Transnational Models of Citizen Participation : The Case of Participatory Budgeting

This article pursues two main objectives. First, it provides a transnational overview and analysis of participatory budgeting, which has been central to the literature on democratic innovations in citizen participation. Second, it combines this broad empirical project with a theoretical approach based on the construction of ideal-types in the Weberian tradition. Namely, it presents six models of citizen participation: participatory democracy, proximity democracy, participative modernization, multi-stakeholder participation, neo-corporatism, and community development. Although these models have evolved from participatory budgeting and the European context, it is our contention that they can help us to understand the sociopolitical and ideological dynamics, contexts and impacts of civic engagement and democracy today at the transnational scale.

This article pursues two main objectives. 1 First, it aims to deliver a transnational overview and analysis of participatory budgeting (PB), which occupies a central place in the literature on citizen participation and democratic innovations. Second, it combines this broad empirical project with a theoretical approach based on the construction of ideal-types in the Weberian tradition. Namely, it presents six models of citizen participation: participatory democracy, proximity democracy, participative modernization, multi-stakeholder participation, neo-corporatism, and community development. Although these models were initially conceived in close connection with PB and the European context, we believe that these concepts can help us to better understand the sociopolitical and ideological dynamics, contexts and impacts of civic engagement, and democracy today at transnational scale.
Participatory budgeting, involving ordinary citizens in the spending of public funds, has been one of the most successful participatory instruments of the past 20 or 30 years. At the beginning of the 2000s, there is hardly an organization or territorial entity which would not subscribe to the virtues of greater civic engagement, at least verbally. In Western democracies, citizen participation is seen as a potential cure against the acute, though enduring, "malaise" or "crisis" of democratic representation (Torcal and Montero, 2006). In other parts of the world, citizen participation is increasingly required in the framework of international development programs or is the result of various bottom-up initiatives. This has led, particularly since the 1990s, to a global diffusion of participatory processes such as citizen juries, deliberative polls, neighborhood funds, and community development projects (Smith, 2009). The global spread of participatory mechanisms, despite their highly variable influence, and the parallel spread of non-democratic dynamics (Crouch, 2004), is still in its infancy, but this development represents more than the latest fashion trend. PB programs are forerunners in this respect, which is the reason why they constitute the starting point of this analysis.
We developed our transnational typology of citizen participation with two goals in mind: First, to conduct integrated fieldwork on participatory budgeting in more than 20 European cities, relying on the same methodology and the same concepts and to extend the methodology to other parts of the world where we can maintain the same definition of PB al., 2008, 2013b); Secondly, to 1 The authors would like to thank Janette Hartz-Karp and Brian Wampler for their comments on an earlier draft of this paper. Special thanks are also due to all the researchers that took part in the research "Participatory budgets in Europe," that has been conducted by the Centre Marc Bloch in Berlin, in co-operation with the Humboldt-University in Berlin and with funds from the Hans-Böckler-Foundation and from the CNRS (France). A previous shorter version of this paper has been published in Sociologias, XXX/14, May-June 2012, p. 70-117.

PARTICIPATORY BUDGETING: THE TRANSNATIONAL DIFFUSION OF A DEMOCRATIC INNOVATION
Participatory budgeting spread first in Latin America during the early 1990s, and then over the entire globe, hybridizing in contrasting ways. Any comparative world view therefore faces a definition problem, with no organization being able to control the label.

