Keywords

Introduction: Participatory Budgeting as Co-production

Participatory Budgeting (PB) is a democratic innovation where citizens participate directly in making collective decisions about how to spend public money. This deceptively simple definition makes PB distinct from other forms of public participation for its focus on political renewal and socio-economic impact. PB’s potential and malleability have turned it into a global phenomenon. In the last 30 years, it has been adopted in thousands of towns, cities and regions in 70 countries, gaining support from governments, civil society and international organisations. There have been an estimated 11,690 cases worldwide, two-thirds in South America and Europe (Dias et al. 2019, 15–16).

PB has only recently permeated the literature on co-production (e.g. Barbera et al. 2016; Brandsen et al. 2018). A co-production lens brings into relief a central aspect of PB, namely, its regenerative work at the intersection between civil society and the state. Bovaird and Loeffler (2012, 1121) define co-production as being about ‘professionals and citizens making better use of each other’s assets, resources and contributions to achieve better outcomes or improved efficiency’. PB illustrates how co-production can generate public value by linking citizen voice and action and by enabling collaboration (and at times productive contestation) between professionals and citizens across communities of place, practice, identity and interest. I use the term ‘citizen’ in an expansive sense, rather than in the restrictive manner of legalistic definitions. By citizen, I mean people who inhabit the various physical or virtual communities where PB takes place. This includes migrants, refugees, children or prisoners, typically excluded in narrow definitions of citizenship. I use the term ‘professional’ also expansively to refer to practitioners working for institutions, networks, organisations and services across sectors.

PB has developed into myriad models and approaches, thus encompassing modes of co-production spanning Bovaird and Loeffler’s (2013, 5) four categories: co-commissioning, co-design, co-delivery and co-assessment. In PB, citizens decide how to allocate a budget to fund policies, services or projects (co-commissioning). But they may also be involved in creating or reconfiguring a service or initiative (co-design), in managing facilities or implementing a service or programme (co-delivery), and in the monitoring and governance of the implementation cycle (co-assessment). PB processes can thus combine several modes of co-production at different stages.

PB spans both ‘user co-production’ and ‘community co-production’ (Loeffler and Bovaird 2016). Sometimes PB’s constituency is the users of a service, and sometimes PB has a broader governance remit, where the entire community shares decision-making power over various policy areas. In any case, PB is a quintessentially collective endeavour, which may help to rebalance, in research and practice, the ‘major gulf between current levels of collective co-production and individual co-production’ (Bovaird et al. 2015, 19).

PB processes are diverse in scope, scale, ambition and impact, and while they can include various modes of co-production, this chapter focusses on co-commissioning because it underpins its transformative potential through the mobilisation of resources. In the more radical versions, PB seeks to tackle health, social, economic and political inequalities, achieve more efficient and responsive governance and produce redistributive decisions (Baiocchi and Ganuza 2017; Cabannes and Lipietz 2018; Wampler et al. 2018). Underpinned by participatory and deliberative ideals (Escobar 2017), the underlying theory of change is that PB processes can:

  • mobilise a range of values, perspectives and knowledges (e.g. local, experiential, emotional, professional, scientific) to work out needs, aspirations, trade-offs and priorities in a given context;

  • develop civic and official capacity to grapple with complexity (e.g. wicked issues), overcome silo-thinking, address urgent problems and enable long-term thinking and

  • contribute to develop a public sphere that promotes solidarity over self-interest, public deliberation rather than just public opinion, and civic education and agency over public apathy.

In doing so, the theory goes, PB enables well-informed decisions for collective action that can transform lives, communities and systems. Increased capacity for effective action should improve legitimacy and trust, which in turn sustain and deepen that capacity, thus generating a virtuous circle for democratic governance. Later sections will examine to what extent, and under which conditions, PB can live up to this theory of change. But these broad ambitions already suggest why PB has inspired so much hope and concern. This type of democratic innovation can challenge political, civic and administrative cultures because—again, in its more radical versions—it entails a reconfiguration of public governance that redefines the roles and relationships of citizens, public servants and politicians.

This chapter provides an overview for new readers, drawing from key sources in the PB literature and my ongoing research and learning with PB communities of inquiry and practice. The next section defines PB and offers a historical account, before moving on to examples of how it works. This is then followed by a review of PB impacts and outcomes, leading to concluding reflections about the future.

What Is Participatory Budgeting? Concept, Origins and Global Diffusion

Democratic innovations are ‘processes or institutions that are new to a policy issue, policy role, or level of governance, and developed to reimagine and deepen the role of citizens in governance processes by increasing opportunities for participation, deliberation and influence’ (Elstub and Escobar 2019, 11). The point of these new processes and institutions is the ‘democratisation of democracy through more flexible, collective and less hierarchical decision-making’ (Sintomer et al. 2016, 6). PB can thus be defined as a democratic innovation where citizens participate directly in making collective decisions to allocate public budgets (e.g. capital investment, service funds, grants). Unpacking this definition highlights what makes PB distinctive:

  • citizens participate directly—PB places citizens (framed as constituents, residents, service users, etc.) at the centre of co-production, in contrast to other forms of participation that prioritise intermediaries (e.g. organised civil society). PB therefore descends from direct forms of participatory democracy, rather than mediated forms of associational democracy (Ganuza and Baiocchi 2019, 78). Civil society organisations still play important roles in PB, but citizens are framed as political agents rather than as mere subjects of intermediation and representation.

  • making collective decisions—Unlike other forms of public engagement, PB participants are not just being consulted, but share in the power to make decisions. PB differs from aggregative models of participation by emphasising deliberative work and collective action. That is, PB is not about aggregating individual demands but about working out public priorities through a collective process of ideation, proposition, development, review, deliberation and vote.

  • toallocate public budgets—Many public engagement processes are symbolic or tokenistic, rather than substantial and consequential. PB’s distinct appeal is that it deals with actual investment, resources and budgets. It goes beyond promises that something may happen and entails commitment to explicitly allocate funds to make something happen.

