Participatory Budgeting: A meta-level review

Participatory Budgeting: A meta-level review

Rebecca Rumbul Alex Parsons Jen Bramley

PB Report

Much has been written about participatory budgeting over the last 30 years. From humble beginnings in Brazil, it has swept over the globe and is considered by many institutions and governments to be an ideal method of tangibly engaging citizens in the operation of their communities. It has, however, developed beyond the original Porto Alegre model, and the evolution, exportation into different cultural landscapes, and digitisation of the model have posed new challenges for implementers, innovators and supporters. This report, conducted primarily for the Hewlett Foundation with the involvement of the Omidyar Network, examines some of those challenges. The research was conducted from a meta-level perspective, seeking not to replicate the many excellent case-studies on individual instances of participatory budgeting, and instead identifying where additional support for participatory budgeting could be targeted to benefit the community of practice as a whole.

This research identified significant challenges in the participatory budgeting sphere, from a very common lack of goals to be achieved through participatory budgeting exercises, to very weak network links and peer support for implementers, to the frustrations of the exercises as a result of political corruption or subversion. The migration to managing participatory budgeting digitally presents the very real risk of the process becoming gentrified, and is just one example of the consequences of scale in participatory budgeting only being achieved at the expense of disenfranchising the most under-represented.

The report makes four key recommendations to target funding into participatory budgeting:

  1. Fund better targeted and comparative research into PB in areas of interest
  2. Fund specific research into citizen trust and attitudes towards governing bodies in areas where PB has failed or been withdrawn.
  3. Establish (either through support of an existing organisation, or creation of a new one) a dedicated PB organising body.
  4. Establish a senior expert PB forum/committee comprised of global PB stakeholders to attempt to build consensus on improving PB implementation and its outcomes through institutional change.

While these recommendations do not deal with the detail of the method of participatory budgeting, they aim to cultivate a better organised, supportive and fruitful landscape for practitioners in which the social and participatory benefits of the model can be fully realised. With targeted funding to support participatory budgeting programmes, positive outcomes can surely be maximised for citizens all over the world.


Participatory budgeting (PB) is almost 30 years old. From its origins as a means of redistributing wealth in Porto Alegre in Brazil it has developed into a global phenomenon, appearing on every continent and is responsible for channelling millions of dollars per year into public projects proposed and chosen by citizens. Participatory budgeting's evolution from small scale innovation for addressing the needs of the impoverished in a city in Brazil, into a global mechanism for citizen engagement and participation has been rapid. While the World Bank has been key in pushing for the adoption of PB in the developing territories in which it works, the European Union and individual cities and regions in the developed world have also embraced PB as a tool to provide citizens with a greater say in how their public funds are spent.

The allure of PB is clear. Governments believe it can be used to increase democratic engagement, to generate public ownership of budgetary decision-making, and to tacitly generate public support for incumbent government. As a condition of aid and development investment, international funding bodies believe it can increase budgetary transparency, reduce corruption and encourage the greater accountability of state to citizen. And for citizens, it could provide a public route to request new or improved facilities or services that may not have been a political priority for governing parties or civil service managers. The potential benefits of PB are therefore significant, but as yet, largely unproven.

The literature examining the operation, evolution and outcomes of PB is surprisingly thin. While a number of case studies have been produced over the last 20 years, with a small number demonstrating a correlation of positive outcomes alongside PB implementation, very few comparative or large N studies have been conducted, and fewer have gone beyond short term evaluative studies to provide evidence of positive outcomes specifically caused by the use of PB. Without a pool of good quality research evidence, it is difficult to assess the merits and risks associated with PB, and problematic to assume that it is an inherently positive mechanism. PB also stands out as an isolated intervention mechanism amongst the wider open governance, transparency and accountability landscape. It rarely features in conversations on these themes at nation state-level, and lacks the global peer support and development network and community that similar interventions and civil society organisations focused on accountability, transparency and open government enjoy. As such, whilst PB is globally practised, the community involved appears to lack the maturity, plurality and professionalism exhibited by similar communities of practice, such as the Open Data community, or the transparency movement.

This report, commissioned by the Hewlett Foundation with the involvement of the Omidyar Network, seeks to identify where there may be opportunities to develop or support the PB community at the meta level. It is not intended that this report be considered a full or exhaustive study or mapping exercise of PB — indeed, such an endeavour would require a vast and lengthy research commitment beyond the scope of this exercise. It does not take a view on whether PB is inherently ‘a good thing', nor does it critique different methodologies of PB. Rather, this research seeks to uncover and understand some of the nuances of the PB world and community; whether there are gaps in funding that if plugged could improve or increase positive outcomes, and the risks and vulnerabilities associated with PB programmes.

Problem statement

Over 1,500 instances of PB have been implemented across five continents as of January 2017. The majority of these instances have been implemented in relative isolation as independent mechanisms unconnected to other PB instances elsewhere in the world, and most are implemented at a sub-state level, either city/municipal level, or regional level. However, this situation is slowly changing. The World Bank is a leader in encouraging PB implementation, and large cities such as Barcelona, Madrid, New York and Paris are increasingly allocating larger portions of their city budgets to PB exercises. Portugal is currently experimenting with PB at a national scale, rolling out a pilot project in which citizens will be able to vote using ATM machines, and Kenya has enshrined PB in its constitution.

This rapid expansion has yielded a mix of concerns across the PB world regarding the cost, effectiveness, and inclusivity of the programmes. While PB programmes targeted at the most impoverished may yield positive results, the cost of running a PB programme for the most disengaged and disadvantaged will be understandably high in order to provide outreach, support and education to participants to enable their engagement. Broader PB programmes conducted at scale or through digital means may attract a high number of participants, but without significant outreach and support may suffer a lack of diversity and ultimately advantage individuals already well placed to benefit from the system. The very aims of PB in many instances are opaque, with implementation viewed as something necessary or positive without clear goals and desired outcomes from the programme beyond engaging citizens in decision-making.

The Hewlett Foundation's Global Development & Population Program and Omidyar Network's Governance & Citizen Engagement Initiative instigated this research because they were interested in whether or not PB may advance their own strategic goals to:

It is through these lenses of increased trust between citizen and government, and improved service delivery, that this study examines PB. Gaps in trust between citizen and government exist across all continents and developed and developing nations, and these gaps fundamentally inhibit the development and delivery of high quality and universal public services.

The purpose of the study is to explore philanthropic grantmaking opportunities to advance the research, impact, efficiency and scale of participatory budgeting. The research focused on four specific areas of opportunity: research, implementation & innovation, institutionalisation, and advocacy, under the following research questions:

  1. What can the existing research literature tell us about the current evidenced outcomes of PB, and where are the gaps in knowledge?
  2. How are changes in trust between citizen and government evidenced?
  3. What are the common issues, risks and challenges associated with PB implementation?
  4. Are there innovative uses of PB that have the potential to increase or improve outcomes?
  5. How is PB institutionalised and what effect does this have on participation and outcomes?
  6. How has PB affected the quality and delivery of services?
  7. How has advocacy shaped the development and implementation of PB?
  8. How have global institutions shaped the PB landscape at the state and meta-level?

In conducting this study we sought to address these sub-questions and used them to shape our discussions with the research participants. The hope invested by enthusiasts in PB derives from its simultaneous use of transparency, participation and accountability. Taken together, these three actions are believed to reduce the gap in trust and promote greater confidence in government through a transparent and participative budgeting process that can yield concrete results which citizens can see with their own eyes. These themes were therefore set at the heart of this study.

Research methods

The research questions posed for this study concerned not only existing evidence of PB outcomes and impacts, but an exploration of the more subtle and nuanced characteristics of PB and its implementation in a variety of different territories. Relevant information was considered to exist in academic and grey literature, and within a range of organisations and institutions responsible for promoting and conducting PB exercises. A mixed methods approach to data collection was developed, which consisted of:

  1. An extensive literature review (conducted and submitted independently by Brian Wampler, Stephanie McNulty and Michael Touchton)
  2. Phone/Skype interview with key stakeholders, practitioners and academics
  3. In-person interviews with key stakeholders, practitioners and academics
  4. Desk-based review and analysis of PB processes and digital tools

The literature review for this report was separately commissioned work conducted by academic PB experts Brian Wampler, Stephanie McNulty and Michael Touchton. This literature review is not included in this report, but covered significant ground concerning this study's four guiding themes of PB (research, implementation & innovation, institutionalisation, and advocacy), and was used to focus the remainder of the study, develop a framework of questions and test assumptions.

