Outsourcing Participatory Democracy: Critical Reflections on the Participatory Budgeting Experiences in Taiwan

This article attempts to critically examine the experiences of participatory budgeting (PB) that have proliferated across Taiwan’s cities over the past three years. It is argued that PB in Taiwan remains an isolated initiative instead of an integral part of a comprehensive administrative reform. What makes Taiwan’s PB experiences theoretically interesting is that, in most cases outside the Taipei City, the tasks of promoting PB, designing the procedure of participation, and organizing and mobilizing lay citizens have been contracted out to NGOs or teams led by scholars. In light of the analytical framework of state power political society civil society, this article argues that (1) PB in Taiwan takes place in the absence of active support from civil society and confronts a political society that is hostile to or skeptical of PB; (2) the “outsourced” model of PB generates incentives for the commissioner to evade administrative and political responsibilities, and imposes structural constraints on the performance of the contractor regarding mobilization, organizing and deliberative quality; and (3) the future of PB in Taiwan depends on whether the current modus operandi of PB will give way to a more comprehensive institutional reform and whether a growing number of active citizens and civil society organizations can fill the new political space created by PB that may otherwise be occupied by vested interests and political elites. Author Biography Poe Yu-ze Wan is Associate Professor at the Department of Sociology, National Sun Yat-sen University, Taiwan. His current research interests include political sociology, social theory, and philosophy of social science. He is one of the first scholars to introduce the idea of participatory budgeting to Taiwan, and he himself was in charge of a PB project in the Kaohsiung City in 2016.


Introduction: Participatory Budgeting in Taiwan
Since 2015, participatory budgeting (hereafter PB) has been successively implemented in all the major cities in Taiwan.Initiators of PB include departments and agencies in both central and local governments and several local councilors.The former have experimented with PB on different themes, scales, and with different sizes of public budget, while the latter have adopted PB in their districts to decide how to spend part of their discretionary fund.Among these cases, Taipei is the only municipality that attempts to institutionalize PB on a city-wide basis (see Table 1 for a classification of the major cases of PB in Taiwan).
Despite some variances, the majority of cases of PB in Taiwan exemplify a "local community funding approach" (Sintomer et al., 2016, p. 131), which understands PB as a way to let the community decide how to spend a defined public budget via an open (and sometimes online) vote.Policy makers and scholars in Taiwan do occasionally invoke the Porto Alegre experience.However, and importantly, what characterizes the Porto Alegre model -its emphasis on social justice and redistribution (e.g., Baiocchi and Ganuza, 2016;Ganuza and Baiocchi, 2012;Sintomer et al., 2016, pp. 13-4;Röcke, 2014, pp. 45-56;Fung, 2015) -was absent in Taiwan's PB from the very beginning. 1evertheless, among both policy circles and the academia in Taiwan, PB has soon become a (or yet another) catchword for a long-awaited type of democratic innovation that aims to decentralize the decision-making structures, strengthen the dialogue between civil society and government, and empower lay citizens.While PB practitioners in Taiwan tend to report what is innovative and progressive about the experiences of PB across the major cities, I hope to put forth some critical remarks on these experiences especially in terms of their weaknesses and pitfalls.I may be in an advantaged position to do this, since I have been heavily involved in the introduction and implementation of PB since the start of 2015.More specifically, I have served as a consultant to several city administrations, have been responsible for training hundreds of public officials about PB, and have constantly talked with high-rank public officials, rank-andfile city staff, case officers, and NGOs and scholars that collaborated with city governments in a variety of PB projects.This research is accordingly based on my extensive participant observation of half a dozen cases across the major cities (mainly New Taipei City, Taichung City and Kaohsiung City).Nearly a hundred semi-formal and informal interviews were conducted in the form of extended conversations with many of the practitioners of PB around Taiwan in 2015-17.From February to November in 2016, I also led a team in charge of a PB project in Kaohsiung, in which I was not only a scholar or an observer, but also a dedicated organizer.