As shown later, these parameters are unevenly met across the world. The reader may also wonder, is this then a form of direct democracy, of delegated decision-making rather than co-production? Not quite; this is why the co-production lens is useful as it puts the spotlight on the complex set of processes and collaborations developed through PB. PB is a form of co-production because co-commissioning takes place through interaction between citizens, public servants, civil society actors (e.g. activists, third sector professionals) and politicians at various stages of the process. For example, public servants may not vote in the final stage, but they may co-develop priorities and guidance for earlier stages; by the same token, third sector professionals may co-develop proposals and eventually co-deliver the implementation of PB decisions. PB thus offers a unique exemplar of co-production as distributed practice across a complex system of co-commissioning.

The PB definition above distils basic normative and practical elements in an otherwise heterogenous field. PB conceptualisations are varied and contested due to the interplay between ideas, contexts and practices (Cabannes and Lipietz 2018; Dias et al. 2019). To understand this, we must track PB’s origins and diffusion. Accounts of the PB journey abound (e.g. Ganuza and Baiocchi 2019; Sintomer et al. 2016) and they start at the same time and place: Porto Alegre, Brazil, 1989. This large city was the experimental cradle for PB, which soon spread across municipalities in the country. The Workers Party, a social movement then morphing into a governing party, developed PB as part of a broader programme of democratic renewal and socio-economic justice after Brazil’s military dictatorship.

Porto Alegre’s administration was the trailblazer in mobilising citizens, particularly from its poorest communities, to participate in decision-making over public budgets. This generated a complex system of participatory governance that evolved over time, including fiscal and administrative reforms to support these new co-commissioning processes. High levels of local government autonomy and fiscal powers were central in enabling PB to allocate hundreds of millions of dollars (Baiocchi 2005; Wampler 2007).

In this original model (see Abers et al. 2018, 7–11), the PB annual cycle starts in open neighbourhood assemblies where, with support from community organisers, residents identify proposals for investment. Open assemblies also take place at regional level covering the entire city. Their role is to scrutinise the government’s implementation of decisions from the previous year and elect temporary delegates for 16 Regional Budget Forums. These forums work over months to define distributional criteria (e.g. using indexes of socio-economic exclusion), prioritise proposals coming from neighbourhood assemblies and categorise investment programmes by policy area (e.g. infrastructure, education, health). Over time, this geographic approach was complemented by 5 Thematic Budget Forums covering municipal-level policies, in order to foster solidarity across the city on issues affecting all inhabitants (e.g. transport, economic development).

The final layer is the Municipal Budget Council, a deliberative body formed by civil society representatives elected for a year by the regional assemblies. It oversees the PB process, ratifies criteria for regional and thematic distribution and approves allocations to government departments. Public servants and other professionals, including the Mayor’s office and relevant agencies, provide expertise to cost proposals or assess feasibility and legal compliance. The result is an Investment Plan taken forward by the Mayor, as part of the full municipal budget, for scrutiny and ratification by elected politicians at the City Assembly. Implementation then follows, overlapping with annual PB cycles. Initially the focus was on capital investment (i.e. infrastructure), but it eventually extended to co-commissioning public services (Abers et al. 2018).

Creating and sustaining this complex model of participatory governance was fraught with challenges and entailed ongoing development and reform. Nevertheless, PB is credited as a catalyst for urban transformation through development and redistribution in Porto Alegre, particularly in its first decade (de Sousa Santos 1998; Marquetti et al. 2012). I will review outcomes later, but there were six types of impacts as summarised by Abers et al. (2018, 11–12): direct and sustained inclusion of the poor and socially excluded in the governance of the city; breaking down clientelism and patronage through increased transparency; redistributing urban infrastructure and public services to under-served neighbourhoods; building and democratising civil society; developing new public administration capacity and advancing a radical form of democratic governance.

For narrative purposes, PB’s diffusion can be roughly mapped onto three waves: national (1990s), international (2000s) and global (2010 onwards). In the 1990s, Porto Alegre inspired adaptations mainly across Brazil (Wampler 2007). In the second wave, PB spread to Argentina, Ecuador, Peru and Uruguay, alongside early adopters in Europe such as Spain, Italy, Portugal, France and the United Kingdom (Sintomer et al. 2016; Talpin 2011). The ‘tipping point’ was the first World Social Forum, hosted in 2001 in Porto Alegre to galvanise an international civil society counter-movement to the World Economic Forum held that year in Davos (Porto de Oliveira 2017, 129). The World Social Forum was a milestone in PB’s diffusion, particularly for the European trailblazers (Rocke 2014). Thousands of activists, public servants, politicians, researchers and social innovators brought back their learning about the Porto Alegre experience to their countries. New international connections, networks and communities of practice began to form.

In parallel, PB caught the attention of international institutions, development agencies and NGOs, attracted by PB’s potential to increase participation, transparency and accountability in the allocation of resources. For example, UNICEF, the United Nations Development Programme and UN-Women funded PB programmes in South America and Africa (Porto de Oliveira 2017, 137); the World Bank has promoted PB through international development programmes; and the Open Government Partnership champions PB transnationally, with countries like Scotland making it central to its Action Plan (Escobar et al. 2018). Discursive work by these organisations and networks propelled PB as a ‘best practice for good governance’ (Ganuza and Baiocchi 2019).

The Participatory Budgeting World Atlas includes 11,690 cases (Dias et al. 2019), showing global reach in the last decade. National PB legislation was passed in Peru, Dominican Republic, Kenya, Indonesia, Philippines, South Korea and Poland (Wampler et al. 2018, 58). PB now goes beyond neighbourhoods, towns and cities, including also communities of interest and identity (e.g. Glasgow Disability Alliance 2018; Hayduk et al. 2017) as well as decision-making at regional and national level. For example, PB has been scaled-up in the regions of Lazio (Italy), Poitou-Charentes (France), Rio Grande do Sul (Brazil), Podlaskie Voivodeship (Poland) and the Federal District of Mexico, while Portugal is pioneering PB at national level (Allegretti and Copello 2018, 45). Capital cities are becoming key drivers. For instance, Taipei has recently institutionalised PB, training over 1000 public servants (Yu-ze Wan 2018, 225); in 2017, over 100,000 New York residents voted on US$40 million of capital funds (Goldfrank and Landes 2018, 164); and Madrid and Paris are allocating up to €100 million annually (Cabannes 2017, 181).