Further to the desk-based research, preliminary interviews were conducted with a number of individuals working on pertinent issues concerning PB. Key stakeholders were interviewed informally, generally via Skype, to produce a working knowledge of the sector and identify key areas of interest for this study. These interviews were semi-structured and were used to further inform the study and identify stakeholders and programmes of interest.

The coordinating teams within the Hewlett Foundation and the Omidyar Network expressed an interest in PB globally, but in particular, in developing contexts. The potential for PB to contribute to increases in trust and improvements in service delivery in countries such as Mexico and Kenya was considered of significant interest, and therefore, these two countries were specifically targeted within the research. Interest was also expressed in the use of PB in programmes used to distribute royalties generated by extractive industries, and therefore Mexico and Madagascar were examined for their programmes in this area.

Field visits were scheduled for Nairobi and Mexico City, however the project fieldwork suffered a number of setbacks that meant that interviews had to be rearranged and conducted over Skype. The unforeseen re-run of the Presidential election in Kenya was eventually scheduled for the same time as the proposed fieldwork in Nairobi, and following advice, the field visit was cancelled due to difficulty in accessing the relevant stakeholders during that important political period in Kenya. The field work in Mexico was unfortunately interrupted by the significant earthquake of 19th September 2017, during which time members of the research team were in Mexico City conducting interviews. They were unable to continue on the ground due to research participant unavailability and safety issues navigating the city.

The semi-structured interviews conducted for this research, both in person and via Skype, focused around the four themes and research questions outlined in the previous section. Participants were split into two groups - academics and interested expert stakeholders, and PB practitioners. Questions asked of research participants centred on the following themes:

These themes were open ended and participants were invited to provide information and signposting to other issues that the research team may not have been aware of. The participant responses to these themes form the basis of this report, and its conclusions and recommendations.

Case studies

Mexico and Kenya were identified by the Hewlett Foundation as countries of interest with regards to PB. This section details the project findings relating to how and why PB operates in these countries, examining the motivation, implementation and impact of PB processes. It then explores the nascent development of participatory budgeting programmes related to the use of resources derived from the extractives industry as a potential growth area for future PB programmes.


The new Kenyan constitution (agreed in 2010 and implemented in 2013) enshrines public participation in the devolved budgetary process. That is not to say that PB itself is enshrined in the constitution, merely that there should be some ability for citizens to contribute their views officially on budgetary matters. There is a structural obligation for citizen participation in budgetary matters in Kenya, and several organisations have used this obligation to move into the PB and participatory space. With the new constitution came significant devolution of powers and budgets to the new counties, representing a shift in the methods of governance away from centralised policy-making, and providing a new level of autonomy for county administration (although some powers that had previously belonged to the smaller local authorities were now effectively at a more remote level). Since 2013, the implementation of the new constitutional requirements has been progressing, but cannot be described as fully integrated or operationalised at this point. Many areas are yet to fulfill all of their participatory obligations under the new constitution.

While the requirement for participation at county level has brought opportunities for citizens to have their voices heard on budgetary matters, the constitution does not specify any particular structures or programmes that should be implemented, providing significant flexibility in how each county operates participatory exercises. International organisations such as the World Bank have encouraged the implementation of PB processes in a small number of counties such as West Pokot and Makueni, and other NGOs and community organisations have implemented differing participatory exercises in other areas, such as the International Budget Partnership in Nairobi. For the majority of territorial community groups and NGOs that have moved into the PB space, PB represents a new component to their existing community activism and participation work. As such, smaller groups supporting citizens to participate in budgetary discussions may not necessarily have the capacity, experience or expertise to make those programmes a success. Some areas currently have no supported participatory structures in place. This has created something of a local lottery, in which citizens of one area may be able to participate in a very well-structured and funded participatory process, while residents in another area may have no structured process in which to participate. While this is currently not ideal for citizens, this does potentially create opportunities for experimentation and innovation in participatory methods, where similar counties could be subjected to an RCT if the county authorities were willing to accommodate it.


Currently a number and range of participatory activities relating to budgeting processes are being conducted at county level in Kenya. Some are supported exclusively by the state, others are being supported solely by NGOs, and further activities are being supported by the World Bank and/or in partnership with the other two. The purely PB exercises tend to depart from PB convention elsewhere, in that deliberative and consensus-based methods are generally used to develop and agree the projects to fund, with voting only occurring rarely or in situations in which an impasse has been reached. This method of PB is contested, with one prominent expert participant suggesting that ‘it's not really, in my opinion, PB at all, if there is no vote'. Several participants also voiced concerns about these processes being co-opted by senior figures within the community, with citizens deferring to individuals, rather than supporting projects they truly wanted, and with the process potentially being dominated by men, thus women and minorities being marginalised. Other participants however, claimed that the consensus-based approach improved community buy-in for projects and provided valuable opportunity to debate the comparative merits of projects and their benefits, which may have been lost in a purely vote-based exercise. The constitution requires that women and individuals with a disability should be provided for within these exercises to ensure their voices are heard and valued. There is evidence that some programmes specifically cater to certain demographics to reduce the potential capture of the process by powerful men. There is however, no consistency in the engagement of citizens at this point.

Beyond the ‘pure' PB programmes implemented, are programmes aimed at meaningful citizen participation in a broader sense. The International Budget Partnership delivers some such programmes in select Kenyan counties with the aim of not only engaging citizens to participate, but to do so meaningfully through deliberative means. A large part of this exercise includes budgetary discussion and debate, but does not confine itself to a rolling project ideas and awards cycle, and integrates significant community education and discussion in the programmes. This programme also identifies much of the relevant necessary information for decision-making and advocates for improvements in the quality and availability of this information for citizens. This programme represents one of the best evidenced PB-type programmes in Kenya examined for this project, and demonstrates clear positive outcomes for individuals as a result of the education component, as well as laying the foundations for citizens to advocate for institutional change. Whereas a ‘pure' PB exercise will engage individuals in the short term, this IBP method of wider deliberation and citizen education removes the narrow focus on a project cycle, and stimulates citizens' ongoing interest in budgetary decision-making and engaging with governing institutions. Participants in this research working outside of this project also supported the idea of placing a greater focus on deliberation than participation in order to achieve more meaningful and long-term outcomes.

This variance in the implementation of programmes has resulted in a current unevenness in participation provision, and poses risks, as well as opportunities, to those interested in the outcomes of PB. Such risks include the political co-option of PB programmes, PB exercises being implemented poorly, low-levels of outreach, and project monitoring and evaluation being non-comparable and/or of low quality. Several participants referred to the issue of political co-option as a pernicious consequence of participation programmes, with political leaders in each county responsible for the design and delivery of the programmes. In one county, an interviewee told us the Mayor ‘simply stuffed the committee responsible for the PB exercise with their own loyal supporters, who made decisions unilaterally on what money would be spent on.' Another participant noted that in some areas county officials have failed to set up any real programmes at all, with officials either conducting fraudulent paper exercises or making excuses for the absence of a participatory mechanism. This is particularly evident in more rural areas where individuals are more dispersed and less aware of their rights and opportunities. There was also the concern that token participation in World Bank sponsored PB programs was being used to "open-wash" other aspects of a county's government, where legally required information was not being disclosed.

Barriers and challenges to participatory budgeting

Reasons such as corruption and political co-option were specifically cited as causes of organisations shying away from involving themselves in PB programmes. Community and sub-national NGOs in particular operate with scarce resources, and concentrate their capacity towards endeavours that will potentially bear fruit, and the state-structured participatory mechanisms are only considered fruitful in areas where the politicians are judged to be sympathetic to citizen voices and not likely to influence the process. Where groups and citizens perceive the process to be partisan or guided by the invisible hand of government, there is a general consensus that it is not worth the effort to engage. Citizens are also accustomed to being ignored by government in a range of other situations, such as if they have reported issues to a county council in the past and nothing has been done. As a result of repeated past failures of government, many individuals and groups wonder why PB processes would be any different.