The focus of this article will be placed upon what I take to be unique to PB in Taiwan: Most of the cases studied here adopt an "outsourced" approach by contracting out nearly all aspects of PB to the private or voluntary sector.The government plays a minimal role, restricting itself to deciding on the amount of resources allocated to PB and to implementing certain winning projects.In the following I will first develop an analytical framework suitable for capturing the dynamics of democratic innovations that necessitate state-society synergy.Then I will utilize this framework to specify the context in which PB emerged in Taiwan.I will also discuss the ways in which the outsourcing system creates incentives for and places structural constraints on the practitioners of PB, before I outline the main factors affecting the future development of PB in Taiwan.Fung and Wright (2003) call "empowered participatory governance" (EPG).I submit that to better understand the dynamics and evolution of EPG, it is necessary to start from the premise that "specific configurations of civil and political society can have markedly different implications for democratization" (Baiocchi et al., 2011, p. 14).Besides, since EPG is characterized by "state-society synergy" (e.g., Evans, 1997;Abers, 2000Abers, , 2003Abers, , 2009)), the state or state power has to be taken into account. 2In other words, the state, civil society, and political society are three key elements of an analytical framework that helps reveal the driving forces behind and the consequences of any democratic innovation that aims to move beyond conventional citizen participation.Let me elaborate on these elements further.Firstly, the state is understood here as "the cluster of institutions, more or less coherently organized, which imposes binding rules and regulations over a territory" (Wright, 2010, pp. 118-9).Its "administrative, legal, bureaucratic, and coercive system" structures and regulates -to varying degrees -the social relations within both civil and political society (Stepan, 1988, p. 4).It should be noted that it is convenient but often misleading to treat the state as a homogeneous whole.Sometimes it is both theoretically fruitful and empirically necessary to distinguish between, for example, different departments of the government, or between the political appointees and career civil servants.In this research, entities that fall under this category include local governments (mainly individual departments and agencies), mayors and their political appointees.

PB is essentially an exemplar of what
Secondly, civil society is loosely defined here as the collection of people, institutions, and practices of voluntary associational life (Baiocchi, 2005, p. 18, Wright, 2010, p. 119;cf. Baiocchi et al., 2011, p. 26). As Somers (1995, p. 230, quoted in Baiocchi et al., 2011, p. 18) argues, the civil society is "a 'third' space of popular social movements and collective mobilization, of informal networks and associations, and of community solidarities that sustain a participatory public life symbolized not by the sovereign individualism of the market or by the state."Indeed, the employment of the concept of civil society can be highly normative (e.g., used as a weapon to defend democratic gains against market forces).For instance, Wright's line of argument is that the state (political power), the economy (economic power) and civil society (social power) constitute three domains of social interaction and power relationships (Wright, 2010(Wright, , 2013; see also Young, 1999).A "civil society" thus understood excludes market relations and capitalist economic power,4 and is afforded the potential of transcending capitalism, as Wright's project of "real utopias" suggests.However, when the research is focused on the dynamics of EPG, a less normative concept of civil society is likely to carry more explanatory weight, since EPG may run counter to the vested interests of economic elites and thus provoke conflicts between, for example, labor unions and business interest associations.In this research, the main civil society actors include militant social movement organizations, the community-building movement (as broadly conceived), advocacy NGOs, local elites, and ordinary citizens.
Finally, the classic definition of political society is the "arena in which the polity specifically arranges itself for political contestation to gain control over public power and the state apparatus" (Stepan, 1988, p. 4).Again, I prefer a loose definition: political society is the collection of people, institutions, and practices related to the pursuit of political leadership and power.It usually involves the contestation over the forms of political representation and legitimacy, the electoral system, the ways of exercising public power, etc.The major political society actors in this research are local political elites, including village chiefs (leaders of the village and neighborhood system) and councilors.