Various political, economic and social factors set the backdrop for PB’s third wave, including the 2008 financial crash, the acceleration of socio-economic inequalities and the climate emergency. There is a growing gap between ‘the politically rich and the politically poor’ (Dalton 2017). These challenges, alongside the global democratic recession, are opening space for democratic innovations across the world (Escobar and Elstub 2019). It is striking that a radical innovation developed in the Global South by the experimentalist left has being adapted to such diverse contexts. Although progressive actors remain key champions, the PB journey is no longer exclusive to the left and includes centrist and centre-right parties, alongside international agencies and neoliberal institutions.

Critical observers argue that PB started as a radical ‘political strategy’, but has travelled as a ‘neutral political device’ to improve governance and public trust (Ganuza and Baiocchi 2019, 78). In contrast, celebratory accounts note that PB has gone from a few experimental processes to myriad institutionalised programmes as ‘part of a social and political movement in defence of participatory democracy’ (Dias and Julio 2018, 19). Both perspectives may be right: PB has created a pluralistic field of co-production practices where competing approaches coexist, coalesce or collide.

How Does It Work? Stages, Logics and Dimensions in Participatory Budgeting

Despite variations and exceptions, PB usually entails various stages:

  • Ideation and development—participants make proposals, which may be clustered and further co-developed with other participants, civil society professionals or public servants.

  • Review anddeliberation—proposals are assessed and discussed according to pre-established criteria; for example, feasibility, legality, health and safety, environment, equalities, redistribution, funder’s priorities, etc.

  • Decision-making—proposals are refined through scrutiny and deliberation are then decided on, for instance through ranking, voting or consensus-building.

  • Implementation and oversight—successful proposals are implemented, sometimes co-delivered by public authorities, civil society organisations and community groups, with participants sometimes involved in co-assessment and oversight.

These stages highlight key practices but are not linear. For example, deliberation can also take place when people are imagining or developing proposals, or overseeing the quality of implementation. Likewise, decision-making can permeate several stages, for example through the early assessment and filtering of proposals.

PB’s diffusion is a story of translations rather than transferences, which has generated striking variation. Democratic innovations can rarely be transplanted across contexts without adaptation. Translating practices entails not only transposition but also transformation (Freeman 2009). For PB to become viable, legitimate and functional, democratic innovators must engage in a developmental process that intertwines civic, political and administrative capabilities, infrastructures and cultures. PB variations result from the interplay between travelling imaginaries (values, stories, ideas) and local idiosyncrasies (beliefs, traditions, systems).

The spectrum of PB adaptations is vast, but there are models that show family resemblance. For example, the ‘community grant-making’ approach in hundreds of PB processes in the UK (see PB Partners 2016) usually entails the allocation of small community or service funds. In Scotland, early PB ranged from £750 to £200,000, with the average expenditure being £28,400 (Harkins et al. 2016, 4). In England, up to 2011, most processes were in the tens of thousands range, with exceptions like Tower Hamlets in London, which allocated £2.4 million in 2009 based on local authority-led proposals (DCLG 2011, 114).

Community grant-making usually proceeds in three stages. First, there is an open call for proposals by community groups and civil society organisations. Second, the proposals are publicised, alongside invitations to attend events or participate online. Third, there may be community events where people review the projects, or engage in deliberation, before voting or ranking. Sometimes the entire process happens online, such as in Dundee Decides (Scotland), which in 2018 included 11,000 residents aged 11 or above, deciding over £1.2 million of the city’s capital budget through a digital platform (Escobar et al. 2018, 329).

Developing a particular PB model can create path-dependency, but ambitions and arrangements can evolve over time. For example, after a decade of community grant-making, Scotland is transitioning towards mainstream public service budgets and potentially capital investment. This follows the agreement between the Scottish Government and the Convention of Scottish Local Authorities that PB will allocate at least 1% of local government budgets by 2021 (Escobar et al. 2018).

The examples so far, from Brazil’s system of participatory governance to the UK’s community grant-making approach, illustrate wide-ranging variation. This hinders parsimonious typologies, despite notable exceptions (e.g. Sintomer et al. 2016, 47–57) and efforts to offer continental overviews (Dias et al. 2019; Dias 2018). Rather than trying to categorise such a heterogeneous field, this section explores variation through two heuristics: the logics that underpin PB approaches and the dimensions that help to interrogate PB cases. I will focus particularly on illustrating co-production through further examples.

Firstly, following Cabannes and Lipietz (2018), there are three logics that usually underpin PB in practice:

  • Political: PB as an instrument to ‘radically democratise democracy and contribute to the building and deepening of a new polity’ (ibid.: 70). This logic was prominent in places like Seville (Spain), the first large city that adopted PB in Europe, adapting it from Porto Alegre. PB was developed to co-commission infrastructure, services and projects in neighbourhoods and across the city, including under-served communities and spending around €90 million between 2005 and 2009. The process enshrined decisions as binding once voted in open assemblies. The implementation of projects was overseen by commissions elected at the assemblies and working with public servants (ibid.: 72).

  • Good governance: PB as an instrument to ‘establish new societal priorities and construct new relationships between citizens and governments’ and to strengthen links in civil society and communities (ibid.: 70). This logic is illustrated in places like Dondo (Mozambique), which pioneered PB in Africa seeking to rebuild relationships between ‘citizens and the local state in the context of post-war reconstruction through an emphasis on good governance processes’ (ibid.: 72). Between 2007 and 2009, US$2.6 million were invested on improving living conditions through public works and services such as water supply, drainage canals, streetlighting and health centres in the poorest areas. Communities identified issues that could be addressed by local people and those that required co-production with government. This informed the proposed budget and the level of co-delivery by citizens and public servants (ibid.: 73).