Outside of urban areas, there is generally a lower awareness of what PB actually is, how it can be used and indeed, how to actually do it. As such, many community groups are not engaging with PB at this stage because they lack the knowledge and expertise. If or when they do choose to engage, the variety of structures available to them may potentially cause confusion, and there are few obvious and legitimate sources of information in Kenya to approach, unless groups are familiar with the development community. Groups unacquainted with the development community face immediate barriers in the language of PB and its delivery, especially those groups and individuals with lower levels of basic skills. The mere term ‘budget' can deter individuals with low levels of confidence in their numerical abilities, and without adequate outreach, any exercises eventually employed may be vulnerable to co-option, whether that be by politicians or by individuals with a higher level of education. This is not solely a rural issue, and one participant noted that the expertise required to make good, sound decisions in a rural area will be different to that in an urban area. Where projects such as clean water and simple infrastructure are easy to identify as important projects in rural areas, the needs of communities in urban areas are less obvious, and this participant stressed that many people do not know what might be best for them in terms of economic or social development in an urban area. In such situations he said he would rather have experts come and talk to people who were deliberating, as there were limits to their imagination and expertise in complex urban planning, and consulting an expert may provide them with better ideas and a better understanding of what might be best. This also becomes an issue where suggested projects are not reviewed by technical experts prior to agreement, so projects that were agreed on may eventually be dropped because they are not practical. As such, groups in Kenya that could potentially benefit from or assist in delivering PB may not, due to a lack of information or assistance across a variety of areas.

Less formal methods of voting are also vulnerable to subversion, for example show-of-hands meetings are recognised as a form of voting that is vulnerable to pressure and subversion in Kenyan PB. When popular participation is low, it is easier for the room to be stacked to the advantage of partial interests. Sheely notes that it's not necessarily easy to fix this simply by increasing participation, as "if mobilisation is successful in increasing participation in planning meetings, it may also cause elites to modify the tactics they use to maintain influence over participatory institutions".1 It was pointed out that even a secret ballot does not prevent collective punishment when voting for projects in small areas. If an area did not support an agreed project, it would be clear in the result, and as such it is clear that the locality of PB projects can work against the effectiveness of a secret ballot when the size of electorate is small. One participant noted in Kenya that collective retribution was a real risk for some citizens as a result of their decision-making. In one case, ‘punishment' had been experienced in a particular area because that area had not chosen projects that the local politicians had made clear they wanted. This punishment was allegedly the deliberate neglect of local services and infrastructure by local politicians in the months following the project decision, and demonstrates one of the more brazen consequences of participation in PB processes that have been structurally compromised.

While being generally positive about the current reception of PB, Fahamu, a pan-African organisation that collaborates with movements to tackle social injustices, points to significant challenges remaining, especially the limited amount of information available to citizens on ballot allocations. The county governments themselves are generally slow to absorb and spend funds (in the 2013/2014 cycle Kajiado County spent only 4% of their budget) and the exact workings of the devolution process remains unclear to county governments and citizens, leaving ambiguity as to where responsibility lives.2 The issue of lack of access to information was brought up by interviewees in terms of other means of public participation, where discussions of budgets were complicated by the way budgets were presented and released: as very long documents only made available shortly beforehand.

Measuring success

The measurement of success or the recording of tangible outcomes arising out of the new PB programmes is currently patchy. While good examples of M&E and impact research do exist, notably from MAVC sponsored research and the IBP, as well as from individual academic projects, these studies stand out amongst a landscape of mostly anecdotal evidence from active project participants. There has been no real reflection in setting up this new wave of PB why the previous wave of PB in Kenya failed (with the suggestion from one interviewee being that the previous round of PB had been too tightly modelled on the Porto Alegre example without adapting to local context).

In general, there is little research produced by practitioners themselves on the success of the various aspects of the PB programme, as many felt that proof of success would be evident in the execution of chosen projects. While it is disappointing from a researcher's point of view that there is currently no larger ambition to examine issues such as influence in the deliberation process, origins of project ideas, budget allocations and expenditures for awarded projects etc, it is positive that participants appear engaged in citizen-monitoring the eventual production of the chosen projects.

In particular, for projects in Embu County, citizen monitoring committees were formed to track the progress of projects and reflect a sense of social ownership of the results of the PB process. Projects currently being awarded through PB processes are generally very publicly visible at this point, with projects concerning the repair or development of clean water and sanitation or the construction of community assets being chosen. The promise of these facilities encourages citizens to keep themselves informed of their progress. At the most basic level, it is likely that fulfilled projects will encourage future participation and build trust between citizen and institution, and that this will provide a foundation for more detailed and academic investigation into the operation of PB in Kenya.


Mexico was identified by the Hewlett Foundation as a country of interest with regards to PB. This section details the project findings relating to how and why conventional PB operates in Mexico specifically, examining the motivation, implementation and impact of PB processes. PB processes in Mexico related to extractives revenue are explored in the following section.

Mexico City is the twelfth largest city in the world, and one of the largest instances of PB globally, both in terms of absolute funds and population affected. While participation levels are low (averaging at a few percent), this still represents several hundred thousand people participating in the process. Although technically there are entry points for the involvement of citizens and civic organisations, the structure and incentives of stakeholders in the PB process in Mexico create a culture that limits the scope of participation.

Development and precursors

The administrative structure of the city was reformed in the in the mid-1990s, resulting in the direct election of the mayor and the legislature. This period also saw the introduction of more direct methods of citizen participation, including limited use of PB. These tended to share issues of low participation and association with particular political projects that limited their effectiveness as a tool of general public participation.

In the late 1990s, the PLC party set up a system allowing for election to neighbourhood committees and plebiscites. In the 1999 neighbourhood elections, 9.5% of registered voters participated and 10 percent of these had invalid ballots.3 Political parties were banned from presenting platforms for neighbourhood committees, but this was widely ignored. This was seen to damage the process, as the loss of party-label shortcuts led to confusion (preventing straightforward participation), and the process was partially discredited through obvious circumventions of this rule.4 Use of plebiscites similarly led to a low turnout. In a 2002 plebiscite on a plan to extend city highways, there was only 6.6% turnout — far below the ? of all registered votes required for the result to be binding. There was a shift towards more informal methods of participation after 2000 such as "consultations by telephone (consultas telefónicas) and neighborhood assemblies (asambleas vecinales)".5 These again were seen as being partisan vehicles, where the agendas for consultation were set by the mayor and where participation was generally low (from <1% to 10%). The then Mayor, Andrés Manuel López Obrador, argued that this was acceptable, as even if a decision was taken by more than one person, it would theoretically be improved. There were general complaints that participants weren't represented, and this was validated by a comparison of surveys, consultations by telephone and a plebiscite on a single project, where the telephone consultations were found to be far more in support of the project. Obrador conducted a consultation on whether he should remain mayor or resign, with a 95% approval rate (which, although he was popular, was in excess of what could have been expected.)6

The assemblies called by neighbourhood committees were later integrated into the mayor's office as part of the social entitlements programme Programa Integrado Territorial de Desarrollo Social (PIT) and their operation was centralised through this. There were two rounds of assemblies per year. In the first, the priorities for the programs are set out, in the second, the citizens are informed of how the resources have been allocated. These assemblies were again seen as partisan, as a high number (?) of attendees were people who benefited from the program. This high proportion was at least partially believed to be because of "fear that failure to attend will lead to the revocation of the PIT entitlement card" and loss of access to the benefits of the programme. Harbers suggesting that this means citizens view PIT as "a favor that requires some kind of repayment rather than their right as citizens".7 This is part of the dilemma when evaluating the success of citizen participation projects: participation in projects associated with particular political projects may well reflect a client-like relationship rather than deeper engagement.