Following the tradition of relational political sociology (Baiocchi, 2005(Baiocchi, , 2009;;Baiocchi et al., 2011), the "configurations" of civil and political society denote the ways in which individual and collective actors (e.g., political groups, political factions, trade unions, social movement organizations, community associations, etc.) in civil and political society position themselves and relate to one another (cf.Baiocchi, 2005, pp. 20, 139-40;2009, p. 120).Important dimensions include the degree to which they are organized and/or mobilized and the strategies with which they cooperate and compete with one another.To fully grasp the theoretical richness of "configurations" means to take seriously human agency and its "environments" (or structural contexts), including cultural, social-structural, and social-psychological environments (Emirbayer, 1996).By way of "the interplay of habits, reflection, and judgment," the agency of individual and collective actors "both reproduces and transforms those structures in interactive response to the problems posed by changing historical situations" (Emirbayer and Mische, 1998, p. 970).
The state, civil society, and political society are three strategic action fields with relative autonomy.That is to say, the actors in each field are "are attuned to and interact with one another on the basis of shared (which is not to say consensual) understandings about the purposes of the field, relationships to others in the field (including who has power and why), and the rules governing legitimate action in the field" (Fligstein and McAdam, 2012, p. 9).That said, the three fields intersect and interact with each other in diverse ways.It is therefore necessary to take into account their interrelations and interactions (see Table 1).The following sections will utilize this analytical framework to identify the key actors in the three domains and the mechanisms that structure the evolution of their (inter)actions.
Figure 1 The state -civil society -political society analytical framework

A Policy Device without Active Support from Civil Society
In the diffusion or transfer of policies, "interpreting, translating and adapting" the objects of diffusion is always necessary (Roggeband, 2004, p. 162, quoted in Röcke, 2014, p. 28).In the process of its diffusion to Taiwan, PB was soon adapted to the modus operandi of both central and local governments.Mainly cast as a way of delivering social services or collecting community-level budget proposals, PB in Taiwan was intended neither as a democratic innovation directed towards empowering lay citizens and achieving social justice goals, nor as part of a set of comprehensive institutional reforms.Therefore, there are remarkable similarities between PBs in Taiwan and those in the U.K. in that both can be conceptualized as "participatory grant-making" or "community grants" (Sintomer et al., 2016, pp. 145-56).But what is probably unique to Taiwan is that both the central and local governments play a minimal role in PB, since in most cases outside the Taipei City, almost the entire structure and process of PB are outsourced.In this and the next sections I will explain why and how this happened.
Like the majority of PBs in Europe, the recent "participatory boom" in Taiwan has been characterized by a stronger top-down than bottom-up mobilization.These participatory practices were exclusively initiated by policy-makers (from the top-down), not by citizens or social movements (from the bottom-up).More specifically, the main impetus for experimenting with PB in the major cities has been the competition between leaders both in the city government (mayors and their political appointees) and in the political society (a few political elites from the two major parties in Taiwan, i.e., Democratic Progressive Party [DPP] and Kuomintang [KMT]).The non-partisan Taipei mayor, Ko Wen-je, first included the idea of PB in his platform during the campaign for the mayorship in 2014, and started to put it into practice in 2015.The New Taipei City (ruled by KMT) and the Taichung City (ruled by DPP) soon followed suit in the same year.They were joined by other cities in 2016.Ganuza and Baiocchi (2012) make a useful distinction between "policy instrument" and "policy device."In its early development in Latin America, PB is closer to a policy instrument because it represents "a very specific way of orienting the relationship between political society, civil society, and the state" (Ganuza and Baiocchi, 2012, p. 2).It is generally anchored in a broader political strategy that aims to radically transform the structures of public administration.By contrast, since the 2000s, PB has been gradually disconnected from a broader set of institutional reforms, turning out to be an isolated, "neutral" policy device conducive to good governance.
To put it briefly, PB in Taiwan is such a policy device introduced from the top down, and so far it has not had significant impacts on civic engagement and associational activities.The main reason is that it (1) takes place in the absence of active support from civil society, and (2) confronts a political society that is generally skeptical of or hostile to popular participation.