  • Technocratic: PB as an instrument to ‘improve financial efficiency and optimise often scarce public resources and service delivery’ (ibid.: 70). This logic drove most of the 274 processes developed in Germany since the early 2000s, particularly influenced by the New Public Management approach. The focus was on cutting costs or improving municipal revenue, thus prioritising managerial solutions over citizen participation (ibid.: 74). For example, the Solingen municipality responded to the Regional Government of Dusseldorf’s request for €45 million in savings by involving residents in prioritising cuts from an online menu prepared by officials. Around 3600 citizens commented and voted, which generated nearly €40 million in savings. Other cities in Germany replicated the approach since 2009 (ibid.: 75).

Clearly, co-production principles and practices permeate the ‘political’ and ‘good governance’ logics, while they are lacking in the ‘technocratic’ approach. These logics are not mutually exclusive and may coexist, albeit uncomfortably, in PB shaped by competing agendas in particular contexts. Beyond these overarching logics, there are eight dimensions that help to analyse a PB process.

Framing

Framing analysis unravels how PB is presented by the convening organisations. What is organised in and out of the discursive frame? For example, participatory democracy and social justice were prominent frames in Porto Alegre, positing PB to tackle socio-economic disadvantage and reconfigure power relationships between citizens and institutions. In contrast, PB in England has often been framed by local authorities and third sector organisations as community engagement to address local issues and build trust (Rocke 2014). This frame has also been prevalent in Scotland until recent discourses intertwined community empowerment, public service reform and tackling inequalities (Escobar et al. 2018; cf. O’Hagan et al. 2019). From another angle, PB may be framed as an ad hoc process dependent on political will, or as a permanent institution legally grounded, as for example in Peru, South Korea and Poland (Dias and Julio 2018, 31). Frames set overall aspirations, expectations and boundaries, thus shaping constraints and affordances for co-production.

Agenda-Setting

The agenda for a PB process usually emerges at the intersection between the priorities of organisers and citizens. For example, a public authority can only open to PB those policies and services under its competence, and some may be excluded because they are statutory or ring-fenced. Likewise, a civil society organisation may seek to co-commission with money ear-marked by external funders. This was the case of the ‘Canny Wi’ Cash’ co-production process convened by the Edinburgh Council of Voluntary Organisations to allocate £56,000 from the programme Reshaping Care for Older People (Harkins and Escobar 2015, 16).

In most PB processes, citizens and community groups make proposals, which increases their capacity to develop initiatives and respond to priorities. It can also foster commitment to PB and motivate people to participate and mobilise others. However, there are cases where the organisers provide a menu of options (e.g. Tower Hamlets), which limits the power of participants to shape PB priorities. It also hinders the creativity, energy and assets mobilised when citizens prepare proposals, and the skills and capacity developed through that process. Nonetheless, this model may ensure that decisions are readily implementable, as the funding body should have already considered feasibility.

There are processes where co-production is actively supported at the agenda-setting stage. For example, in 2016 there were initially 3200 project proposals in Paris, which were filtered according to pre-established criteria by a municipal commission. Then, organisations, groups or residents proposing complementary projects were invited to co-design workshops in collaboration with public servants. This resulted in 219 larger and stronger proposals for public vote (Cabannes 2017, 189–190).

Scope and Scale

PB can be organised geographically (e.g. city, region) or thematically (e.g. environment, housing). Thematic variations still take place in a geographic area, but the budget is for a specific policy or service. For example, in the French region of Poitou-Charentes, 93 high schools take part in co-commissioning including thousands of students, teachers, professional services and parents allocating €10 million annually (Sintomer et al. 2016, 90–93). Some processes focus on a neighbourhood or community of place, as most cases in Scotland so far (Harkins et al. 2016). Others have broader scope and include municipal, regional or national levels. For example, since 2002, half of Portugal’s municipalities carry out PB, and since 2017 the national government has pioneered three country-wide PB processes, for the general population, young people and schools (Dias and Julio 2018, 22, 25).

Multilevel PB has often been inspired by Brazilian experiences, and developed in European municipalities such as Morsang-sur-Orge, Cordoba and Seville (Talpin 2011). PB designed as a multilevel co-commissioning process promotes attention beyond local priorities and towards broader strategic considerations and solidarity. That is, making decisions that consider priorities beyond one’s own patch, thus grappling with trade-offs in allocating limited resources. This goal to overcome parochialism and foster solidarity explains much of PB’s global appeal (Sintomer et al. 2016, 3). In Seville, for example, city-wide cycling infrastructure co-commissioned through PB has particularly benefitted low-income residents, ‘dramatically improving their mobility and access to places of work and education’ (Cabannes and Lipietz 2018, 71).

Inclusion

Participatory processes face multiple barriers to citizen mobilisation because key resources (e.g. time, self-efficacy, wealth, education) are unequally distributed in society. Proclaiming that a process is open to everyone doesn’t mean that it is inclusive. Inclusion requires purposeful efforts to mobilise people by lowering participation barriers across communities of place, interest and identity (Lightbody 2017). PB organisers in Seville worked to include traditionally excluded groups such as low-income urban framers, young people and migrants (Cabannes and Lipietz 2018, 70–71). A much-praised feature of PB in Porto Alegre was that it mobilised some of the poorest and most excluded people in the city, including shanty-town residents (Baiocchi 2005).

In their analysis of European cases, Sintomer et al. (2016, 190) note that gender representation is often better in PB than in traditional political arenas. In Seville for example, childcare facilities were available during PB assemblies, and there were programmes co-commissioned and co-delivered to help women access the labour market (ibid.). Glasgow City Council has recently completed PB pilots focussed on inclusion in both process and outcomes (Harkins 2019). The £1 million initiative included four deprived districts, each prioritising an area of disadvantage: child poverty, work and unemployment, ethnic minority experiences and young people. The process was co-designed and co-delivered by four citizens’ panels in collaboration with public and third sector partners.