PARTICIPATORY BUDGETING: A DEFINITION
There is no recognized definition of participatory budgeting, either political or scientific, explaining the minimum criteria they must satisfy. Procedures called PB in some places would not get that label in others. Hence, there needs to be a definition that includes a set of minimal requisites to clearly differentiate this participatory procedure from others, while giving sufficient leeway to enable different specificities. Basically, PB allows the participation of non-elected citizens in the conception and/or allocation of public finances. However, five further criteria need to be added al., 2008, 2013b) : (1) Discussion of financial/budgetary processes (PB is dealing with scarce resources). All participatory devices might concern financial questions (for example, any participatory process related to urban planning will have an impact on costs if projects become bigger or smaller than previously planned). In PB, however, the participatory process is centrally based on the question of how a limited budget should be used.
(2) The city level has to be involved, or a (decentralized) district with an elected body and some power over administration and resources (the neighborhood level is not enough). In fact, we can observe a growing number of neighborhood funds where citizens can decide about a concrete amount of money, but without having any influence on issues that go beyond this level of a single neighbourhood.
(3) It has to be a repeated process over years. Consequently, if a participatory process is already planned as a unique event, we would not consider it as PB: one meeting, one referendum on financial issues are not examples of participatory budgeting 2 .
(4) Some forms of public deliberation must be included within the framework of specific meetings/forums. This means that if citizens are invited to discuss budgeting in local councils or in parliaments, we would not view it as sufficient, because PB should include specific institutions and therefore a new public sphere. Thus, PB should be based on some kind of deliberation. This is the reason why we do not consider a survey on budgeting issues in which citizens would remain without contact with one other as PB. However, PB deliberation does not necessarily directly lead to decision-making.
(5) Some accountability on the results of the process is required. We have observed that in many participatory processes, participants never get feedback about whether or not their proposals are accepted. This should be different in participatory budgeting, through annual meetings or publications where organisers provide information about the realization of the proposed projects.
With these criteria in mind, globally, there were between 795 and 1,470 participatory budgets in 2010. Around 200 cases were in Europe (Sintomer and al., 2010). This is the result of a very rapid development: Ten years before, in Europe, there were only a handful of PB programs. Similarly, in Asia, participatory budgeting is accruing great interest. The 2010 number of 40-120 Asian experiments is steadily increasing. Latin America had the highest number of participatory budgets at the end of the 2000s (510-920). In Africa, new procedures of citizen participation in budgeting are constantly being developed. From a global perspective, growth is considerable. When participatory budgeting emerged in Brazil in the 1980s, the country was transitioning from dictatorship to democracy, and was characterized by one of the greatest income gaps in the world. The huge social movements that shook Brazil for over nearly two decades were clamoring for both political and social changes. The new constitution adopted in 1988 was very progressive and open to citizen participation, but the political system remained characterized by corruption and clientelism. The context for Porto Alegre is also quite specific. Porto Alegre (with a population of 1,4 million in 2007), the capital of the state of Rio Grande do Sul, has always been dissident towards the central government. The standard of living was above the average of Brazilian cities, and social movements, especially urban movements, had been the most important in Brazil (Baierle, 2007;Avritzer, 2002Avritzer, , 2009. The city was also one of the strongest places of the Workers' Party (PT).
After some previous experiments in smaller cities, participatory budgeting crystallized in Porto Alegre due to a 'window of opportunity' which opened in the aftermath of the electoral victory of the Workers Party in 1988 (Santos, 1998;Abers, 2000;Fedozzi, 1999Fedozzi, , 2000Fedozzi, , 2007Baiocchi, 2001Baiocchi, , 2005Allegretti, 2003;Gret and Sintomer, 2005). It was not only the new left-wing local government that pushed the new participatory process. Civil society, in particular community associations, also demanded more co-decision-making capacity. The invention of PB was, therefore, the result of a conjunction of top-down and bottom-up processes. It was a pragmatic move, and not the application of an intellectual or political design. From 1989 to 2004, when the PT lost the office of mayor to the opposition after 16 years in power, PB was sufficiently institutionalized that the new government did not dare abolish it.
The Porto Alegre process is an institutional invention. The basic idea is to permit non-elected citizens to have a role in the allocation of public money, with direct decision-making power at the local level, the power of co-decision at the city level, and oversight capacity at all levels. The participatory pyramid has three levels: assemblies open to all in the neighborhoods, assemblies, and a participatory council of delegates in the districts, and a general participatory council at the city level. In addition to the meetings that are organized on a territorial basis, a complementary process that focuses on thematic topics (i.e., housing, urban infrastructures, healthcare, education, youth, culture, sport, and so on) takes place. The aim of the assemblies is to discuss priorities and to elect delegates who follow up on the development of suggestions put forward. Any individual who wants to participate in the public meetings can do so. Neighborhood associations have no special privileges, but they do have a decisive role in the organization of citizens. The municipal assembly, although it is entitled to accept or reject the municipal budget, has, de facto, a marginal role in participatory budgeting. Delegates are tightly controlled by the grassroots, can be removed, have a one year mandate, and their re-election is in theory limited. These features much reduce their autonomy and make them very different from usual elected representatives. At the city level, the PB council convenes once a week for two hours. Although there are accountability problems because leaders may not follow their supporters' interests, PB councilors should in principle ensure that the priorities of the districts are taken up in the budget to the largest extent possible. Independent NGOs train the representatives of the participatory budget to enable them to co-plan with the administration. The process has a oneyear cycle. Apart from technical control of the feasibility of the public works proposed by citizens, the funds that are at the disposal of each of the investment areas are distributed among the districts taking into consideration (a) the local list of priorities with the majority principle 'one person, one vote'; (b) the number of residents; (c) and the quality of the infrastructure or the service available (Genro and Souza, 1997;Herzberg, 2001). The embodiment of a principle of social justice has been one of the most original achievements of the experiment.
Finally, despite continuing challenges, the overall results have been surprisingly positive. Participation increased over time, peaking in 2002, with 17,200 persons taking part in the main district meetings, and many more at the neighborhood level. The social characteristics of participants are even more striking: lower income people tend to be more involved than others, women have become a majority in the assemblies, and young people are very active (Fedozzi, 2000). PB gives the floor to those who previously had been outsiders in the political system. It has led to the empowerment of civil society and, most notably, of the working class (Baierle, 2007). Clientelistic structures have largely been overcome, and relations between the political system and civil society have improved considerably (Avritzer, 2002). In addition, PB has led to a reorientation of public investments towards the most disadvantaged districts, at least those investments decided within the participatory process (Marquetti and al., 2008;Mororo, 2009): primary health care was set up in the living areas of the poor, the number of schools and nursery schools was extended, many streets in the slums were asphalted and most households now have access to water supply and waste water systems. This has come about because of the significant working class investment in the process, and because it has contributed to an improvement of public services and infrastructure.
The process has also led to a better government. Corruption, though not high in Porto Alegre, has been made more difficult. PB has been an incentive to reform public administration: a strong planning office has been created to enable discussions with the PB council, there has been more cooperation between administrations, new budgeting methods have focused on products and services, and the relationship between technicians and users has improved (Fedozzi, 1999(Fedozzi, , 2000. The main weakness is that the focus on annual investments has tended to side-line long-term investments, with the associated risk of PB decisions incurring expenses in the long run (maintenance and salaries) that are not sustainable (World Bank, 2008), or making it more difficult to develop a different urban form (Allegretti, 2003).
Despite these limitations, Porto Alegre has been the most important transnational reference for participatory budgeting and has remained one of the most fascinating experiments. It has convinced alter-globalization activists as well as local governments and advisors from international organizations such as World Bank and UNDP to support PB. To understand this success, we need to see it in the broader context of Latin-America, where PB emanated.

PARTICIPATORY BUDGETING IN LATIN-AMERICA
In Brazil, the progression has been impressive: in 2008 there were around 200 participatory budgets (Avritzer and Wampler, 2008;Wampler, 2010;Marquetti, 2005). The development in large cities has been even more remarkable: in 2005-2008, 41% of the cities with more than 100,000 inhabitants had set up a PB process; in total, more than 44 million people were living in places where the local government had decided to implement PB. This was due to the fact that in 9 of 14 cities with more than one million residents, participatory budgeting was established. PB spread throughout Latin-America, and by 2010, it has become one of the most popular instruments of citizen participation. Between 510 and 920 cities (out of 16,000) have introduced PB. This geographical dissemination has affected nearly every region in Latin America. Although it has been only very partially implemented, Peru's constitutional reorganization requires that PB should be implemented by all local governments (McNulty, 2011).  1989-1992 4,3% 1993-1996 9% 1997-2000 18% 2001-2004 40% 2005-2008 41%