Participatory budgeting in Mexico City

The idea of PB had existed in Mexico for some time, the "broadly representative" Municipal Development Planning Councils to make investment decisions since 1983; however in practice these powers were rarely used.8

In Mexico City, there were schemes run in Cuahtemoc and Tlaplan in the early 2000s, but PB in Mexico City in its current form starts in 2010 with the passage of the Citizen Participation Law of the Federal District (LPCDF). This mandates a process that each of the 16 local authorities will run for neighbourhoods in their areas, with the voting administered by the IEDF. This was amended in 2012 to introduce a mandate for 3% of each local authority's budget to be spent on PB.9

Typical projects include street lighting, railings around public spaces or adding a public noticeboard. Most projects relate to parks and open spaces, with others including a health centre and several long term projects that are completed in phases (with new additions budgeted and added each year). The IEDF holds a competition to encourage innovative projects, and some recent projects noted through this were the refurbishment of a library with the addition of cooking facilities; acquiring furniture for a retirement home; and several clean water projects.10

Even objectively speaking, 3% of all the respective city budgets adds up to a large pool of money. In 2016 this was $ 874,920,194 pesos or around US$ 47 million11. However, because Mexico City is also large geographically and contains a very dense population, the budgets often translate into quite a small amount of money when divided out. From the point of view of the authorities, for whom much of their budgets are tied up in salaries, 3% represents a much larger proportion of the remainder, and is therefore a substantial imposition on the actual services budget. The money spent on PB is (depending on who you ask) both too small and too big, creating difficult expectations for the process. Several interviewees expressed sentiments about PB being a small commitment of the local authorities, showing that the full budget distribution is either not shared openly or not understood by citizens — hence failures of the process being all the more corrosive of public trust in the capacity/willingness of authorities to manage even such, seemingly to citizens, relatively minor programmes.

The distribution of the funds over the city is also uneven, as money is divided equally between neighbourhoods rather than per capita. While larger areas have larger budgets, they might also have more component neighbourhoods relative to their population, resulting in more areas with smaller budgets. This system makes no allowances for the circumstances or demographics of these neighbourhoods in a semi-random distribution, often to the disadvantage of the poorer citizens.

Similarly, as previously mentioned, the funds are often restricted to certain types of project. These fall broadly into the following categories:

  1. Works and services
  2. Equipment
  3. Urban infrastructure
  4. Crime prevention
  5. Recreation activities
  6. Sports activates
  7. Cultural activities12

However, the categories are inconsistently applied to project types (street lighting is variably urban infrastructure or crime prevention for example). Wiemann calls the combination of these factors "pre-structuring" — substantial decisions on what kind of projects in what area are effectively possible are decided by the rules of process before any projects have even been proposed.13

Similar problems applied to the PB programme in the state of Oaxaca, where projects were less successful in "Oaxaca's smaller, more remote villages", which "turned out to have fewer significant projects and more project failures than the municipal centers. In terms of observed impact, three-quarters of projects in town centers were successful (74 percent), in contrast to a 50 percent success rate in outlying villages." For an explanation of this, "[i]mbalances in project budgets appear to have been quite relevant; in 1992 average projects in town centers received about three times the average amount for projects in outlying areas."14

As a function of the total budget, there is substantial year-on-year variation in the amount of money dedicated to PB, further impacting which projects are and are not possible in various locations.15 As only one project can win, there is also no incentive to bid for less than the total amount, making long term planning more difficult and sustained involvement in the process more complex.


Participation in PB is generally low in Mexico City. On average the participation from citizens in the scheme has been under 3% of the city population except for two cases: once in 2013 where participation reached 12% of the population; and once again in 2015 where it was around 5% of the population. We also heard that when PB projects are held at the same time as the local neighbourhood council elections, there is a corresponding increase in turnout.

While the aggregate number of people participating is not small, those involved are all taking part in separate votes at the neighbourhood level – so local engagement is relatively restricted. A citizen group we talked to won their project with 40 out of 90 votes in a potential electorate of 9,000. This same group also had a very low turnout at a community event to inform the public about the projects. Several interviewees talked about the arbitrary nature of neighbourhoods leading to situations where wealthy gated communities and slums could co-exist in the same PB programme while having very different needs, and another pointed out that classism is a factor, in that mixed settings made participation from different groups less likely.

In addition to low participation, Callas identified three additional problems in the current PB process: lack of co-ordination between various bodies leading to impractical or incomplete projects, lack of planning by the local authorities, and a lack of planning to reconcile actions at lower levels with priorities higher up. There is a general problem in the lack of experience and training for citizens in designing projects, and while there is expert feasibility testing, this is not detecting all future problems. Callus gives examples of wheelchair ramps constructed that are too small or obstructed, or where a project adding security cameras does not account for the additional staff costs associated with making them effective.16

Corruption and subversion of the participatory budgeting process

In Mexico City it was evident that there are issues with the project proposal process. In theory, the process is for people to submit projects on a form, after which expert councils — made up of academics and other such technical experts — gathered by each local authority determine which projects can proceed to a vote. In reality, many projects are ‘citizen-washed', that is, projects are proposed by the local authority itself, using a citizen as an intermediary. This completely shortcuts any benefit of local knowledge in the citizen sourcing of proposals and provides citizens with a choice of projects that are likely to have been implemented by the authority without the PB process. In cases such as this, the local authority can remain in effective control of its budget if the choice of projects is controlled.

Even the perception of corruption is enough to damage the civic capital-building aspects of PB. We were told that a project to deal with infestation in local trees (which was in itself recognised as a good thing) was viewed suspiciously because of a rumour — it was stressed that the truth of the matter was unclear — that a friend of the government was providing the chemical required. The truth is almost incidental to the fact that even projects generally supported by the community were marred by suspicion of malpractice over time. A citizen group noted that they were frequently accused of being given money or apartments when talking to people about their projects. The mere perception of corruption undermines the development of civic capital with the assumption that money or favours are trading hands behind every positive initiative.

Another concern raised in Mexico City was that the written formalised process (while providing large scope for transparency — all proposals are released online) might disadvantage projects proposed by illiterate or less literate citizens. In poorer areas of the city, this might especially disadvantage citizen proposals against those being citizen-washed by the council. Similarly, in Solo, Indonesia, Grillos found that the poorest subunits were less likely to submit proposals overall.17 In circumstances where the first round of the proposal is written, proposals might have to be actively solicited and advice given to ensure even participation. Without strong citizen engagement at the start and neutral but comprehensive support from programme implementers, it is easier for the process to be captured by technically superior official projects.

As participation is low, effort required to affect the outcome is relatively minimal. Specifically mentioned was the problem of vote buying, where hundreds of street vendors might be registered in a single building by organised crime groups for the purposes of generating voting cards in an area. For electronic voting in Mexico City, initially all that was required to vote was information on the voting card. We were told of incidents where authorities or third parties asked for information from residents to register them, but then discovered when they came to vote that their vote had already been cast. There was a substantial problem with online fraudulent votes in 2015, with the security additions in future rounds reducing the number of online votes from around 100,000 to 5,000. This should act as a note of caution in looking at instances of high rates of participation as an unambiguously good thing, as this could just signify very successful vote buying (for instance Vicente's study in Sao Tome and Principe, finding that an experiment to reduce vote buying also reduced turnout18).

Turid studied an indigenous village in Mexico City, in part to reconcile low trust in elections with a high rate of participation and pointed out a possible counter-initiative relationship where perceptions of fraud may encourage people to use their vote (to prevent the vote being stolen by illegitimate votes).19 Corruption and participation are not necessarily opposites, but low trust should make us suspicious of drawing positive conclusions from high participation statistics.

Once a project is decided on, there remains the process of converting it from idea to reality. Without sufficient monitoring, there is scope for funds to go missing or for the project to fail to materialise altogether. Plata argues that one of the deficits of the Mexico City system is "the lack of clarity about the institutional mechanisms for citizens to participate in the implementation and evaluation" of the projects.20 After the vote, there is no formal role for the submitters of the project. It is essentially in the hands of local authorities to arrange contracts and complete.

Only one project can win in each district, and so there is no benefit to a project not claiming it costs as much as the total project budget available to the neighbourhood. This means that the choice of contractor can serve as a venue for corruption where contracts may be obtained through bribes rather than the quality of the bid. As an ambiguous point, the local group we talked to complained about an incorrect (and cheaper) construction technique being used in a successful sidewalk repair project a few years earlier. Corruption isn't required for people to cut corners on construction projects, but in the context of amorphous projects like road repair, there is significant room for money to go missing. Examples such as this would only comprise corruption as part of the transaction cost of completing a project, however in other cases, the local authority has reported that the project is complete when in fact it wasn't even started. The interviewees mentioned that costs for projects would often differ from neighbourhood to neighbourhood for the same project, and the justification would be the use of a different contractor or suppliers. This leads to the suspicion that the extra money might be going elsewhere.