Deliberation

A fundamental question in any participatory process is ‘what kind of citizen are citizens invited to be?’ Some processes invite people as spectators and consultees, while others invite them as deliberators and co-producers (Escobar 2017). PB usually entails the latter, but there are various possible designs. There are three approaches to participation in PB, namely aggregative, deliberative or combined. In the aggregative approach, participants express priorities by voting. Ranking is usually the preferred form of voting because it reduces the ‘popularity contest’ effect, where participants vote only for initiatives they know. Having to rank options incentivises learning about alternative proposals and priorities. The aggregative approach is characteristic of PB in community grant-making.

In the deliberative approach, the emphasis is on discussing priorities, options and trade-offs. Participants learn about the issues and initiatives at stake and deliberate with other participants, proponents and stakeholders. This typically entails assemblies, forums or committees, where participants share evidence and reasons to justify their arguments and re-examine their views in the light of perspectives presented by others. Making decisions via deliberation entails building consensus, which can be difficult without a resolution mechanism such as voting—especially when the process involves large populations. Purely deliberative PB processes are thus unusual, with the notable exception of PB in Antwerp (Belgium) where decisions are currently made through consensus-building in sequenced forums (see https://www.demsoc.org/2019/02/28/antwerps-consensus-based-participatory-budget/).

More common is to combine deliberation and voting, so that decisions are not just a reflection of public opinion (aggregative) but of public reasoning (deliberative). This combined approach can be found worldwide, but it’s not universal (Dias 2018; Harkins et al. 2016). Combined approaches enhance the democratic quality of PB by enabling exploration, discovery, learning and scrutiny, which in turn generates more robust, informed and considered decision-making. This guides current work by Fife Council (Scotland), where £22 million in transport services will be co-commissioned through PB designed for deliberative quality (see https://fif.communitychoices.scot/legislation/processes/1/debate). It will comprise four stages: Discover, Dream, Design Together, Decide. In the first two, participants will review current public transport services and gather ideas and proposals through online and offline forums. The third stage will include a ‘deliberative mini-public’ (Escobar and Elstub 2017) to assess, develop and prioritise proposals. The final stage will entail a public vote on service designs, with changes to policy, spending or service delivery built into the mainstream budget cycle.

Facilitation and Support

PB typically mobilises expertise across public authorities, civil society organisations and communities. Depending on scale and governance context, various forms of knowledge must be brought into co-productive relationship: from local, experiential and professional perspectives, to scientific, technical and policy know-how. At its most ambitious, PB also requires expertise to work through, and possibly change, existing systems and cultures. Organising the process takes a varied skill-set, including project management and coordination, PR and communication, process design, IT, community organising and forum facilitation. When coordinators and facilitators belong to a large institution or service they can draw on organisational resources. Sometimes there are teams dedicated solely to PB, as in Seville, where the municipality employed five PB officers, or Porto Alegre, with a large team of official community organisers (Talpin 2011; Baiocchi 2005). In Paris, the permanent PB team includes nine staff based at the Vice-Mayor Office, with access to political, administrative and IT capacity; for example, around 300 civil servants have contributed to the feasibility study of proposals (Cabannes 2017, 184).

In smaller processes, PB may be only a portion of the workload of in-house facilitators, as it is typical in Scottish processes like Kirkcaldy Kanes or Leith Decides, delivered by neighbourhood planning officers and community workers (Harkins et al. 2016). This adds considerable pressure to an already stretched public and third sector workforce, thus undermining the sustainability of PB (O’Hagan et al. 2019). Sometimes capacity is multiplied by working in partnerships between public officials, third sector professionals and community volunteers. Glasgow City Council, for instance, is developing a partnership approach with stakeholder organisations and community networks (Harkins 2019). Often, in-house staff are trained and supported by external facilitators such as PB Partners in the UK (see https://pbpartners.org.uk). There are also cases, like Rome’s Municipio XI, where an independent non-profit organisation has responsibility for impartially facilitating the public forums (Talpin 2011).

The question of how the process is facilitated and supported is crucial for effectiveness, legitimacy and fairness. For example, community organisers play an important role in mobilising people who are usually excluded. Other support may be needed during the development of proposals, so that less organised community groups are not disadvantaged compared to professionalised civil society organisations. Ongoing support may be also needed for people involved in implementing and monitoring successful proposals.

Digitalisation

PB is experimenting extensively with online platforms, mostly complementing but sometimes substituting offline processes. Building effective digital infrastructure can enable large-scale participation at key stages: crowdsourcing ideas, co-developing proposals, deliberating on options, prioritising and voting. In Iceland, since 2012, the Better Reykjavik online platform hosts the My Neighbourhood PB process, which allocates €3 million annually: citizens can make and discuss proposals, the City’s Construction Board analyses feasibility and costs, citizens vote online, and the City implements the projects—with 600 delivered by 2017 (see https://participedia.net/case/4225).

Madrid Decides is the largest digital-only PB process, allocating €30 million for city-wide projects and €70 million across 21 local districts annually (Cabannes 2017, 185). The platform, which hosts other participatory processes besides PB, allows residents to submit, review and vote on policy proposals and community projects over a three-month period. There are access points at 26 Citizen Advice Offices for people without internet. The proposals are reviewed and costed by officials before the voting stage; 45,000 people participated in the first year and there are now over 400,000 registered users (see https://www.involve.org.uk/resources/case-studies/decide-madrid).

Digital infrastructure is particularly important for regional and national PB. In the pioneering case of Portugal, since 2017, people can make proposals online or at face-to-face assemblies. Proposals are assessed by the relevant ministries before public voting through facilities such as ATMs or online (Dias and Julio 2018, 25). In Scotland, by 2018 there had been 50,000 people participating in online PB, including over 5000 young people via Young Scot’s platform to allocate £60,762 to youth projects in North Ayrshire (Escobar et al. 2018, 325).

All the usual caveats about online participation and exclusion apply. Rumbul et al. (2018, 42) warn that the digital migration of processes may contribute to the ‘gentrification of PB’. Combining offline and online approaches seems thus prudent. Paris, the largest PB operation in Europe, combines online and physical voting, placing over 200 ballot boxes across the city—50% of them mobile across streets, public squares, schools and markets (Cabannes 2017, 189).