Source: Calculation of the authors based on data provided by Leonardo Avritzer and Brian Wampler
The results of 30 years of PB in Latin America vary in direction and scope. Firstly, when it is well designed and implemented, PB increases transparency in the use of public money and reduces corruption (Zamboni, 2007). Secondly, it reduces clientelism and helps to fight corruption which further reduces clientelism (Avritzer, 2002). Thirdly, and crucially important to Latin America, PB is a powerful instrument in the redistribution of wealth towards the poor. A series of quantitative studies have shown that poor neighborhoods have tended to receive more investment than well-off ones (Marquetti and al., 2008;World Bank, 2008). Fourthly, although less frequently noted, when PB is articulated with a broader concern for the modernization and efficiency of public administrations, the two processes can reinforce one another (Fedozzi, 1999;Herzberg, 2001;Gret and Sintomer, 2005).
Summarizing, there is a wide spectrum of experiments. At one extreme, exemplified by Porto Alegre, "empowered participatory governance" (Fung and Wright, 2001) is characterized by strong political will together with bottom-up movements, and a methodology aimed at the devolution of power to local communities. This empowerment is part of a broader and deeper transformation of society and politics, and as a consequence, the massive inequalities that characterized Latin America during the last centuries have been called into question. In this way, participatory budgeting can be seen as a dimension of a larger process that has shaken Latin America, shifting the continent from dictatorships implementing neoliberal policies to democracies in which new governments try to promote another kind of development. At the opposite extreme -leaving aside the numerous 'fake' experiments -many examples of participatory budgeting in Latin America are primarily top-down and are not based on the mobilization of civil society. They involve limited amounts of money and have hardly any impact on the redistribution of resources. It is true that they can bring more transparency, social accountability and responsiveness, and reduce corruption. They may also help to mitigate inequalities. However, although formally they may be inspired by the Porto Alegre methodology, they are not geared towards political participation and empowerment. The World Bank, which in 2000 agreed to foster 'pro-poor policies', is playing a substantial role in the proliferation of these types of PB programs (See Goldfrank article in this special issue). Between these two poles, numerous PB projects are steered by proponents with a left-wing orientation or by NGOs that really want to change the development model, but lack bottom-up mobilization and a wider political perspective. Here, PB schemes have ossified into mere routine. Some early proponents of PB have denounced such schemes as examples of "participatory budgeting light" that have lost their soul (Baierle, 2007).

PARTICIPATORY BUDGETING IN EUROPE, AFRICA, AND ASIA
In Latin America, participatory budgeting procedures have been for a long time hybrids coming of the original Porto Alegre model. This has changed when the idea of PB has spread to other continents, first to Europe, and later also to Africa and Asia. The logic of diffusion has been combined with more complex learning processes starting from local experiments that were originally very different from the Porto Alegre model. What are the similarities and differences between PB experiments in these three continents? Are the dynamics and outcomes we have analyzed for Latin America reproduced elsewhere?

THE RETURN OF THE CARAVELS: PARTICIPATORY BUDGETING IN EUROPE
In Europe, the landscape differs significantly from that of Latin America. One might say that the caravels that carried the discoverers to the New World at the beginning of the modern age have now returned, bearing an innovation that brings citizens, elected officials and civil servants closer together. The demand for such innovation certainly appears to be strong: a relatively high degree of electoral abstinence and political disaffection is putting pressure on political systems in the Western world to demonstrate their legitimacy anew, and in many countries, local governments are struggling with financial problems (Dalton and Wattenberg, 2002), particularly since the global financial crisis. A growing number of municipalities are responding to these multifaceted challenges by developing participatory budgeting. Although their reference point is mainly Porto Alegre, the methodologies that are proposed most often differ from the original one al., 2008, 2013a,b). PB has spread rapidly in Europe, largely as a result of NGO activists and also local government politicians attending social forums in Porto Alegre. A particularly important role was played by those who attended the Local Authorities Forum, a parallel event of the World Social Forum. Proliferation in Europe has been impressive.

Figure 2: Number of participatory budgets in Europe and population involved
Source: Sintomer et al., 2013b At the end of the first decade in the new millennium, despite the demise of some participatory budgets, their geographic and numerical proliferation is notable. At the beginning of the second decade, the most dynamic further developments are probably in the UK and Portugal (Röcke, 2010). In contrast, Italy experienced a significant decline.  1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 > 200 >8.000.000 (2010) Source: Sintomer et al., 2013b The closest adaptations of the Porto Alegre model are found mainly in Spain and Italy (Ganuza, 2007). The most widespread approaches in Europe, particular in France, Italy, Portugal, and Scandinavia, are based on neighborhood meetings to seek to improve and strengthen communication among administrators, politicians, and citizens. Initially, in Western Europe, the social democratic or post-communist left-wing parties were those mostly involved in the dissemination of PB. Now, conservative governments are also active -and in Germany, PB has been non-partisan from the outset (Herzberg, 2009). In most West and North-European countries, various local government networks and state organizations also support the introduction of PB. In Eastern Europe, however, PB is mostly promoted by international organizations, such as the World Bank, UNO, USAID, or the German Society for International Cooperation (GIZ), in cooperation with their local partners. This type of PB, coming from the outside, has had the primary objective of mobilizing citizens and promoting good local governance. Processes of this kind often begin with the transparent preparation of public budgets. In a number of cases, a clear PB structure is displayed, such as Svishtov in Bulgaria or Elbasan in Albania (Shah, 2007;Co-Plan, 2005. Another common feature of participatory budgets in this region is that most of them involve pilot projects that come to an end when international support ceases.

Figure 3: Map on participatory budgets in Europe
As in Latin America, a clear link can be found between PB and the demand for more transparency (Shah, 2007). Although this has applied to the overall budget situation as well as allocations for participatory budgets, so far, transparency has not been sufficient to actually enable citizens to control the finances of the city. It is only in some cities, for instance in Spain, that PB constrains the administration to describe performance as well as administrative activities in a transparent and straightforward manner to both insiders and outsiders (Ganuza, 2007). One important criterion of sustainable development of PB could well be the link between participation and a comprehensive modernization process of public administration.
The potential political consequences of PB are diverse. In many cases, it has contributed to improved communication among citizens, administration and the local political elite. However, it is questionable whether it plays the same intermediary role as political parties have done in the past. Additionally, the widespread expectation that voter turnout will increase with PB has not been supported by empirical research. Although PB has positive influence on the political culture and competences of participants in Europe, its real long-lasting impacts is still unclear (Talpin, 2011). To date, European participatory budgeting has had less political overall impact than in Porto Alegre.
The contrast with the Latin-American situation is even sharper with regard to social justice. The small Italian city of Grottammare is one of the only examples of fundamental social improvements through participatory budgeting. However, subaltern groups have been mobilized for the process in some cities such as Albacete in Spain. Here, the ethnic group of Romani as well as migrants hold permanent seats in the participatory council and were able to achieve the construction of a community center that meets their needs. To achieve broad social justice, PB must include the participation of different groups and different social strata (e.g. through appropriate procedures) and distributive criteria. However, this has been the case only on a modest scale, in some experiments (Sintomer and al., 2013b).