Project monitoring post PB process is an integral and often forgotten part of the PB cycle, and an important one in retaining public confidence and trust in the programme. In Mexico City the IEDF have built a website that encourages social media reporting of project statuses so that there is some room for popular monitoring of the construction process21, and have also started to work with Lab Para La Ciudad to create other online monitoring methods for the process.

The utility of this tool in Mexico City is limited by the fact that IEDF has no legal powers of enforcement, but it has partnered with other offices of the government of Mexico City to attempt to hold corrupt officials to account. Where budgets or projects were changed (for instance if a local authority says they have constructed a computer center in a school, but it was never built), files are passed over to other offices and around 40-50 people in the local authorities have been sanctioned or removed from office as a result of this process. In this case, the customer of transparency is not specifically the public (although that is a goal), but the wider government administration, creating a useful pipeline to hold officials to account through government partnerships.

Civic involvement

Given the problems of participation and corruption (and the corrosive effect of the suspicion of corruption), there are limited incentives for CSO actors to involve themselves in the PB process. While there are technically entry points for the involvement of citizens and civic organisations, the structure and incentives of actors involved create a culture that limits the scope of participation.

While the process certainly generates civic activity, there is significant overlap with clientelistic relationships, where "citizen-washing" of projects makes it difficult to determine when projects can be attributed to participation or the local authority. And while there are genuine groups engaged with the process, the extent of this appears to be limited, with CSOs generally unengaged in the process, preferring to focus on other aspects of participation or budgetary activity where they feel they can have most impact (for instance, the national budget) or where their expertise is better suited.

There is difficulty in engaging in a process where political actors play such a large (but concealed) role. The structure of participatory budgeting, with responsibility split between the sixteen local authorities and the Electoral Institute (IEDF), creates lines of responsibility that make the correct place to engage unclear to civil society actors. Several different interviewees attributed roles to the IEDF that belong to the local councils, matching complaints from the IEDF of fielding complaints about rejections of projects (which is the responsibility of the authority).

The two most significant areas for potential interventions are the pre-process in providing training or resources to citizen projects (especially to overcome literacy issues) and in the post-vote monitoring process.

In the former, the group Wikipolitica provides some assistance to people proposing projects related to road safety and mobility in the urban environment. In the latter, scope for involvement is more limited as the path of appeal is limited. Currently local authorities are held to account via the IEDF making use of agreements with other areas of the city government to investigate and punish corruption in the PB process as they have no direct powers in this area. While her research pre-dates this approach and so doesn't pass judgement directly on its success, Sánchez recommends the law related to participatory budgeting be amended to add explicit penalties on officials for corruption or non-compliance in the process. In the absence of these, evidence gathering by third-parties may have limited utility.22

Extractives-related participatory budgeting

PB programmes focused exclusively on the distribution of income originating from extractive activities are currently rare. Madagascar represents the most established programme, having had several cycles of project development, voting and award in the eleven communities in which the programme has been held (in the south of the country where mining is conducted) since 2004. However, the programmes have not been consistently sustained, and only two communities have programmes operating at the time of writing. Just as the PB processes in Madagascar have undergone evolution, so too the projects funded by PB have evolved. Initially, the projects were focused on infrastructure and capital projects. After two to three years, citizens were able to broaden the project eligibility criteria to encompass scholarship, economic development and agricultural development projects. Madagascar is interesting because the state does not have an integral centralised role in the transfer of funds between the mining companies and the communities. In one sense this could reduce corruption through minimising the number of ‘hands' the funds have to pass through before being awarded to projects; however, it also means that the monetary transfers are inconsistent and subject to the operations of the mining companies involved. Currently several communities that have experienced a significant time gap in payments are unhappy with the situation, but have no formal method of recourse.

Mexico has recently (2014) established a mining fund into which royalties from extraction are funneled, with the goal of making the flow of monies more transparent for citizens, and to ensure that states affected by extraction are the ones that benefit from the monies generated from those activities. The fund receives approximately $150 million per year, although this fluctuates due to variations in extraction volumes and market values. The World Bank has supported a pilot project to establish a PB programme exclusively for the distribution of extractives royalties in one area (Cananea in Sonora), and this area has received approximately $8.5 million to conduct the programme. This programme only began in early 2017, and at present is in the project voting phase. It uses some digital tools including Empatia, but individuals have to attend in person to cast their vote, and the programme has invested in a civic education element to ensure that individuals understand the rules concerning eligibility of projects and how to vote. This programme is more bureaucratic than the one operating in Madagascar, and relies on a regional development committee to approve the projects for the programme and to develop the project ideas into viable activities. The committee contains representatives from the state level and the extractives companies, as well as indigenous communities and local politicians, and theoretically has the power to veto projects that won the public vote. The number of citizens engaging in this first cycle is approximately 2,500 out of a population of 35,000.

A key issue with the use of PB for extractives revenues is the inconsistency of funding amounts and of the timing of those receipts. While the fund in Mexico is large enough to sustain regular cycles of small PB exercises, this was not the case in Madagascar, and interruptions in the operation of these kinds of civic engagement programmes can ultimately cause a certain amount of disengagement and disenchantment with the whole process. The dilemma of who and how to manage the money also has no easy answers. In countries in which corruption is endemic, it may appear a better idea to ensure monies are kept outside of government influence. However, it was evident in Madagascar that this caused a significant power imbalance between communities and companies, with no legal routes for communities to address lack of payment. Having a PB process exist outside of government influence is positive in reducing the politicisation of the fund and empowering non-political community actors. Given the rarity of these programmes at this point, there is no easy solution to this issue.

It should also be noted that PB is not the only policy that helps remove corrupt governments from the flow of extractives money. Standing recommends direct cash transfers for funds derived from Ghana's extractives industries as an alternative to social accountability projects23.

These kinds of programme are arguably more transparent in terms of citizens being able to detect fraud (the money either arrives or it doesn't) and has fewer costs associated with the process itself. It would however, avoid the deliberative and civic capital aspects of a PB programme and may be less effective when the funds available per person are lower. Just as the form of PB should follow from the circumstances of the area, there may be instances where the benefits sought from a PB process may be accomplished through an entirely different kind of programme.

Thematic findings

This section reviews some of the key themes that emerged from this study on PB. It addresses specific weaknesses and vulnerabilities of the PB system, identifies gaps in knowledge and practice, considers interesting thoughts and arguments concerning PB, and discusses specific aspects of PB, such as the use of digital technology and its use in extractives contexts.

1. A lack of clear vision

The implementation of new civic programmes can usually be traced back to an identified problem and a need for change. Severe public health issues arising from the use of dirty water can be addressed by investing in clean water infrastructure; inefficiencies in public bureaucracy can be reduced through better administrative management; and increases in skilled economic activity can be addressed through better educational provision. The implementations of PB examined for this study however, rarely followed this logic. While the original success story of PB in Porto Alegre was developed and implemented in response to specific issues, this element of the process has fallen away in the transfer of PB to other areas. A common theme detected amongst research participants in this study was an absence of specific identified issues or problems to be addressed, and a lack of desired goals and outcomes attached to the decision to implement PB mechanisms. In almost every instance of PB examined, PB implementers were unable to identify what social, attitudinal or structural outcomes/changes they would wish to see in citizens as a direct result of implementing PB. No participant who had implemented PB had decided to use that particular mechanism in order to achieve a specific goal, and because these implementers had not started their decision-making on PB from a position of identifying a need or desiring a specific outcome, no other mechanisms to achieve the same outcome had been considered. One participant noted that PB implementations had drifted away from the original principles of social justice, and had become tokenistic watered-down items in the ‘participation toolbox'. PB therefore often emerged within this research as a solution seeking a problem.

Variance in the understanding of the role of PB hasn't appeared to be an obstacle to adoption. The Public Policy Institute for Wales talk about the key question for PB being "What are the public being asked to do and why?"24 - and this question is answered in different ways in different places (and sometimes not asked at all). The general case for the global vision of PB is that "transparency, democratisation, and official accountability came to be seen as the most appropriate means to reduce corruption and ensure greater efficiency and responsiveness in the use of aid and other economic resources".25 However, other reformers "adopted participatory budgeting as a way to advance other objectives such as civic education and popular legitimation."26 PB is sufficiently broad a policy that it can be promoted by very different groups for very different motivations, which goes some way to explaining how it can be apparently successful while there is such great disagreement as to what it is for. PB can easily be ‘all things to all people', but lack of clarity and vision in achieving specific goals will inevitably inhibit examinations of its impact.