Institutional Fit

Where PB is a form of participatory governance embedded within institutional arrangements, it often requires administrative and fiscal reforms (Ganuza and Baiocchi 2019, 79). In contrast, where PB is conceived as ad hoc community engagement, the institutional fit is of lesser concern because the process is not seen as core to the democratic system. This is usually the case in community grant-making or PB by third sector organisations.

The institutional fit comprises dimensions such as syncing PB with the overall budgeting cycle for the public authority in question. For example, in Rome’s Municipio XI the PB process starts in January with public assemblies electing delegates for each neighbourhood (Talpin 2011, 54–55). The second phase in February–May entails meetings by working groups to develop proposals across thematic areas. The third phase in June involves voting for priorities through a public assembly in each neighbourhood. Finally, prioritised proposals go to the Budget Office in October for integration in the annual Municipal Budget.

The institutional fit has implications for the role of public servants, experts and politicians, who in some processes review the feasibility of proposals and deliberate with citizens or their delegates. Elected politicians sometimes help to mobilise communities, contribute to deliberative forums or connect PB decisions to institutional procedures. There are, however, frictions between democratic innovations in participatory governance and established institutions of representative democracy. For example, in Brazil and Spain many PB processes ceased due to changes of administration and lack of cross-party support (Dias and Julio 2018, 22; Abers et al. 2018). Party politics can override participatory politics and undermine emerging institutions like PB.

Impacts and Outcomes of Participatory Budgeting

This section illustrates why PB is credited with transforming lives, communities and systems, while also explaining impact disparities worldwide. Effective PB processes are driven by the principles of co-production articulated by Loeffler (2016, 323): citizens and communities have ‘capabilities and resources, not just needs’ and can ‘contribute significantly to services and outcomes’. Moreover, politicians and public servants ‘should work collaboratively with citizens and communities, rather than paternalistically’, and the focus should be on outcomes rather than just on ‘better or cheaper services’ (ibid.). At its best, PB can generate collective capacity to address health, social, economic and political inequalities.

PB funding comes from a variety of sources, including public services, capital investment, partnership budgets, community revenue, philanthropic schemes, supra-national funds or development aid. PB impacts and outcomes must be assessed in proportion to investment; for example, community grant-making is not comparable to mainstream budgets. Money is not everything in PB, but it is a catalyst that provides focus and incentives to mobilise assets and resources by all involved. Although PB cases are multiplying, the amount of investment is actually shrinking (Cabannes and Lipietz 2018; Cabannes 2015). In the early 2000s, some processes were investing up to US$ 400 per capita annually, whereas now even flagship cases like Cascais, Paris or Madrid barely reach US$ 40; investment has decreased from up to 60% of total budgets to less than 10% (Allegretti and Copello 2018, 45).

The emancipatory impetus that inspired earlier models has often being lost in translation. Social justice approaches, for instance through redistributive criteria that prioritise marginalised communities, have been central to PB’s impact (Wampler et al. 2018; Baiocchi and Ganuza 2017). Deploying PB merely as a participatory device, rather than as a form of participatory governance, compromises its transformative potential. Key conditions for substantial PB impact include: public administration support, adequate resources and strong civil society (Wampler et al. 2018, 68). When factors align, PB can enable the co-production of public goods and outcomes:

  • Socio-economic goods. These include public infrastructure, services, policies, programmes and projects in a range of areas from health to education, transport, environment and economic development, generating outcomes in terms of personal and community well-being.

  • Democratic goods. These include civil society goods such as political equality, citizenship development and collective capabilities, as well as institutional goods such as transparency, efficiency and accountability, generating outcomes in terms of public trust, governance capacity and democratic renewal.

This parsimonious distinction recognises that participatory governance is guided by prefigurative politics: it matters both what gets done and how it gets done. As with other forms of co-production, PB has an important developmental dimension: it is not just about getting things done, but about improving capacity to do things better. Of course, these dimensions are interrelated. For example, increasing transparency of public accounts can generate efficiencies that liberate funding for improving services; and increasing public understanding of budgets can enhance critical citizenship that helps to make better-informed decisions.

PB’s global diffusion was partly motivated by the outcomes reported from Porto Alegre. Evaluations of that case are very positive, including multiple impacts resulting directly from co-commissioning by residents, public servants and civil society organisations. The first decade provides the following highlights (de Sousa Santos 1998; World Bank 2003; Baiocchi 2005; Marquetti et al. 2012):

  • Shifting priorities. PB reoriented public investment to the poorest communities and under-served areas in the city. Between 9 and 21% of the city’s capital budget was allocated via PB annually. Transparent criteria were applied to support redistributive justice.

  • Improving services. New housing for poor families increased, from 1,714 families in 1986–1988 to 28,862 in 1992–1995. The amount of schools and nurseries grew from 29 in 1988 to 86 in 2001, which enabled doubling the number of children in schooling. There was also higher expenditure on health, particularly on building health care facilities in the poorest areas.

  • Improving infrastructure. There were 30 km of new pavements in excluded neighbourhoods yearly, sewage systems reached almost full coverage and 98% of residents had running water by 2001, compared to 75% in 1988.

  • Improving governance. Inflated administrative costs were reduced, and so were levels of corruption, which had hindered progress for decades. Enhancing transparency helped to increase tax revenue by 50%. PB also improved relationships between residents and officials, and across administrative levels.

  • Improving citizen participation. The number of participants grew annually, up to 100,000, particularly from the most disadvantaged groups in the city. Women, ethnic minorities, and low income and education citizens tended to be overrepresented when compared to the city’s population.

Beyond Porto Alegre, evidence across Brazil indicates a positive association between the existence of PB and social justice-inspired change, in particular tackling health and social inequalities (Gonçalves 2014). A longitudinal study spanning 20 years across 253 Brazilian cities shows that PB programmes are strongly associated with increases in health care and sanitation spending, decreases in infant mortality and increases in civil society capacity (Touchton and Wampler 2014). This connection strengthens dramatically when PB remains in place over time, enabling sustained investment in social policies (Gonçalves 2014; Touchton and Wampler 2014).