As scholars on Taiwan's politics and civil society point out, Taiwan has witnessed the "revival of civil society activism" since 2008 (Hsiao, 2016), when KMT returned to power after being in opposition for eight years.The dissatisfaction with KMT (and especially its pro-China and pro-business policies) culminated in the Sunflower Movement in 2014, which occupied the legislature for 24 days and involved more than half a million people in Taiwan (for details, see, e.g., Wan, 2015;Ho, 2015;Hsiao, 2016).
Importantly, much of this burgeoning social activism was transmitted into electoral politics, as KMT faced consecutive defeats in the 2014 local elections and in the 2016 presidential and parliamentary elections (Hsiao, 2016).The victories of DPP in the presidential (56.12 percent of the votes) and parliamentary (68 out of 113 seats) elections were overwhelming.
From all this it seems to follow that there exists a vibrant civil society in Taiwan.I believe this holds true, at least in part.A high degree of associational activity can be discerned from the rapid increase in the number of registered civil associations -from 3,960 in 1980 to 18,465 in 2001 and 59,181 in 2015.As regards the political and economic pressures China exerts on Taiwan, some even assert that the only way to counteract the detrimental effects of the "cross-Strait political and business alliance" is to form a "cross-Strait alliance of civil society," so that such "universal values" as democracy, human rights, civility, and multiculturalism can be defended (Wu, 2012, pp. 55-60).And here comes the question: why didn't this civic effervescence in Taiwan translate into a demand or support for PB?I submit that there are two main reasons.
First, in the words of Fung (2015, p. 520), the problem lies in "the lack of a broad popular articulation and agreement on the role of non-electoral public participation in contemporary democratic institutions."Since electoral participation remains what Rosanvallon (2008, p. 20) calls "the most visible and institutionalized expression of citizenship," this renders it difficult to champion alternative forms of public participation (Fung, 2015, p. 521).While social activism fueled electoral-representative politics in Taiwan in the past few years, leading to a landslide victory of the opposition party, new forms of democracy such as PB have not garnered equivalent attention and support among lay citizens. 5econd, the more progressive sections of civil society in Taiwan, especially the advocacy NGOs and social movement organizations that have played important roles in the democratization process (e.g., Hsiao, 2003;Wang, 2007;Ho, 2010;Chuang, 2013), are not particularly attracted by those village-or communitybased cases of PB in Taiwan.On the one hand, the transformative value of such a "community grant" version of PB is dubious, especially when the size of the budget is extremely small (usually less than 6,000 U.S. dollars for each winning project).On the other hand, these organizations are mainly concerned about specific values and policies (e.g., environmental protection, workers' rights, long-term care, same-sex marriage, etc.).But there is virtually no space for debates over medium-and long-term policies during the PB process in Taiwan. 6ven in thematic PBs that took place in several cities, the discussion is generally geared toward short-term projects instead of policies. 7nother vibrant force in Taiwan's civil society is the quite heterogeneous community-building movement, which can be traced back to the "Comprehensive Community Development Policy" proposed by the Council for Cultural Affairs in 1994 (see e.g., Lu, 2002).Organizations grouped under this label include, among others, community colleges, community-building organizations, and community development associations.To uncover the devil in the details, the role of community development associations, which exist in a large number of villages, requires some clarification.According to the Civil Organizations Act in Taiwan, there can only be one such association in each community area, 8 and in practice, the majority of them have origins in local elections and factional politics (Chiang and Chang, 2016, p. 382).Some of these associations have to compete with the village and neighborhood system for resources (usually via clientelistic networks and/or community project funding from the government), but sometimes it is the village and neighborhood system that pulls the string. 9Importantly, many of these associations, controlled by local (political) elites, simply don't (even try to) establish themselves on a popular basis, and are highly dependent in their relationship to the political society and the state. 10In fact, the very raison d'être of many of them is to access funding from the government and to campaign for the elections of village chiefs.It is thus not surprising that the aims of community empowerment and participation would fall outside their purview.This has been a deep-seated problem for local democratic governance in Taiwan (see Chiang and Chang, 2016, pp. 371-94 for a cogent critique).