AFRICA: A 'DONOR' LOGIC?
In Africa, a continent in which representative democratic structures and cultures are weak, some social movements and a number of local authorities have engaged in the process, but it remains highly dependent on the action of international institutions and NGOs (Sintomer and al., 2010). Progress has been slow, limited by its decentralization, due largely to the initial premise that innovation should be developed at a local level. However, a series of political reforms in the late 1990s drove attention to a wide range of management tools that could open the way for participatory democracy (Olowu, 2003). PB experiments are often 'catalysts' supporting and even accelerating the effectiveness of decentralization reforms and associated principles of transparency and responsiveness (often demanded by international donors), as well as pre-existing traditions of citizens' participation. This is perhaps why the second half of the 2000s has seen a visible acceleration supported by powerful institutions, such as the World Bank and the United Nations (especially the Habitat agency, based in Nairobi). It is impossible to deny the existence of a dose of 'neo-colonialism' in the way in which the idea of participatory budgeting entered the African political debate. Cases, such as Fissel in Senegal, where local governments and citizens movements have led the initiative are exceptions. Local adaptations are difficult to classify. Especially in Anglophone Africa, participatory budgeting has merged with other tools, whose main objectives are the 'demystification of budgeting', the 'traceability of investments' and the 'consensual development planning'. Many processes follow the logic of multi-stakeholder participation (see below), with aims ranging over a multiplicity of governance principles linked to the improvement of decentralization and the fulfillment of the UN-defined Millennium Development Goals (UN-Habitat and MDP, 2008). The main limit of these practices is their 'donor-based' perspective: processes respond mainly to the goals of the donors (of the international community), i.e. the transparent management of budgets. The 'rights of citizens' that could increase the overall level of democracy constitute only a secondary goal. The path that the Latin American radical movements fear is globally the one that has been taken in Africa. At the same time, the hybrid nature of African participatory budgets could play a positive role, opening new possibilities for poverty alleviation strategies and consolidating decentralization. This could lead to new models conceiving democratization as a substantive issue based on resource redistribution, access to education, knowledge, power, and the 'right to the city'. The Cameroun experiment of Batcham considered as a good example in this direction.

PARTICIPATORY BUDGETING IN ASIA: BETWEEN AUTOCHTONOUS DEVELOPMENT AND INTERNATIONAL EXCHANGES
Compared to Africa, participatory budgeting in Asia emerged later, but has shown an important growth. In contrast with other continents, initial experiments tended to be mainly autochthonous innovations rather than the result of transnational transfers, although their principles and methodologies had similarities with those of Latin America or Europe (Liu and Traub-Merz, 2009). This was especially the case with the Kerala experiment in India. Overall, the differences in methodologies and political significance of PB in Asia make it difficult to draw a global picture. In addition, political structures are much more heterogeneous in Asia than in Europe or Latin America, with a spectrum that includes federal and centralized states, constitutional monarchies with parliamentary governments, unitary presidential systems and single-party states. The diversity in cultures and standards of living is striking. A common factor is that the birth of PB took place in a period of accelerated economic development and, to a lesser extent, in a phase of progressive decentralization (UCLG, 2008). Overall, PB in Asia is something of a mixed bag (Raza and Thébault Weiser, 2006). International exchanges have increased in a second phase, but their contribution to a unified notion of participatory budgeting is unclear. The terms 'participatory budget' and 'participatory budgeting' started to be used only around 2005 in Asia, with explicit reference to Brazil. The first experiment that entered directly in contact with the European or Latin American debates was that of Kerala state (India), which received an international recognition through left-wing scholars (Fung and Wright, 2001) and alter-globalist movements. The Kerala participatory process took shape in 1996 (Neunecker and Mastuti, 2013). The idea came from younger party leaders of the Marxist CPI-M party. The launching of the process was a political decision, but it opened the door to a huge social movement that gave shape to the experiment. Nowhere else, except in some places in Latin America, has PB been a channel for such mass mobilization. People elected delegates to follow the process in every phase, having a say in prioritizing, implementing, and monitoring the consensually-elaborated demands to be inserted in local and supra-local development plans. Over the 13 years of its existence, the 'plasticity' acquired by Kerala's experiment (Heller, 2001;Chaudhuri and Heller, 2002) enabled it to survive the political changes which twice changed the state government (Jain, 2005), and the results of this Indian state as regards health care and life expectancy can be compared with European ones.
While China shares with India a number of economic, and social features, its political structure is completely different, and the growing interest in participatory budgeting is embedded mostly in top-down processes (He, 2011(He, , 2013Wu and Wang, 2012). The concept was discovered in the mid-2000s and widespread interest apparently grew after the 'Sunshine Finance' revolution that championed the development of budgetary transparency in order to enhance the performance of government. In such a huge country, the main difficulty in identifying examples of PB in a comparative perspective is the ambiguity of the Chinese concept of 'participation'. In a context in which information often remains the monopoly of the executive and Communist Party leaders, the notion is not necessarily related to the direct involvement of the people in public policies, but is often used instead to designate practices of inter-institutional dialogue involving members of the legislature (the Local People's Congress deputies have traditionally been excluded from the determination of the municipal budget), information disclosure, public notification and -in the best case -legislative hearings, public opinion polls, inquiries, and surveys (Sintomer and al., 2013a). In some cases, participation also implies negotiations with organizations such as private enterprises, residents' committees or NGOs. Only a few experiments rest on the active involvement of 'ordinary' citizens and can be considered 'real' participatory budgets, the best example probably being that of Zeguo (He, 2013). This initiative mixes Porto Alegre's notion of getting citizens to decide investment priorities with randomly selected citizens' assemblies (adapted from the methodology of deliberative pooling invented by Fishkin, 1998).
China is important, but not because the trend towards citizens' budgets is especially strong there when compared from a transnational perspective. Instead, China is significant because the ruling CCP abjures political pluralism and prefers to modernize the state administration and develop local participation under autocratic conditions (He, 2011). As in other areas of social policy, when presenting participatory budgeting in China we come up against the qualification 'with Chinese characteristics'. Generally speaking, this means that the CCP's monopoly of political power is sacrosanct. Administrative reforms have priority. The new obligation on authorities to disclose their budgets to the public and, for example, to make them accessible on the Internet is intended to improve accountability and limit the scope for corruption. Budget consultations are aimed at improving the position of the People's Congress with regard to the Political Executive. Deputies of the People's Congress, who have hitherto not seen itemized budget breakdowns and were asked merely to rubber-stamp whatever was sent down from the Executive, can now carry out consultations with citizens and experts in pilot projects in order to improve their competence in investment planning. One might call them public budgeting rather than participatory budgeting, since decision-making is shifted from the administration closer to the People's Congress, not to the population. However, Zeguo, which has been an exception, could well be a 'best practice' example for new and dynamic PB experiments that have begun to take place in the 2010s. Chengdu, most notably, could become a model for 'weak democratic countries'. In this 14 million city, with 2000 villages, PB implies budgeting transparency, a significant amount of money, and even the possibility to decide of spending immediately or putting aside the money for building bigger public facilities in the next year.
In Japan and South Korea, two rich countries that are members of the OECD, the social, economic, and political context does not have much in common with that of India, and even less with that of China. PB has developed as a tool to address problems linked to resource scarcity, incomplete decentralization, weak accountability, and limited responsiveness of elected institutions to the needs of their citizens (particularly the poor). Japan's constitutional monarchy shares a number of problems with Korea, such as the strong influence of national parties on local elections, declining local election turnouts (below 50 per cent), rising corruption among officials, and the inflexibility of national transfers to local budgets (UCLG, 2008). In Japan, local governments have broad functional responsibilities and account for over half of total public expenditure and 10 per cent of GDP. This strong formal role is matched by the extensive power given to citizens to call for local referendums, improve or eliminate ordinances, audits, and even the dissolution of the local assembly, as well as the dismissal of the mayor, council members or officials. Despite this, citizens' participation in public policy-making is infrequent, especially in the field of financial planning. The first attempt to involve people in budget issues was perhaps the active involvement of some grassroots organizations, which were allowed to legalize their status in 1998. After 2003, various processes got under way involving citizens and grassroots organizations in the discussion of public budgets. The Coalition for Legislation to Support Citizens' Organizations distinguishes between a number of types: transparency in budget-making processes (sometimes merely informational); counter budgetmaking by citizens' committees; delivering the budget to the community; citizens carrying out budgeting; and 1 per cent of residential taxes handed over to nonprofit organizations for projects on which citizens vote. The city of Ichikawa stands out in this respect, where the participatory budget uses 1 per cent of residential tax revenues for non-profit projects. Other Japanese cities have been inspired by the Ichikawa experiment. The term 'participatory budgeting' is entering into general use. In 2009, Ichikawa organized a '1 per cent summit' with a view to developing a network to share experiences of this exceptional approach to community development participatory budgeting (Matsubara, 2013).
In Ichikawa, the use of funds in the citizens' budget is determined by taxpayers. Linking tax payment with participation brings to bear, in a modified form, the early principle of 'No taxation without representation', which was directed by British colonists in America towards the English monarchy. But should involvement in citizens' budgets be bound to the possession of a taxable income or the right of residency alone? Ichikawa has left open a backdoor for housewives, the unemployed, students, and schoolchildren, the main groups excluded by the tax qualification. Community service points are distributed for voluntary community work which, converted into money vouchers, entitle their bearers to vote on citizens' budgets.
At the beginning of the 2000s, South Korea is the Asian country in which PB has developed most, and, as far as this experiment is concerned, it is indeed one of the most dynamic countries in the whole world. In South Korea, citizen participation has a strong tradition, as mass mobilizations were a decisive factor in the progressive democratization of the country in the 1980s. Nevertheless, the development of PB is also linked to a framework in which the size of local government debt and the capacity of local authorities to borrow has been under strict central government control since 2000, when President Rho Moo-hyun emphasized participation as a means of bringing about 'trust and confidence' in institutions. His mandate (2003/2008) was labeled 'Participatory Government' and contributed much to the rapid expansion of PB in the country. The concept was introduced as a bottom-up process, but its diffusion has been stimulated topdown by the national government. One of the most outstanding examples is Dongku (Songmin, 2013). Formally, the key principles of PB in Korea have been imported from Porto Alegre, but have been locally adjusted, giving birth to a 'reduced version' of the Porto Alegre model in that it lacks the social mobilization. PB is a joint decision-making process in which all ordinary citizens can take part in making proposals at a first stage, while groups of delegates follow up on the remaining steps. It consists of locally based meetings in which every resident in the area can participate, and a city assembly that gives a pivotal role to a citizens' committee on participatory budgeting (subdivided into 5 thematic committees). This committee is appointed partly by means of open recruitment and partly through recommendations made by community organizations. All members are trained for their tasks at a so-called 'participatory budgeting school'. A number of tools (such as internet surveys, online bidding, cyber forums, online bulletin boards, public hearings, and seminars) have been provided in order to foster non-exclusive processes for all citizens in every phase, and the tradition of citizens' budget schools and budget policy seminars is one of the most important South Korean contributions to the rest of the world in this field.

