endure or if new mechanisms, including more effective deliberation, will have to be considered. We turn to this discussion in the next section.

Reimagining citizen involvement in the SUS
In this paper I have departed from the literature calling attention to the potential of mobilization, participation and deliberation in contributing to the democratization of the decision making process and its increased accountability to citizens. I have also reviewed a number of studies that researched empirical experiments concerning citizens' involvement in the policy process. While the empirical findings confirm the relevance of the theoretical approaches to explain successful cases, they call attention to the fact that in most cases these perspectives have been studied separately. They also point out to the fact that there are several case studies while there are few systematic comparative studies in these fields.
In order to approximate these debates and allow for more systematic analyses, I have presented an innovative comparative research framework designed to make inquiries into the relations between forum features, social mobilization, the capacity of these forums to influence policy, and to promote distributive impacts. Research has been focused on health councils that are part of a national framework for citizen involvement in health policy implemented during the mandate of political parties that supported the cause of greater citizen involvement. These councils were conceived as public spaces that could promote deliberation between civil society, public managers and service providers on health policy.
This research framework was tested with a small selection of cases. Despite the number of cases analyzed, the results suggest that in a context of growing public health expenditures, federal inducement of health programs and governments that are supportive of the cause of participation, a positive association did occur between, on the one hand, mobilized social actors, their participation in health forums and the building of alliances with health managers, and on the other hand, a growing offer of health services to poor areas. This mechanism is probably similar to those that linked the health movement, health councils and health authorities throughout Brazil over the last twenty years and helped to promote greater equality in inter and intra regional distribution of public health services. However, the results also call attention to a negative association between both mobilization and deliberation, as well as to a slight increase in inequality in the distribution of basic services within the poorest areas of the Municipality of São Paulo.
These results should not be read as an indication of the need to insulate the councils from more politicized actors. To the contrary, mobilization brought dynamism to the forums and proved important in guaranteeing the inclusion of women, non-whites and non-educated people in the health councils, as well as in promoting the councils' connection with state, social and market actors, which helped to disseminate the debates and struggles. The result was an increased ability to negotiate and bring health services to poor areas, contributing to diminish inequalities. Nevertheless, the results also suggest that deliberative dialogue is very much needed as it can help to tackle the risk of exacerbated horizontal inequalities between mobilized and non-mobilized poor communities. Deliberation may also amplify the capacity of these forums to deal with 'second order' themes such as, for example, issues of quality, of priority setting in a context of less elastic resources, and of more controversial ethical problems related to life and death.
The remaining question is how a more positive relationship between mobilization, deliberation and distribution can be established. The question is not an easy one since a number of studies have already noted that in practice, rather than being spaces for the convergence of different interests, forums designed for citizen involvement in the policy making process are themselves an expression of specific coalitions (Coelho and Favareto 2008). In this scenario I would suggest that the answer cannot be sought exclusively in the daily routine of the political arena. An important part of the work will have to be done in another arena: the early education of public officials and politicians in deliberative techniques. Once they learn more about the rationale and procedures associated with deliberation, they may work towards the inclusion of a broader spectrum of actors in participatory arenas debates, as well as towards reducing the asymmetries between them, contributing to fostering genuine deliberative dialogues between experienced political actors and the less politicized.
In addition, the training of public officials and politicians can bring a decisive contribution to the development of more systematic connections between the councils and government bodies. These connections can be facilitated, for example, by the ways by which the debates and recommendations made by the councils are synthesized and presented (for example, format and scope). In the councils under discussion, for instance, there is still very little work that has been done to bring together the myriad of suggestions and criticisms worked out by the councilors. More investment in this area would go some distance in helping to rescue the richness of involving citizens, managers, researchers and service providers in policy debates as well as in preparing these materials to be used more effectively during other stages of the policy process.
Also core to strengthening the capacity of those forums to attract a broader spectrum of actors is the political decision of making the connection clearer between participation and the capacity to influence the channeling of resources in order to deliver public tangible benefits.
Another arena that deserves attention is the early education of citizens concerning their right to participate. Over the last twenty years, Brazil has built an incredible participatory governance structure, but citizens still know very little about its existence. One possible way to fill this gap is to include information on educational curricula, from primary education to higher learning, concerning the variety of accessible channels provided to citizens so that they may participate in the process of discussing, drafting, and monitoring a wide range of public policies.
In short, the Brazilian experience with health councils offers important lessons concerning the possibilities of building a national network for citizen involvement in health policy as well as the distributive effectiveness of mechanisms that link social mobilization, participation and public officials. This experience suggests that insofar as the political parties and system managers acknowledge social actors as partners they gained an important ally in their struggle to overcome a biased and inequitable distributive profile. This disposition to acknowledge social actors as partners is not a trivial one; also, the very existence of organized social actors interested and capable of acting as partners cannot always be relied upon. Nevertheless, the case also calls attention to the risks of these mechanisms in reinforcing the exclusion of the less mobilized while promoting greater horizontal inequalities. As a practical suggestion, in order to maintain gains and avoid risks, I would recommend that more attention should be given to the early education of citizens on their right to participate and to training public officials and politicians in deliberative techniques, since this may help in fostering the features of participatory governance related to inclusion of ordinary citizens, along with their values, positions and preferences, in the political process.