When community-or village-based PB is introduced to communities dominated by these elites in local civil and political society, it may be expected that PB will play a role in counteracting their powers by bringing ordinary citizens or more progressive forces (e.g., those more experienced in community organizing) into the decision-making process.This did happen in a few cases (e.g., the Zhengjue community in Tainan City 11 ) .Yet, as will be discussed in the next section, generally the existing power relations tend to be reproduced rather than questioned.
Without active support and commitment from the more militant sections of civil society, in the words of Baiocchi and Ganuza (2016, p. 155, emphasis added), "the most sensible path for PB is its being implemented through the path of least resistance; that is, through budgets of less importance to bureaucrats and powerful interests."However, what matters is not only the type and amount of resources allocated to PB, but also the way PB is designed, promoted, organized, and understood in the context of local politics where clientelistic practices have 8 Up to 2016, there were 6,881 community development associations in Taiwan.(Source: Ministry of Health and Wealth, Taiwan.) 9 The village and neighborhood system is the basic administrative unit in Taiwan, and it straddles civil society, political society and the state.Village chiefs (or ward chiefs) are elected public officials, many of whom are active in local civic organizations.A substantial portion of village chiefs (especially in rural areas) are intertwined with patronage connections in local politics.The clientelistic networks formed around this system best embody what I call the logic of elite-mass relations. 10Even those organizations more committed to community empowerment and grass-root democracy were sometimes criticized as depending too much on government-led projects."The ways in which communities work betray a top-down manipulation" (Wang, 2015; see also Liu, 2008). 11I am grateful to Prof. Kuo-Ming Lin for bringing this case to my attention in the conference "Participatory Budgeting in Taiwan: A Dialogue from Within and Without" held in Taipei, Feb. 18-19, 2017.prevailed for decades.This in turn brings us to the other important aspect of the pitfalls of PB in Taiwan.

Problems of "Outsourcing" Participatory Democracy
Anyone who wishes to evaluate the experiences of PB in Taiwan has to take seriously the striking fact that, in many cases (mainly outside Taipei), the entire structure and process of PB before the final implementation of certain winning projects, including procedure design, promotion, mobilization, deliberation, and voting, are outsourced to NGOs or teams led by scholars (the "contractors" in legal terms).I suggest that this "outsourced" model of PB, arguably unique to Taiwan, deserves more critical attention and has broader comparative implications.
This trend of outsourcing everything, including participatory democracy itself, should be analyzed in terms of "government by contract" (Freeman and Minow, 2009), or the market model of management (Box et al., 2001), which has become the dominant methodology for the provision of public services in Taiwan since the 1990s (see e.g., Tang, 2004). 12The point is not just that this trend, informed by the philosophy of New Public Management, turns "public goods into commercial goods and citizens into consumers or clients" (Brodie, 2007, p. 102), or results in "the fragmentation of the sector and the intensification of competition that result in detrimental effects for the 'public ethos'" (Bartocci, 2016, p. 2).Even more worthy of note is the way in which outsourcing almost everything about PB to "participatory democracy experts" and NGOs becomes a path of least resistance and risk taken by certain local governments.
Laura Pin (2017, p. 131) notes in a recent paper that PB in Chicago "relies on extensive volunteer labor, with some paid support from aldermanic staff, but minimal support from municipal staff."Things are similar but more striking in Taiwan when considering the fact that the main initiators of PB are not aldermen (councilors), but (municipal) governments themselves.It is understandable that undertaking a comprehensive institutional reform may risk a strong backlash from within the administrative machine, and to restrict the amount and scope of PB and the degree of its institutionalization by outsourcing also serves as a political signal that PB does not pose a challenge to the power of political elites; hence a path of least resistance and risk.But PB without sufficient support from the government is far from a "state-society synergy," but a shirking of responsibilities in the guise of public-private partnership.