2. Poor quality of impact / evaluation research

The lack of initial vision or desire for specific PB outcomes across the PB spectrum is problematic for the study of PB, in particular in conducting explorations of changes in attitudes, behaviours and socio-economic well-being, and in determining wider and enduring impact. No PB exercises examined in this study had included any baseline research prior to implementation, and frameworks for monitoring and evaluation were generally simple, developed by the implementing organisation and tending towards confirmation bias. Within the academic literature, as described by Wampler, McNulty and Touchton in their submission, there are few comprehensive bodies of work that are able to identify consistent, comparative or concrete evidence of impact from PB implementation. There are studies showing a distinct effect of PB, for example, in a cause and effect relationship between PB and increased spending on health issues that positively impact local health. However, studies that go further to argue that there is a special efficiency bonus (where the same increase in funding to health potentially leads to superior health results when paired with PB) are less convincing. The majority of studies available use case-study methodology, and while these projects provide rich insights into one particular context, they cannot be used to understand the wider impacts, commonalities or opportunities of PB. As Goldfrank argues that there are a number of factors that make a country more or less suitable for PB, comparative studies on large numbers of municipalities in Brazil (judged to have a number of good PB pre-conditions) may not travel well to other contexts.27

An additional complication for researchers is that PB is not duplicated identically in each of its implementations, and therefore the variables involved in comparison of PB programmes prohibit effective analysis. PB reveals itself as a very slippery concept to examine, because in each context it has been adapted to local legal, political, social and cultural frameworks. Even in developed contexts, PB programmes differ significantly in their reach, their structure and their methods of engagement, and there is disagreement amongst practitioners about the parameters and definitions of PB, with many suggesting that voting is an absolute requirement of the PB process, while others incorporate deliberation and consensus-building instead of voting. In addition, only a small number of studies on PB have examined it as a global phenomenon or examined it at a more meta level to consider the global PB community itself, the influence of funders and development agents, and the proliferation of the mechanism across very different social and political contexts.

Whilst the current lack of research is disappointing, there is definite potential for distinct pieces of comparative research to be conducted. It would be possible within countries such as Indonesia and Kenya, where individual PB programmes are operating at local level as mandated by a legal or constitutional requirement, to conduct in-country comparisons of differing PB methods. Large-scale PB programmes operating under the OGP sub-national pilot could also provide valuable comparative insights. With regard to case-by-case impacts and outcomes, it would be possible to conduct large scale N studies and randomised control trial experiments with willing municipalities to examine the real impacts of PB processes on citizens; however, this would not overcome the problem of case studies examining differing circumstances being difficult to compare.

3. Risks of top-down prescribed participatory budgeting programmes

In discussions concerning the motivation of public institutions to implement PB, the majority of participants pointed either to commonly held assumptions about PB (that it is an inherently ‘good thing' to involve citizens to participate in decisions), or to coercion (a higher authority required its implementation for political or development-linked reasons). In both cases, the call to implement PB came from positions of power. Historically, PB programmes were fundamentally redistributive, and aimed at providing more for the poorest areas. This principle has not always endured. Current PB programmes do not necessarily redistribute, and where they do, that redistribution may be as a result of decisions taken internally by those running the exercise, rather than by citizens. One participant noted that:

"the biggest redistribution is often before the exercise begins – when the formula for deciding which areas get funding is decided."

The design and development of PB programmes in the majority of places is conducted by civil servants within government (although many civil society organisations also lead on this, such as those operating in Kenya), and the first contact that citizens have with the programme is when they are invited to participate in a decision-making process. At this point of first citizen engagement, the conditions for participation, decision-making and expenditure have already been determined by the state. As such, inequality, inefficiency or inadequacy in the mechanisms may already have been ‘baked in', and while citizens may go on to participate effectively, the benefits may be minimised due to pre-existing faults with the mechanism. This can be seen in areas of Mexico City, where decisions on budget allocations for each area are done inside government based on political boundaries, and are therefore blind to the separate needs of those living in extreme poverty in what are considered to be generally affluent areas. Similarly an interviewee told us that Kajiado County in Kenya's PB programme gives equal funding to each of its 25 wards regardless of relative prosperity.

Within this study, and contrary to much of the marketing of PB as a bottom-up approach, PB very much emerged as a top-down process imposed upon citizens, rather than designed and implemented in partnership with them. While it is necessary for an organising force to develop and deliver the PB programmes, and for that to at the very least include government, the lack of citizen involvement in the creation of the programmes feels incongruous alongside the underlying participation and social justice principles of PB in general.

4. Funding of participatory budgeting

An under-discussed aspect of PB programmes is where the money that is being spent originates from. When a PB programme is introduced it will either be to spend new money (such as an extractives fund, grants, or an increase in tax revenue) or a redistribution of existing money. In the Wampler et al Phase 1 submission, it is also noted that decentralisation frees up money for PB programs, as the money is temporarily unallocated and has no defender.

When PB is implemented without an influx or with unallocated money, it is likely to have been channelled by legal change from above, without the buy-in of authorities in question. In these instances, these authorities will be strongly motivated to subvert the system to retain control of their budget. In Mexico City, although only 3% of an authority's budget is spent on PB; once staff salary costs are taken into account, the percentage of the remaining budget this represents is much larger. Given this, it is understandable that effort is expended on retaining control of it .

Although rarely emphasised in the literature, one of the key features of the Porto Alegre PB was a massive increase in the spending ability of the city. Ganuza and Baiocchi detail the scale of the financial problems and changes in Porto Alegre at the time:

"The reforms introduced tax progressivity in the two most important Brazilian municipal taxes, the taxes on real estate, and on services; different utility rates were updated and indexed to inflation; control was also increased over tax fraud. Real estate taxes became increasingly important, going from 5.8% of the volume of municipal revenues to nearly 18% of total revenues, while the services tax went on to account for 20% of municipal revenues (Santos, 2002: 68). Tax evasion was also significantly reduced over ten years later (Baeirle, 2003).These reforms essentially doubled the city's income over the ten years[emphasis added]. To put it in context, while roughly two thirds of municipal budgets in Brazil, overall, come from transfers from higher levels of government during the time period, Porto Alegre´s own revenues accounted for almost 60% of its budget."28

By minimising this part of the story, Porto Alegre is used to support a very broad portrait of the benefits of participatory budgeting, where "[i]nstead of pointing to fiscal reforms as a pre-condition to PB, increased revenues were now sometimes framed as an outcome of PB".29

In this reversal in the common narrative the question is lost as to whether PB was a useful tool in securing this additional funding. Just as PB was self-consciously used as a selling point by the Workers Party in Porto Alegre when pitching for external funds, was the same pitch useful when raising taxes from its own citizens? In systems with a history of corruption and poor public financial management, people might be happy to pay more tax in theory, but in reality are deeply sceptical of the utility of it. PB functions as a way of demonstrating what the additional money is being used for. In the Mexican state of Jalisco a PB system is being implemented explicitly with the aim of incentivising higher tax-payers, with a greater say in participation given to those contributing the most tax.30 While from one view it is the opposite of the Porto Alegre model in that more votes for the wealthy is regressive, it shares a similarity in using PB as a tool of public trust while increasing public spending and reducing tax evasion.

For induced PB projects (where they haven't emerged spontaneously, as is case with the most studied examples in Brazil) it could be posited that there will be less push-back from authorities when it is paired with an influx of new funding. While individuals may still attempt to profit, there is less institutional requirement to regain control of old funds. This suggests that PB is potentially a good idea when a windfall comes to an area with low trust in government, or when an increase in the tax base is needed. However, new legal requirements for PB in the budget might backfire by raising the motivation of affected authorities to act against the process. Any presumed financial efficiency benefit of PB is instantly lost if the authority is incentivised to mobilise and capture the process. The costs required to commute this might negate any efficiency gain.