SIX MODELS OF CITIZEN PARTICIPATION
How can these highly different developments and adaptations of the Brazilian process be integrated in a systematic framework? How is it possible to go beyond the specific case of PB in order to present a more general analysis of citizen participation?
A DESCRIPTIVE OVERVIEW It is obvious from the panorama we have drawn that there is no uniform model in any continent to which the others could be compared. Overall, a descriptive overview can identify three different trends. At the most radical level, we see participatory budget programs that aim to fundamentally change prevailing conditions and are one component of a broader movement for renewal. They are based on interaction between governments and grassroots movements. These budgeting procedures are about overcoming social injustice and achieving sustainable development. This means breaking with established traditions of patronage and corruption. When civil society is mobilized, the pressure it exerts helps to achieve this. This typifies many Brazilian (particularly Porto Alegre) and other Latin American examples. The outcomes of participatory budgeting in Europe and Africa appear to have less radical impacts. In Asia, Kerala, where the Porto Alegre model has been combined with rural development, is one of the few examples of the more radical approach.
The second trend involves the use of PB to drive a reform agenda. Although this does not involve a break with former practices, these initiatives do have a real impact. The local government is the lead player here, but citizens are not absent. There are some clear rules, or a routine that allows newly established practices to become the rule. Aims include administrative modernization, deepening decentralization processes, and social impacts, such as improving the lives of socially disadvantaged groups, while retaining the basic structure of the system and existing patterns of allocation. The greatest impact of reform, however, involves an improvement in relations between local governments and their citizens. This includes that local governments are willing to implement suggestions put forward by citizens, which can be seen as a confidence-building and trust-inducing measure. In the global south and in Eastern Europe, this kind of PB is often supported by international organizations.
The third trend is evidenced when PB is largely of a symbolic nature and in which there is a yawning gap between the proclaimed objectives and the reality. Here the aim is no longer really to consult citizens. Meetings are used rather to legitimate a path that has already been embarked upon. For example, citizens' suggestions concerning the design of an austerity package are not sufficiently analyzed and/or incorporated into the policy-making process, and there are no ways to improve citizens' capacity. Symbolic participatory budgets are found both in established democracies and in authoritarian regimes. Particularly in the latter case, they are intended to express an ostensible openness that in reality does not exist; participation is designed to placate the population and/or international financial donors.