At What Works Scotland we conducted a scoping review of PB evaluations to assess evidence of impact on health and wellbeing worldwide (Campbell et al. 2018). We found both positive and mixed results from studies of political, economic, health and social outcomes. Highlighting the patchy nature of the evidence available, we concluded that only in the case of Brazil it has been possible to establish systematic longitudinal research. A lack of evidence on impact is not necessarily evidence of a lack of impact, and positive assessments of PB abound, but the scarcity of systematic evaluations remains a pressing concern.

Cabannes (2015) reviewed PB in 20 cities worldwide, small and large (e.g. Chengdu, China, 17 million inhabitants), covering 20,000 projects worth US$ 2 billion. Basic services such as water, sanitation, waste management, transport and electricity attracted half of the funding, while the rest went to infrastructure, economic development, amenities and health and education facilities (ibid.: 259–260). In some cities, PB created new social services, helped to increase tax revenue, mobilised community resources, optimised scarce funding and generated matching funds from other tiers of government and international agencies (ibid.: 266). Cabannes concludes that PB has significantly improved ‘basic service provision and management, with projects that are usually cheaper and better maintained because of community control and oversight’ (ibid.: 257).

PB provides incentives to mobilise resources by all involved, which helps to build trust and capacity to act. In Dondo (Mozambique), co-commissioning through PB built confidence in communities and improved their relationship with public authorities and services, thus increasing capacity to collaborate on issues beyond PB, e.g. security, HIV/AIDS (Cabannes and Lipietz 2018, 73–74). In a similar vein, a survey of experiences in Russia found that PB helped to address direct needs, develop governance institutions, increase public trust in local authorities and change residents’ attitude to their role in local development (Shulga and Vagin 2018, 440).

In their assessment of PB across Europe, Sintomer et al. (2016) observe that the overarching focus has been the modernisation of public services. They identify six trends, with varied incidence across cases: improved services, better responsiveness, problem-solving, devolution, joined-up-thinking and transparency (ibid.: 173–174). Despite exceptions, they found limited impact on social justice outcomes, due to factors such as lack of political will to make PB a redistributive instrument, and a focus on social outcomes through other arrangements between state and civil society (ibid.: 185–191). They also noted limited impact on political culture and the ‘democratisation of democracy’, despite successes in strengthening civil society and increasing citizen participation (ibid.: 203–206).

Studies of the UK’s community grant-making model echo these findings. Rocke (2014, e-location 2067) argues that systemic impact has been limited, but there are positive local results: participant satisfaction, improved self-confidence in individuals and organisations, improved intergenerational understanding, increased volunteering and formation of community groups; better awareness of local councillors, increased confidence in service providers, and more community control over some resources. An evaluation of PB in England (DCLG 2011) indicates that PB can attract additional funds into deprived areas by offering an effective means of distributing resources, generating innovative projects and disrupting the status quo of top-down initiatives and services. PB is credited with opening up new communication channels between the public sector and communities, but sustained support depends on the successful co-delivery of projects.

In Scotland, PB is currently framed as a key driver for improving public services, empowering communities and achieving better outcomes for disadvantaged populations (Escobar et al. 2018). However, the evidence so far tempers expectations (Harkins et al. 2016). O’Hagan et al. (2019) argue that PB has been more about transaction and transference than about transformation. That is, much PB activity has focussed on transactional relationships (demand-supply) between service providers and communities, and to some extent transference of decisions and resources, rather than a more transformational shift towards co-producing collective action and social justice outcomes.

Nevertheless, some experiences in Europe offer new hope in that direction. The City of Paris made a strong commitment by earmarking €500 million Euros over 5 years, inspiring what seems to be a particularly creative and solidary process (see Cabannes 2017). In 2015, there were 5000 proposals, of which 1500 were deemed feasible and resulted in 8 Paris-wide projects and 180 at district level. These covered transport and mobility, environment, public spaces, culture and solidarity programmes for vulnerable groups (ibid.: 193). In 2016, the top voted priority by Parisians was ‘Solidarity with the Homeless’, a complex and innovative programme focussed on homeless people and migrants, including various projects to increase access to services and facilities, provide health kits and emergency accommodation and develop new housing solutions (ibid.: 195). All in all, PB is challenging and changing public administration in Paris, as multiple directorates must collaborate among themselves and with various communities of place, practice, interest and identity, in order to co-deliver expediently (ibid.: 201–202).

But for all its promise, PB is not exempt from pitfalls and undesirable impacts. For example, much attention goes to early PB stages (ideation and decision) but its impact and sustainability depend chiefly on the implementation cycle that brings proposals to fruition. A key challenge, observed for instance in Italy, is that instead of becoming a stimulus for better administrative performance, PB can end up ‘dragged by the slowness and inertia of bureaucratic procedures’ while public trust is eroded (Allegretti and Copello 2018, 42). Public administrations are rarely built for participatory governance or co-production and thus implementation problems are found in PB worldwide, including Porto Alegre (Cabannes 2015; Abers et al. 2018).

There is also the risk that politicians hijack the process and, for instance, in Spain and Brazil party politics has undermined PB (Dias 2018; Sintomer et al. 2016). Furthermore, there have been notorious cases of clientelism and corruption in Mexico City (Rumbul et al. 2018, 22–25). It is easy to rekindle cynicism even among pioneers in democratic innovation. This has been the case in Peru, one of few countries that mandated PB for all subnational governments. In the early years, PB was the most successful aspect of decentralisation reforms, increasing citizen participation and directing resources ‘toward improving quality of life in Peru’s poorest neighbourhoods’ (McNulty 2018, 147). Fifteen years later that optimism has disappeared, with most PB becoming a formality lacking robust deliberation, expedient implementation and civic trust (ibid.: 148). There are measures that can help avoid these pitfalls and enhance PB’s effectiveness; for example, following Wampler et al. (2018, 65–70):

  • Social justice requirement; redistributive criteria for the overall funds and subsequent selection of projects, for example using indexes that map under-served communities

  • Incentivising inclusive participation; e.g. removing barriers, introducing positive action

  • Training public servants for technical challenges but also participatory approaches and culture change

  • Peer-to-peer learning between practitioners in public authorities and civil society

  • Strong organised civil society, effective in co-production as well as scrutiny and contestation

  • Binding decision-making rules

  • Transparency and monitoring for expedient implementation

All in all, for PB to transform lives, communities and systems, democratic innovators must consider structural issues which may require fiscal reform, redistributive policies, and profound changes to governance and public services.