5. Political vulnerability and corruption

One of the core vulnerabilities of PB in analysing its value in increasing trust and improving service delivery is that it exists by the grace of politicians. As mentioned in the earlier section, the implementation of PB is normally a top-down political decision, either taken at the behest of the incumbent or newly elected party, to satisfy legal requirements from higher levels of government, or adopted to unlock development and aid funding from international institutions. Because the monies being budgeted are, in all but a small number of cases, held by government, the incumbent legislature essentially holds the power over the process and its enduring existence. A lack of political buy-in to the PB process can render it toothless and starve it of resources and necessary support. In Mexico City, participants noted that in some areas there seemed little incentive for authorities to promote or engage with PB. A lack of local authority buy-in has therefore led to a corresponding lack of investment in the PB programme, and greater incentives to subvert it. Information on the process and where responsibility for it rested was not easily available to citizens, causing prolonged confusion and eventual disengagement. While PB is often billed as a way of increasing trust in government and reducing corruption through openness, as a mechanism of public participation it by necessity includes openings that allow the possibility for these supportive mechanisms to be recaptured.

The issues can emerge at the very start of the process. Individuals in structural positions of power are able to disrupt, subvert or co-opt the PB system when it is in operation. A participant commented on the Kenya system:

"There are Mayors here that aren't politically invested in the system, even though its mandated in the constitution. I know of cases where the local Mayors have set up a PB process, but then stuffed the group with their friends who will do whatever the Mayor wants. It ends up just an exercise on paper"

A number of participants commenting on the Kenyan PB processes identified issues with the representatives responsible for decision-making being politically installed, amongst other instances of political subversion of PB processes. In Mexico City, the previously discussed problems of local authorities submitting projects through citizens sidesteps popular input. In circumstances where the first round of the proposal is written, proposals might have to be actively solicited and advice given to ensure even participation. Without strong citizen engagement at the start and neutral but comprehensive support from programme implementers, it is easier for the process to be captured by technically superior projects from official sources.

Similar problems can be found in other countries. One participant pointed to an instance in Brazil where local police officers were actively stopping citizens en route to vote, completing the participation paperwork themselves and directing citizens how to vote. In countries with a recent or ongoing history of electoral corruption, PB might be controlled or subverted using the same toolkit used to deliver elections (and for similar clientelist reasons).

As previously discussed in the Kenya case study, show-of-hands voting is also vulnerable to pressure. When popular participation is low, it is easier for the room to be stacked to the advantage of partial interests. Ultimately, institutionalised and embedded practices will adapt to preserve the status quo, even in situations where disruptive interventions such as PB are implemented. In PB more deliberative meetings might be structured to limit access to materials, or with a schedule that limits discussion of outside projects. It was pointed out that even a secret ballot does not prevent collective punishment when voting for projects in small areas. If an area did not support an agreed project, it would be clear in the result, and as such it is clear that the locality of PB projects can work against the effectiveness of a secret ballot when the size of electorate is small.

Ultimately, institutionalised and embedded practices will adapt to preserve the status quo, even in situations where disruptive interventions such as PB are implemented. In PB more deliberative meetings might be structured to limit access to materials, or with a schedule that limits discussion of outside projects. One participant noted in Kenya that collective retribution was a real risk for some citizens as a result of their decision-making. In one case, ‘punishment' had been experienced in a particular area because that area had not chosen projects that the local politicians had made clear they wanted. The punishment was allegedly the deliberate neglect of local services and infrastructure by local politicians in the months following the project decision, and demonstrates one of the more brazen consequences of participation in PB processes that have been structurally compromised.

6. Monitoring

Once a project is decided on, there remains the process of converting it from idea to reality. Without sufficient monitoring, there is scope for funds to go missing or for the project to fail to materialise altogether. Project monitoring should be an integral component of the process, and an important one in retaining public confidence and trust in the programme. That said, it is often a forgotten part and where present is the result of ad-hoc solutions rather than planning.

In Mexico City the IEDF have built a website that encourages social media reporting of project statuses so that there is some room for popular monitoring of the construction process. The utility of this tool in Mexico City is limited by the fact that IEDF has no legal powers of enforcement, but it has partnered with other offices of the government of Mexico City to attempt to hold corrupt officials to account. In a PB process in Embu, Kenyan citizens formed what they called ‘project monitoring committees' to oversee the process. These processes reflect that while monitoring becomes recognised as important and efforts are made, these are rarely considered from the outset in PB programs and given the authority required.

7. Costs of managing opposition to participatory budgeting

Where actors are actively working against the process, creating a virtuous PB cycle with positive social spillover has been shown in this research to be extremely hard work. In systems where there are competing efforts both to capture participatory processes and to prevent capture, the cost of waging this battle increases progressively for both sides. As seen in the example of online voting in Mexico City, the more the cost is raised for subverting the process, the higher the cost of participation also. You need to spend more time holding secret votes rather than shows of hands, and to ensure projects happen, you need to pay for or support monitoring. The objective of popular participation and encouraging trust in the process may be far more expensive than first appearances indicate. If being deployed as an anti-corruption function, monitoring and anti-subversion mechanisms need to be priced in from the very start.

Recently, the complete withdrawal of the PB programme in Porto Alegre, the spiritual home of PB, has been identified as a political decision by the government to avoid the need to engage with citizens wishes and to impose its own programme of activities. That PB has been halted in the very place that inspired the global proliferation of the process should act as a cautionary tale that PB is difficult to permanently institutionalise, and that while it relies on government funds and government buy-in, it will remain vulnerable to the whims and institutionalised attitudes of political and public administration actors.

8. Diffusion, networking and knowledge sharing

Participatory Budgeting is an enormously successful policy, in that its diffusion has been global within the space of 30 years. It is discussed as an example of ‘fast policy', rapidly spreading from one location, but transformed and evolved in independent pockets along the way. PB became "associated with a cluster of material (and, on the face of it, replicable) practices, from indicative schedules for neighborhood assemblies to refined rules of deliberation and voting, all of which endowed the model with a concrete, tangible form".31

Porto de Oliveira tracks the diffusion of the idea of PB in his book ‘International Policy Diffusion and Participatory Budgeting'.32 In particular he draws attention to the role of the World Bank, the EU and the UN as international organisations that propel the diffusion of PB, and suggests that there are also general expertise-sharing networks and peer-to-peer contact between municipalities. These represent both organisations looking for better knowledge, and also those actively positioning themselves as conveners. These might be individual municipalities looking to Porto Alegre as an example to emulate, NGOs promoting PB in their area, or international organisations actively looking for best practice to fulfil their governance goals. For larger organisations, there may be different PB programs operating from different offices, for instance Porto de Oliveira found that in addition to the Agency for Human Settlements (UN-Habitat), there were also experiments in PB funded by UNICEF, UNDP and UN-Women in Latin America and Africa.

While new organisations might look to PB to consider whether it is a fix for their existing approaches (or problems with those approaches), organisations connected with PB can have interest in dedicating resources to advocating for PB. One aspect Porto de Oliveira draws out is the work done by officials in Porto Alegre to trade on PB to make a name for itself as a leader of democratic innovation. This produced dividends not only in global standing, but in continued funding to support the project. More "than US$240 million was secured (both national and international) for the construction of infrastructure, between 1993 and 1998, 150 million being in the form of loans"33. Porto de Oliveira's interviewees highlighted the emphasis put on funding bids by the city on the presence of PB "when the Mayor's Office sought funding from international organisations, it was always mentioned that public policy went through a process of participatory governance, that is, PB"34. Porto Alegre secured over US$240 million for infrastructure between 1993 and 1998 (mostly in the form of loans). The Inter-American development fund explicitly designated around US$21 million to PB work, so while in some cases PB requirements are attached by the funders, in this case it was also used by cities themselves as an indicator of responsible and innovative government.35

Whatever the motivations for implementing PB, there is often an expertise gap locally which needs to be filled in order to get the PB programme off the ground. In such situations a global peer network and centralised organisation that holds information and advice on best practice in PB would be extremely useful. However, such resources do not really exist, and advice and knowledge on PB tends to be dispersed amongst a few civil society and practitioner websites. While these sites are useful, they do not provide comprehensive guidance, and because they lack any official or accredited status, they are not a sufficient resource from which to develop a new PB programme. In the course of this research, there appeared to be a very small number of well known PB consultants and champions in the global PB space who filled this expertise gap, and while their contributions to developing PB programmes are significant, this global reliance on a small number of private actors is not ideal for the sector as a whole or for its sustainability and prosperity.