SIX CONCEPTUAL MODELS
These three trends, however, do not reflect the complexity of participatory budgeting in the world -let alone other participatory devices. Hence, we propose a conceptually more complex typology, following a Weberian approach that aims to develop ideal-type models. Ideal-types never completely reflect the empirical data, but they allow classifying and systematizing the puzzling variety of real cases. They compose a conceptual map on which one can situate empirical cases. Invariably, the models need to be combined in order to explain a particular experiment -as with a road map, typically, you don't travel precisely towards North, South, East or West, but the existence of these cardinal points help so you don't get lost (see Figure 4). Our models are constructed by showing differences around six criteria: Socio-political context, normative orientations, participatory rules and procedures, the dynamics of collective action, the relationship between conventional politics and participatory processes, and the strengths, weaknesses and challenges of each participatory model. 3

PARTICIPATORY DEMOCRACY
We have chosen to call the first model participatory democracy. This term 'participatory democracy' is often used as a catchword, referring to the majority of approaches that in some way bring non-elected citizens together in the decision-making process, even if this is purely consultative. We decided to give the notion a narrower meaning: for us, it means that traditional mechanisms of representative government are linked to direct or semi-direct democratic procedures -in other words: non-elected inhabitants (and their delegates who are invested with a 'semi-imperative mandate') have de facto decision-making powers, although de jure the final political decision remains in the hands of elected representatives. This is the meaning that most contemporary social scientists have attached to the term, and the notion is an explicit normative frame.
Alongside anti-authoritarian socialism, it constitutes the inspiration of our first ideal-type.
This model is mainly characterized by the simultaneous emergence of a 'fourth power' (participants have a real decision-making power, different from the judiciary, the legislative and the executive) and a 'countervailing power' (the autonomous mobilization of civil society within the process leads to the empowerment of the people and the promotion of cooperative conflict resolution). The traditional mechanisms of representative government are linked to direct democratic procedures, where non-elected inhabitants (and their delegates who are invested with a 'semi-imperative mandate') have de facto decision-making powers, although de jure the final political decision remains in the hands of elected representatives. In this model, participation has real repercussions in terms of social justice and relations between civil society and the political system. Essentially, the countervailing power in combination with the political will of the government contributes significantly to an inversion of priorities in benefits of the poor. Here, the logic and general orientation of distribution is transformed, which is much more than the mere involvement of marginal groups, as may be possible in the models of proximity democracy. These effects are most likely to occur in countries of the Global South, while in the Global North the support of marginalized groups is usually limited to selected neighborhoods. The model of participatory democracy is based on the participation of the working class and not just the middle classes, thereby creating an emerging plebeian public sphere. This creates a positive equation between conventional and non-conventional politics, as the dynamics of the two can combine. Local governments are active in the launching of the process but also in the implementation of decisions. In such a model, citizen participation is a left-wing flag and is conceived as an alternative to neo-liberalism and as part of a broader social and political reform process. However, the modernization of administrative action is not always take center stage, as has been the case in Porto Alegre or Belo Horizonte in Brazil.
A number of Latin-American participatory budgets exemplify this model. Seville (Spain) and Dong-ku (South-Korea) share some of its characteristics, but could be called a 'light' version. Kerala fits to a certain extent, but shares some dimensions that refer more to the community development model described later. The same can be said for Fissel (Senegal), Villa El Salvador (Peru), and to a lesser extent Cotacachi (Ecuador). Beyond PB, this model also reflects other citizen participation processes like the constituent assemblies in countries like Bolivia, Ecuador, and to a lesser extent, Venezuela.
Abstractly, one could argue that this model is the most politically and philosophically stimulating, because it combines a strong participation with effects on social justice. However, historical and sociological analyses demonstrate that it can work only under specific circumstances, and that other models could therefore be more appropriate in certain contexts. The weakness of participatory democracy is that it requires strong political will and a mobilized and independent civil society that is ready to cooperate with local governments. It won't last if confined to the local level only. The main challenges include efforts to successfully link civic participation to administrative modernization, and avoid the risk of co-opting the mobilized members of civil society into the institutional framework (which would cut them off from their own grassroots).

PROXIMITY DEMOCRACY
The key characteristic of the second model is that it showcases proximity both in terms of geographical closeness and increased communication between citizens, public administrations and local authorities. Although local governments have some real power, their public administrations are not necessarily involved in a strong modernizing process. Proximity democracy is based on 'selective listening': its logic is that the decision-makers cherry-pick citizens' ideas. Proximity democracy is grounded in informal rules and leaves civil society with only marginal autonomy. Over and above ideological rhetoric, it constitutes more of a 'deliberative turn' of representative government than an inroad into a new kind of democracy -a deliberative turn that will perhaps not be recognized by the theoreticians of deliberative democracy because of the low quality of deliberation which often characterizes these devices. Although left-wing local governments tend to commit a bit more easily to proximity democracy than right-wing ones, there is no clear dividing line. Proximity democracy is not an instrument of social justice, even if it may guarantee some degree of solidarity (for example, by limiting real estate speculation or introducing policies of urban renewal). As the process is merely consultative and civil society does not have much independence, a fourth power or a cooperative countervailing power seems to be excluded. Proximity democracy is essentially top-down. It often addresses individual volunteer citizens, but NGOs also play a considerable unofficial role; moreover, a number of participatory instruments that fit this model use random selection to select 'ordinary' citizens. It would be hard for proximity democracy to generate any remobilization within the framework of conventional politics (a claim that has often been made by local governments), because it deals mostly with 'small things' that seems far away from the competitive party system. Proximity democracy is the most widespread model in Europe, and is supported by neighborhood funds and councils. The same could be said in North-America, Australia, Korea or Japan. In the countries of the Global South, this model is also widespread. In some cases, such as Roma XI (Italy) or Lisbon (Portugal), proximity democracy experiments can be viewed as very "light" versions of participatory democracy and can move from one model to the other according to the political evolution. Other experiments, such as the French region Poitou-Charentes or Dong-Ku Ulsan (Korea), are in between proximity and participatory democracy.
The proximity democracy model is characterized by a low degree of politicization and a low level of mobilization (particularly of the working class). Its main strength is improving communication between citizens and policymakers. Its weaknesses lie in the essentially arbitrary way in which policymakers 'selectively listen' to (cherry-pick) people's perspectives. In this case, the government only accepts the proposals that are in line with their own plans. Hence, participation only has a legitimizing function for decisions that have already been made before. Furthermore, the NIMBY perspective constitutes a problem if discussions are only centered on neighborhood issues without any consideration of the interests of other neighborhoods or the common good of the whole city. The main challenges of this model are to ensure that participation is effectively coupled with decision-making in order to enhance the quality of service delivery; and to combine proximity with state modernization beyond neighborhood level.