The logic of contracting out involves cost control and efficiency enhancement, which is at least different from (if not totally opposed to) the rationale behind participatory democracy, i.e., popular sovereignty, civic activism, and empowerment (see e.g., Box et al., 2001;Baiocchi and Ganuza, 2016).More specifically, there are at least two problems with this outsourced model.First, the extent to which lay citizens are mobilized and involved in the PB process depends excessively on the performance of the contractor, which in turn has to do with the contractor's understanding of PB and its working method.This involves the following questions, to name a few (cf.Baiocchi and Ganuza, 2016, pp. 145, 149): a.Is PB a tool for collecting "sophisticated" budget proposals, or an instrument for empowering lay citizens?b.Which to prioritize: courting support from local political elites, or organizing those outside the existing political networks?c.Is it necessary or desirable to involve participants in a deliberative process leading up to the decision by, for example, creating a series of mini-publics?Or is voting all that matters?If deliberation is important, should these mini-publics become a point of contact between city officials and citizens?
Most importantly, the existing power relations in both civil and political society will remain intact if the contractor adopts a more elitist and non-deliberative approach.Take a city-wide thematic PB in a southern city for example.The contractor (a team led by university professors) did not focus on organizing lay citizens, but took a shortcut by contacting a semi-official organization that consisted of public officials and local elites (e.g., village chiefs and community leaders) and had existed in that city for more than a decade.It turned out that the participants of the PB workshops were mainly the members of this organization, one of whom even reported that she was mobilized by the organization to attend the workshop and had no idea what PB was about. 13A number of similar cases were found where the contractors chose to bypass ordinary citizens, casting PB as yet another channel for local elites to access funding opportunities and policy information.The existing elite-mass relations in local civil and political society therefore remain unchallenged.However, sometimes the contractor is not the one to be blamed.For instance, my own team was asked by the commissioner from time to time to make sure that the local political elites (mainly village chiefs and councilors) would support the PB process.This means that the commissioner prioritized its relationship with the city council and thus attempted to avoid conflicts with these political elites.
Even if the contractor intends to mobilize and organize as many ordinary 13 Interview with a member of this organization, May 2017.citizens as possible14 and enhance the quality of deliberation,15 it will face substantial financial difficulties, since it is generally the lowest tenderer that is awarded the contract (in the spirit of competitive tendering).There were indeed impressive cases.For example, in New Taipei City, the Ludi Community College quite successfully organized the immigrant residents in a thematic PB on energy-saving.Similarly, in Taoyuan city, the Serve the People Association and the Taiwan Reach-Out Association for Democracy worked together to organize and empower migrant workers in another thematic PB.But these successful cases were more a consequence of the self-exploitation of these NGO workers and project assistants16 than a proof of the superiority of the outsourcing system.Therefore, the point is not that contractors can never realize the core values of participatory and deliberative democracy, but that the outsourcing system puts structural constraints on how far they can go.
Second, according to the Government Procurement Act in Taiwan, the commissioning entity is obliged to direct and monitor the contractor.This means that after signing the contract, the contractor and the commissioner are no longer on an equal footing.Importantly, the NGO or the scholar-led team, now as a contractor, is not in a position to intervene in how things are done within the commissioning agency, not to mention the entire public administration.
One consequence of this is that the contractor, left to itself, is unlikely to initiate administrative reforms necessary for upgrading the transformative capacity of PB.For example, unless the commissioning agency or the whole city government recognizes its role and responsibility, cross-agency/sector collaboration is unlikely to occur.It should be remembered that crossagency/sector coordination is extremely important in PB, because the implementation of the proposed projects often requires collaborative efforts of two or more agencies/sectors.To my knowledge, some of the winning projects in Taichung and Kaohsiung were not implemented precisely because there was not sufficient cross-agency coordination.
In some extreme cases, the commissioner even refused to send city staff to neighborhood meetings because this was not required by the contract.The contractor was therefore relegated to a marginal position, striving to make a difference to the logic of bureaucratic conduct but in vain.Besides, my fieldwork indicates that tensions may arise between (1) the contractors that attempt to challenge the clientelistic practices (or more generally, the logic of elite-mass relations) prevailing in local politics and (2) the commissioning entities that aim to adapt PB to the existing power relations.These conflicts have led to considerable frustration and disappointment among a number of scholars and NGOs that had served as contractors for PB projects.