The Future of Participatory Budgeting

PB is arguably the most widespread and shape-shifting democratic innovation of the last three decades. Its global spread has been enabled by conceptual and practical malleability, which allowed adaptation worldwide following disparate logics and with varied levels of impact. Its transformative potential has often been tempered by depolitisation, that is, by ‘rendering technical’ what were intended as ‘radical projects of political renewal’ (Newman 2014, 3301). As public authorities drive the institutionalisation of PB, civil society organisations have crucial responsibility as partners but also watchdogs. There is ‘greater likelihood that PB will evolve into a robust institution when citizens more actively demand their rights’ (Wampler et al. 2018, 52).

PB illustrates the emergence of new interfaces between state and civil society where co-production practices are developing. It provides a liminal space for the renegotiation of meanings, values and practices, which can lead to struggles and contestations. This is to be expected, as models of participatory, deliberative and representative democracy try to coexist and coevolve (Escobar 2017). For all its emphasis on collaboration, effective co-production must also accommodate productive contestation, and PB provides a capacious arena. The future of PB depends on overcoming difficulties aligned with those that co-production faces more broadly (Loeffler 2016, 332–333; Bovaird and Loeffler 2012). I conclude by reflecting on six overarching challenges.

Cultural challenges. PB requires reshaping mindsets and ways of working so that participatory governance can take hold. This entails learning and commitment from public authorities, civil society organisations, elected representatives, community groups and citizens. New forms of facilitative leadership are also necessary—i.e. the ability to bring people together across divides to engage in collective problem-solving, deliberative decision-making and creative co-production. In many ways, this amounts to a countercultural movement to change political, civic and administrative systems that go back decades, if not centuries.

Capacity challenges. PB requires skills including process design, communication, knowledge brokering, mediation and facilitation. It takes local and systemic knowledge and the know-how to build trust, negotiate competing agendas and create space for robust deliberation. PB must be supported by properly resourced and trained teams, including community organisers and deliberative practitioners. Inequalities faced at large in society—related to education, disabilities, resources, self-efficacy,responsibilities (work and care), etc.—often prevent people from taking part in participatory processes. The know-how of facilitators (Escobar 2019) is essential to minimise barriers to participation and support inclusion, which underpins the legitimacy and effectiveness of PB. Building digital infrastructure and capabilities is also instrumental to large-scale participation and deliberation.

Political challenges. PB brings a new type of participatory politics that may clash with established dynamics and challenge the status quo of organised interests. It can also clash with electoral politics and it is difficult to build cross-party support to give PB a stable framework for long-term development. There are frictions between innovations in participatory governance and traditional representative institutions. PB is sometimes discontinued due to changes of government and party politics. Elected representatives have a crucial role not just within PB processes but in broader systemic reform and culture change. Political champions have been central in the PB story, but democratic innovators remain a rare breed among politicians.

Legitimacy challenges. Any participatory process runs the risk of tokenism, that is, being deployed for symbolic reasons rather than substantial purposes. In the current financial context, PB may be used for public spending cuts, which can undermine its legitimacy. PB that fails to mobilise substantial resources and commitment will see a distraction from other initiatives, thus losing support from people who want to make a difference. Consequently, PB must be worth people’s time and contribution. The legitimacy and transformative potential of co-commissioning through PB depends greatly on two democratic dimensions: deliberative quality and social justice. Inclusive deliberation enables learning, scrutiny and problem-solving, which generates better-informed decisions, policies and services. In turn, a focus on solidarity and redistribution improves outcomes for citizens and communities.

Sustainability challenges. PB requires sustainable funding, long-term commitment, ongoing learning and adaptation, and sometimes institutional reform. PB can take years to develop and bed in. Paris has illustrated remarkable commitment by earmarking €500 million for 2014–2020, providing precious stability for PB development. Co-commissioning through PB may entail changing arrangements in governance, procurement, budgeting and other administrative systems. At its most ambitious, it may also comprise broader fiscal reform. In the early years of PB, Brazil underwent wide-ranging taxation reform to increase revenues. Porto Alegre had substantial fiscal autonomy, generating almost 60% of its budget (Ganuza and Baiocchi 2012, 5–6). Yet, not even iconic cases like Porto Alegre have guaranteed sustainability. Since 2004, its impact dwindled due to changes in funding and government commitment, and failures in implementing decisions, resulting in steady decline until suspension in 2017 (Abers et al. 2018).

Research challenges. The evidence base on PB is large, but it is skewed towards case studies, features few mixed methods approaches and lacks adequate datasets for longitudinal and systematic comparison worldwide (Campbell et al. 2018). Given the global proliferation of PB, the challenge over the next decade is to investigate the social and democratic goods generated in the medium and long term (e.g. most effective and sustainable approaches, impact on institutions and services, outcomes for citizens and communities). A stronger evidence base is crucial to inform democratic innovation and reform driven by public authorities and civil society, and evaluation should be an integral part of PB processes.

There has been remarkable progress building PB networks of practice and inquiry around the world, including the crowdsourced platform Participedia (https://participedia.net), the LATINNO database (https://www.latinno.net), the OIDP repository (https://oidp.net) and the new Global PB Hub (https://www.participatorybudgeting.org/globalpbhub), as well as resources like the PB World Atlas (Dias et al. 2019). I therefore conclude by inviting interested readers to go beyond this introductory overview. Experimenting with, and learning from, democratic innovations like PB will be central to reimagine the relationship between citizens and authorities. We need traditional institutions to be better and new institutions to be different. At stake is nothing less than setting the foundations for the governance of the future.