PARTICIPATORY MODERNIZATION
The key feature of the third model is that participation is only one aspect in New Public Management strategies, in a context in which the state is trying to modernize in order to become more efficient and legitimate -and in some cases in order to resist the pressures to privatize. Viewed from this angle, the participatory process is top-down, is not political and has only consultative value. In contrast to proximity democracy, modernization is not necessarily focused on the neighborhood level, but includes also the central administration and its main service providers. Participants are considered clients; hence there is no interest in the integration of marginalized groups or in the launch of social policies. Civil society has only limited independence and there is no space for either a fourth power or a cooperative counter-power. What is at stake here is quite different: participation is first and foremost linked to good management and is aimed at increasing the legitimacy of public policies. Politics remain in the background, so that users or clients of public services are of concern, rather than citizens. The people involved are mainly middle class, except when specific procedural measures are used to improve the sociological diversity of the participants. The normative frames on which this model is based are closer to the participatory versions of New Public Management than of participatory democracy as an alternative to neoliberal globalization.
In terms of participatory budgeting, this model is influential in Germany, and to a lesser extent in Northern Europe. Experiments such as Bagira (one of the three municipalities of Bukavu, Congo RDC), Zeguo (China) or Cologne (Germany) are in between participatory modernization and proximity democracy. Other participatory tools to improve management reflect this model (for example, consumer charters, panels and inquiries, as well as hotlines). Similarly, neighborhood councils and neighborhood management can be part of this perspective. Countries outside Europe have also taken advantage of this approach. For example, many experiments in China that try to integrate ordinary citizens can be interpreted in the context of participatory modernization, because users can contribute to the improvement of public services. The public transport of Shanghai has already set up an elaborate process of consumer participation (Hu, 2009). Many municipal authorities of very different political affiliations have introduced PB procedures that reflect this model. The most outstanding difference may be that participatory budgeting does not need to be linked to representative democracy, as we have illustrated with the examples from China.
The strength of this third model is the close link between the modernization of public administrations and participation, and the fact that crossbench political consensus can easily be achieved. The flipside is that there is only a low level of politicization, which makes it difficult to introduce broader questions, particularly that of social justice; processes close to the model tend to be purely managerial in nature. The challenges include how to increase the participation and autonomy of civil society and develop a genuine political dimension (instead of becoming a merely technocratic procedure) in order to provide politics with renewed impetus.

MULTI-STAKEHOLDER PARTICIPATION
The main characteristic of the fourth model is that the citizens who take part constitute just one of the many different actors, together with private enterprise and local government. In this model, local politics appears to have only limited room to maneuver, compared with economic forces and where the donors call the tune. The approach is weakly politicized and the major development issues of local politics can be discussed only peripherally. Although participatory procedures may well have decision-making powers, they remain caught in a topdown approach that does not enable a cooperative countervailing power to emerge.
Rather than an emerging fourth power, participatory instruments of this type represent an enlargement of governance mechanisms (whereby private economic interests gain an institutional influence in the decision-making process). In the multi-stakeholder participation, civil society is weak and has little autonomy, even if the rules of the decision-making process are clearly defined. It is essentially middle class individuals who take part, and the projects are aimed at active citizens or NGOs, who are supposed to be the spokesmen of local residents. International organizations such as the World Bank or the United Nations play an important part in dissemination. In this model, participation often serves policies that have incorporated the constraints of neoliberal globalization: in the best case it represents only a small weight with which to counter these forces; but there is no intention to change the orientation of power relations and the logic of distribution of public resources. The dominant normative frame is a hybrid of the rhetoric of governance and the theme of citizen participation.
As far as PB programs are concerned, this model exists in Eastern Europe, for example in Płock (Poland); the donor-based participatory budgets of Africa could also share some features of this model, especially when external actors like United Nation Organizations or National Development Organizations try to support the financing of projects defined by the local population. African experiences share often also some features of the modernization approach, especially when PB is linked to processes of decentralization. With regard to other participatory instruments it has considerable influence in the Anglo-Saxon world.
The multi-stakeholder participation model includes private companies that are fundamental to local development but which tend in other models to remain outside the participative process. However, this comes at the cost of private enterprise having the upper hand in a process in which they have voluntarily become involved (and on condition they clearly profit from their involvement), whereas civil society is limited to a subordinate role and is not able to question the dominant economic and political framework. This is why it seems to be diametrically opposed to Porto Alegre. The challenges facing this model are how to link participatory instruments with the core business of municipal politics: that is, to stabilize the financial flows on which they depend (these flows are precarious and do not operate under clear legal constraints); to balance the weight of the various stakeholders involved in the process, and to open up to topics that are relevant to them; and to counter the pressure to transform NGOs and associations into quasi-governmental organizations or semi-commercial entities.

NEO-CORPORATISM
The distinctive trait of the neo-corporatist model is that local government plays a strong role by surrounding itself with organized groups (NGOs, trade unions, and employers' associations), social groups (the elderly, immigrant groups and so on) and various local institutions. In this model, government aims to establish a broad consultation with 'those who matter' and tries to achieve social consensus through the mediation of interests, values, and demands for recognition by the various factions in society. In this model, the political leanings of local governments vary, as does the dynamics of modernization of the public administration. The normative frames are linked to neo-corporatism and certain variations of the concept of governance: the World Bank, for example, does not really distinguish the corporate and the NGO 'private' stake-holders and does not make the relative strength of local inclusion a dividing criterion.
In the neo-corporatist model, the participatory rules may be formalized, while the quality of deliberation is variable. In most cases, local neo-corporatist processes are essentially consultative. Even though civil society does play a considerable role in them, its procedural independence is fairly limited, and they are essentially top-down processes. This is why the emergence of a cooperative countervailing power -or of a fourth power -is unlikely to occur. The outcome is more a reinforcement of traditional participation than a virtuous circle of conventional and non-conventional participation (or the substitution of the former by the latter). At national level, the classic neo-corporatist approaches, particularly those used to manage the health care system, often work in very different ways: they may be highly formalized, have real decision-making authority and confer decision-making powers on the social partners.
The neo-corporatist model is dominant in Local Agenda 21 processes (where different local stakeholders meet to discuss common topics but have no power to realize their proposals), or in participatory strategic plans (where governments invite different groups to round table talks). In the context of PB, this model has had only limited influence, most notably in Spain, where the blend with the Porto Alegre approach has given birth to forms of associative democracy (most notably in Albacete) and the appeal to sectorial participation in various aspects of municipal public administrations might lead to the development of original forms of participatory budgeting.
International organizations play a considerable role in disseminating this model. Its main strength is the linkage between the main organized structures of society, which facilitates social consensus around certain aspects of public policies. However, it is characterized by asymmetrical relationships of power and non-organized citizens are excluded. The main challenges are linking participation and modernization, avoiding co-optation of associations or NGOs (that may become cut off from their roots) to public management, or going beyond a simple cherry-picking approach and successfully discussing the most controversial matters.