In short, the outsourcing system is a mechanism that tends to create incentives for the commissioner to avoid administrative and political responsibilities, and put structural constraints on the performance of the contractor (see Figure 2).I call it a mechanism in the sense that it consists of "entities (with their properties) and the activities that these entities engage in, either by themselves or in concert with other entities" (Hedström, 2005, p. 25).The outsourcing system brings different actors together in an institutional context that regulates the actor's choices and their interactions, and thus tends to inculcate certain patterns of behavior that have the causal power to bring about change or resist it.It should be further noted that the mechanism-based approach does not aim to identify law-like regularities (e.g., outsourcing always leads to undesirable outcomes) but tries to analyze how a variety of mechanisms interact to produce the "outcome patterns" (Pawson, 2006, pp. 17-37;Wan, 2011).
Therefore, while the outsourcing system is causally efficacious, it alone does not determine the outcome in a particular context, in which there are always multiple mechanisms at work.As we have seen, the causal powers of the outsourcing system may be counteracted by other causal powers, such as the dedication and self-sacrifice of NGO workers and project assistants.Or, to take the PB in Taichung in 2016-17 as another example, the contractor, with very limited resources, attempted to collaborate with the local community-building center not only to involve those more popular-based community-building organizations in the PB process but also to pile pressure on the commissioning agency.
Figure 2 The outsourcing system as a mechanism for "defanging" PB The only city in Taiwan that has not adopted the outsourced model is Taipei, where more than 1,000 civil servants have been trained for and/or taken part (however reluctantly) in PB since 2015, and the internal division of labor of the city administration was slightly modified to meet the needs for implementing PB.However, the mobilization of citizens in Taipei relied too much on the district offices and the village and neighborhood systems, since it is the Department of Civil Affairs (the competent authority in charge of these offices and villages) that is responsible for organizing PB.
The experiences of PB in Taipei may suggest that it would be better for experienced NGOs to assume the responsibilities of mobilizing citizens and organizing deliberative processes. 17I have two points to make here.First, it is undoubtedly true that progressive forces in civil society are more experienced in mobilization and deliberation, but this does not in itself prove that contracting out is the only viable method.In fact, the Taipei City Government has been experimenting with other methods of collaborating with civil society organizations.The main problem is that it could not have allocated sufficient resources to these efforts as a result of some city councilors' objection (another indicator of how the political society is skeptical of PB).
Second, PB is not only about mobilizing citizens and designing deliberative mini-publics.It is essentially a democratic innovation that necessitates statesociety synergy.In other words, the current structure of public administration has to be challenged before any meaningful progress can be made in the way city staff interact with citizens and deal with budget issues.If the government plays a more direct role in PB instead of contracting it out, it will more or less be forced to learn how to do it well by, for example, systematically training the 17 I am grateful to one reviewer for pointing this out.
city staff about participatory and deliberative democracy, reorganizing its internal division of labor, and taking seriously the necessity of cross-agency collaboration, and so on.Nothing of this kind can be expected or demanded of a contractor.
My overall worry is that the transformative value of PB will be seriously limited if the city governments continue to prefer outsourcing to institutional reform.Practiced in this way, PB in Taiwan may turn out to be a "toothless radicalism," or a "defanged" version of democratic innovation, which is likely to lead to what Mazeaud and Talpin (2010, p. 369) call "politicized" and "apolitical" disappointments among citizens.On the one hand, those more politicized and progressive may consider PB to be a state-led attempt to reinforce clientelism.On the other, those who tend towards political apathy or cynicism may denounce PB as a continuation of, rather than a break from, old ways of doing politics, and conclude that participatory institutions make no difference.

Conclusion: Whither Participatory Budgeting in Taiwan?
Figure 3 summarizes the relationship between PB, the state, civil society, and political society in Taiwan.Of course, it simply attempts to capture the general situation, since there are variances between different types of PB and between (and even within) cities.The solid line represents substantial influence, and the dotted line indicates that the influence is weaker.The main pitfall of PB in Taiwan is the absence of active engagement among the more progressive sections of civil society, the shifting of responsibilities from the state to the contractors, and the possible domination of PB by the political society.