COMMUNITY DEVELOPMENT
The dominant characteristic of the last model is that participation includes the phase of project implementation, in a context that dissociates municipal politics and a strong participatory process driven as much by a bottom-up dynamic as topdown. The margins for representative politics are fairly small in this ideal-type.
The fourth and cooperative countervailing powers that emerge are therefore not closely linked to local institutions, which is an aspect that distinguishes community development from the participatory democracy model. The influence of Porto Alegre is blended with older community traditions. There are fairly clear procedural rules and a relatively high quality of deliberation. The most active participants are the upper fraction of the working classes or middle classes, because they are involved in running the community associations. The role of NGOs is often decisive, with participation being aimed at disadvantaged or marginalized groups with a view to inclusive action rather than at an overall form of distributive justice. In a configuration such as this, the partial substitution of non-conventional participation linked to community activities for conventional participation (party membership and voting in elections) is fairly likely to develop.
In the field of participatory budgeting, this model has developed in the Anglo-Saxon world, for instance in Canada (with the Toronto Housing Company), or in the United Kingdom, where it predominates (the experiment of Tower Hamlets, London, can be seen as emblematic). It is also exemplified in other countries of the Global North, for instance the Japanese 1% PB of Ichikawa, and in many countries of the Global South, for instance in indigenous towns such as Cotacachi (Ecuador), in rural villages such as Fissel (department of M'bour, Senegal), or in poor suburban communities such as Villa El Salvador (Peru). Some experiments, such as Seville (Spain) or the state PB of Rio Grande do Sul (Brazil), may less easily be located on our conceptual map, because they mix various influences and methodologies. Other forms of community development have emerged in the World and have become one of the most particularly widespread instruments of citizen participation, from the Community Developments Corporations in the United States to the various forms of community organizing, both in the North and in the Global South, and to the neighborhood councils in Venezuela.
The political inclination of local government is not a decisive factor; the normative frames are those of empowerment, of Saul Alinsky's community organizing, but also of guild socialism, left liberalism, Paulo Freire's pedagogy of the oppressed and sometimes the traditions of local communities, particularly of indigenous ones. This participatory model has clear advantages in a context in which local government is weak and where, conversely, civil society has genuine independence and a real tradition of organizing that enables the community sector to manage local projects by themselves. The weakness lies in the fact that it is difficult to build an overall vision of the town, as well as the tenuous links between participation, modernization of the public administration and institutional politics. The challenges that the model faces include trying to keep the management of community organizations free from managerial influence and to stop them from turning into Para-public bodies; moreover, processes of this type need to look beyond the micro-local level and contribute to the transformation of institutional politics.

CONCLUSION
In this article, we have described the spread of participatory budgeting across the continents with a complex mix of transfers, adaptations, and autochthonous innovations. As Esping-Andersen observed, the institutional shape of welfare state has taken different forms, it is apparent that the PB model invented in Porto Alegre has emerged in diverse models throughout its journey around the world (Esping-Andersen, 1990). This applies to the process of exchange between Latin America and Europe but also to Africa and even more Asia, where original experiments are under way with Kerala's participatory development, Chinese deliberative polling PB, Japanese taxpayers' budgets, and South Korean versions of the 'Porto Alegre' original model. It seems clear that there is no single telos toward which participatory budgets in the world are moving. The six models we proposed (participatory democracy, proximity democracy, participative modernization, multi-stakeholder participation, neo-corporatism, and community development) tend to confirm this analysis. In an ideal context, the participatory democracy model has clear advantages: it constitutes a real innovation in the institutional framework of representative democracy in providing ordinary citizens with real power, and has a strong dimension of redistributive justice. Yet, it doesn't embrace all the possible advantages a participatory process can bring in (and most notably, in Europe, the link to administrative reform), simply because this is impossible. Moreover, there is always a certain trade-off between the implementation of radical innovations, which are far-reaching and dependent on favourable political conditions and might therefore be abolished after a change in government, and less radical innovations, which are more easily supported by a broader political consensus and are therefore potentially more sustainable. A process that can be combined with certain existing traditions of participation might lead to more transformative results than an 'artificial' process with no links to existing structures. On the other hand, radical innovations seem necessary to challenge the present asymmetric power relations within most common participatory devices and in society. This dilemma is not easy to resolve, and it is one of the reasons why there are multiple ways towards more just and more democratic urban development in the world, depending on the situation, rather than one 'royal road'.
Participatory budgeting is only one important example of a larger diffusion of democratic innovations. This is one of the reasons why we do not limit our typology to PB. It seems fruitful to build models that go beyond empirical descriptions, and especially that do not focus only on the instruments of citizen participation and take into account factors such as the socio-political context, the normative frames and dynamics of collective action, or the relation between conventional politics and participatory processes. We would have largely succeeded if other works make use of -and probably modify -this typology in order to better understand other participatory devices. It would be of special interest to explore how other instruments relate to the features of participatory budgeting that we discovered in the different parts of the world.
A mere dichotomy (such as authentic vs. fake, or radical vs. neo-liberal, or bottom-up vs. top-down PBs) is inadequate to understand the complexity of this trans-national mosaic. Some more general questions are however worth asking. Will PB and more broadly citizen participation only become another tool of participation in the agenda of international organizations, state and local governments? Will they be part of a broader movement of social and political change? Will they really modify the relations between local citizens and the municipal government, as well as between the later and the central state?
The future is open. It seems highly probable that no one answer will be given, and that further developments will add to the design of a complex mosaic. Future developments will depend on the national and local contexts, on transnational transfers of experiences, on the political will of national and local governments -but also and fundamentally on the involvement of civil society and grassroots social movements.