Figure 3 The state -civil society-political society analytical framework applied to PB in Taiwan Upon its introduction to Taiwan, PB was cast as an "external tool" that may influence governance from outside the institutional architecture, and its very existence depends almost wholly on the will of individual political appointees or mayors, because the system of public administration is not reformed or restructured correspondingly.The critical description of PB in the U.K. offered by Sintomer et al. (2016, p. 155) applies just as well to Taiwan: Although the growth in the diversity, number and scale of new experiments is still very positive, the relatively small sums involved in many of the initiatives claiming to be participatory budgeting and the limited scope of many of the projects, coupled with limited evidence of institutionalization at the local level make it still uncertain whether participatory budgeting can make a meaningful difference to the power and influence available to ordinary citizens or even constitute a vector for a more just society.
As discussed in this article, the "outsourced" model of PB in Taiwan has an adverse effect on the further institutionalization of PB.PB as an isolated policy device is easy to replicate, but PB will not live up to its radical promise unless it is embedded in a larger framework of empowered participatory governance or in a multi-level "deliberative system" (Parkinson and Mansbridge, 2012).The ideal of "democratizing democracy" (Santos, 2005) will be attained only when the logic of elite-mass relations gives way to that of solidarity, equaliberty (Balibar, 2014), and self-governance (autogestion; see e.g., Karalis, 2014;Howard, 2000).
Despite all the pitfalls, one reason to be optimistic about PB in Taiwan is that even a "defanged" democratic innovation "might open the Pandora's Box of real citizens' involvement and deep democratization" (Peck and Theodore, 2015, p. 227).I submit that the future of PB in Taiwan depends on two factors.First, it depends on whether the current modus operandi (outsourcing the implementation of PB) will be replaced by genuine democratic reforms of the state apparatus.As argued above, the outsourcing system tends to leave intact or reproduce the existing power relations in civil and political society, since it generates incentives for the commissioner to evade administrative and political responsibilities, and imposes structural constraints on the performance of the contractor regarding mobilization, organizing and deliberative quality.
Second, it depends on whether a growing number of active citizens and civil society organizations (especially progressive social movement organizations) can fill the new political space created by PB that may otherwise be occupied by vested interests and political elites.And this in turn depends on whether PB will remain an external tool that deals mainly with small community grants, or turn into a platform in which a wide range of municipal policies can be discussed, debated, and decided on.
As Rosanvallon (2008, p. 12) remarks, "the power to vote periodically and thus bestow legitimacy to an elected government is almost always accompanied by a wish to exercise a more permanent form of control over the government thus elected."The past decades have witnessed numerous experiments with both participatory and deliberative forms of democracy, of which PB is a pronounced example that purports to establish the link between participation, deliberation, and decision-making.Once the link is consolidated, the ideal of citizen empowerment will be accomplished and the full potential of democracy unleashed.
Like what usually happens in the diffusion of policy innovations, PB now takes a variety of forms, operates on different scales, and has been employed to serve diverse aims that may have little to do with social justice or citizen empowerment.This article thus attempts a constructive criticism of the "outsourced" model of PB in Taiwan, which I find highly unique compared with those well-known cases across the world.In its current form, it is not an integral part of a more comprehensive set of institutional reforms.Furthermore, it has neither threatened dominant interests nor led to a substantial democratization of power relations in both civil and political society.
This practice of "outsourcing participatory democracy" is a conscious choice of the path of least resistance and risk, which in turn results from the lack of active support from the progressive sections of civil society, the skeptical or hostile attitude of political society towards PB, and the logic of contracting out that dominates the public administration.As the evolution of any policy innovation is not predetermined, it remains to be seen to what extent PB in Taiwan will break away from the current modes of practice and reinvigorate its emancipatory potential.

Table 1
Participatory Budgeting in Taiwan: The Major Cases 3The two projects launched by the Ministry of Culture are both thematic (related to community-building) and village-based.