While there are few international networks dedicated explicitly to PB, with the International Observatory on Participatory Democracy (IOPD) being the main example, international diffusion often occurred through connecting PB to events and networks for leaders and governments of municipalities. There is then the variety of more local co-ordinating networks that have been identified in previous studies; however, these networks demonstrate significant fluidity and a lack of permanence. De Oliveira notes that there are separate existing networks for PB in Brazil, Chile, Colombia, the Dominican Republic, Portugal and Argentina, as well as IOPD. There are references in both de Oliveira's book and the IOPD website to an African branch which suggests that a network existed at one point, but those URLs now lead nowhere. Similarly, references to PB resources produced by the PGU-ALC in earlier PB research point to another case where former networks and resources are now absent. The "Radically Democratize Democracy" network (RDD) which acted as a major vector for dissemination of PB in Europe is now inactive, while Mercocities was superseded by the Forum for Local Authorities for Social Inclusion and Participatory Democracy (FLA) network. People involved in these organisations might often be found in successor organisations or networks, but this reflects the difficulty of sustaining international PB co-ordination over time, and the lack of an enduring and centralised organising network that can sustain and grow the PB community at a global level. One potential solution to this is the development of the OGP sub-national initiative, which, while in its early stages, could provide a stable international community space for PB practitioners and academics. The current pilot stage of OGP subnational has 15 regions involved in implementing a variety of open government initiatives, including PB. If the pilot is successful, many more regions are enthusiastic to join, and this could form the beginnings of a nascent international PB community that is able to integrate more widely with linked communities of interest, such as organisations focused on themes of transparency, accountability, tech and public administration.

9. Longterm issue for networks

One of the longterm problems for horizontal networks is that while groups like the World Bank have a more technocratic understanding of PB, when arising organically it is often associated with the election of a particular party or mayor, and as such is understood as a measure owned by that political group. One of de Oliveira's interviewees argued that "the FLA had, in Porto Alegre, not only its brand but a protagonist"36 and so was disrupted by the Workers Party's loss of control of the city. This means where PB arises more organically (and with less external pressure) it is more likely to be a vulnerable local project and sustained co-ordination over time becomes more complicated. Even where PB efforts are legally required, less friendly administrations have many potential tools to undermine the process.

The translation of a book describing Porto Alegre's experience into multiple languages reflects how PB was spread, but also the problem in subsequently sharing experiences. The amorphousness of PB has allowed its adaption to multiple contexts and forms, but these different forms might then be less intelligible to each other. As Porto de Oliveira put it "[w]hile PB circulates, the polymorphism of its models increases"37. An example of this spread of PB, in which different local implementers adapt PB to local contexts, can be seen in UN-funded PB. Departments within UN-Habitat created two separate guides, one for Francophone Africa, and one for South and East Africa, to reflect the different linguistic and administrative features of the countries. Advice and guidance differs between the two, creating a situation in which PB implementers from different regions may end up with very different understandings and experiences of PB, and not only speak different territorial languages, but speak different PB languages too. This makes networking and peer support more difficult, but could have the potential of improving PB practice if PB users in different areas were able to share evidence of what kinds of system work and what doesn't. The key issue here is a the absence of a strong and enduring network governed and funded by a central organisation. Diffusion of programmes such as PB only has to happen once, however constructive networking and knowledge-sharing is an ongoing process, and will have most value when the participants are able to develop a similar frame of reference and to engage on an ongoing basis with the wider PB community.

10. The World Bank influence

An additional problem of co-ordination is the different understandings groups have of the purpose of PB. Not all descriptions of its diffusion are positive, and the amorphousness of PB is often held against it. Terms like "depoliticised", "stylised and defanged", "fig leaf" or "PB-lite" are used describe the version of PB that spread around the world, in comparison to the PB originally practiced in Porto Alegre which was part of a larger series of explicitly socialist reforms. There are concerns about the motives and abilities of the World Bank (now the largest promoter and funder of participatory budgeting schemes) in promoting PB, specifically around the almost blanket requirement for funding recipients to implement a form of PB. However, there is also praise for the World Bank championing it as a policy. As Goldfrank draws out the different camps: "some on the left have celebrated the Bank's funding and advocacy for PB as signifying the legitimacy or mainstream success of the process, while others see the Bank's endorsement of PB as a sign that participatory budgeting is becoming watered down and losing its transformative potential, if it ever had such potential."38

The attachment of PB to wider flows of development cash adjusts motives for why an area might adopt participatory budgeting, for instance, to meet the participation requirement for debt relief in HPIC II. The entanglement of PB in the World Bank's wider goals is part of the scepticism of its role in the spread; for instance, Cammack notes: "while the Bank's commitment to poverty reduction is real, within limits, it is conditional upon, and secondary to, a broader goal. Its principal objective is the systematic transformation of social relations and institutions in the developing world, in order to generalise and facilitate proletarianisation and capitalist accumulation on a global scale, and build specifically capitalist hegemony through the promotion of legitimating schemes of community participation and country ownership"39. This could be considered to conflict slightly with the socialist origins of PB in Brazil.

Tempting though it is to anthropomorphise organisations, Goldfrank finds that the World Bank isn't monolithic in its view towards PB. There are different understandings of the purpose of PB from different individuals inside the organisation, and similarly a variation in levels of belief of its effectiveness: "Within the World Bank one can find both kinds of PB advocates – those who believe in PB's democratizing potential and those who use the language of participation as a kind of Trojan horse for their own marketizing agenda – but that the Bank as an institution is not committed to PB in the same way that it has embraced Poverty Reduction Strategy Papers. In other words, despite the Bank's importance in the globalisation of participatory budgeting, PB is not very important to the Bank."40

The World Bank puts its emphasis on the waste-reduction potential of PB, both in selection of projects and the corruption promoting potential. It is also useful as a system that helps address criticism of outside actors micro-managing aid interventions. "Theuer (2010) documents the rising importance of participation in World Bank publications over the course of the 1990s and shows how the Bank was both responding to civil society actors' demands for participation and good governance and also attempting to avoid cost overruns and corruption in its projects".41

This can be seen in the way the World Bank promotes success of PB projects in Kenya, highlighting the project of a milk cooler that exists as a local result of more women engaging in PB process. This example demonstrates how PB makes funding responsive to local demand, as well as acting in a way where the investment can produce a profit.42 Or in another example, where a PB process led to an abandoned borehole being repaired to supply a village with water, a need which would have been completely overlooked by someone trying to allocate funding for providing clean water centrally. There are clearly examples of positive outcomes of PB around the world; however, the volume of demonstrably positive outcomes that would have been unlikely to have been achieved through other processes is questionable, and there is, as noted previously, no evidence of areas considering alternative programmes to PB, because the World Bank promotes PB as standard.

This report does not suggest that the World Bank is wrong to promote PB as a tool for participation and corruption reduction, but the research undertaken demonstrates that the PB developed in response to the World Bank's funding conditions is, in many places, very much inferior to the original Porto Alegre model it was based on. There also seems to be a lack of consideration of alternative methods for waste-reduction and participation, in particular because the most effective PB programmes are often the ones with the highest support and administration costs. Additionally, as discussed in the earlier section on political and administrative ‘buy-in', without meaningful political investment in the programmes beyond satisfying the terms of funding, PB is vulnerable to becoming a paper exercise, and no amount of World Bank encouragement and monitoring can prevent programmes from becoming listless or irrelevant in that situation.

11. Going digital

The migration of PB onto digital platforms has been widely embraced in developed municipalities, with some developing regions also beginning to follow suit. Very few instances of PB done digitally have been running for more than one to two years, and so evidence of effectiveness is still elusive. There are, however, several observations on digital PB that research participants were able to share. A very commonly noted positive aspect of using digital platforms for PB was the ability to scale programmes to involve thousands of people, rather than purely in-person initiatives that can only handle a fraction of that number. In Madrid, Paris and Barcelona, PB implementers were very proud of the number of individuals participating digitally, with reports of between 1-5% of the populations of those cities getting involved in the PB process. Of course, relative to total population, these figures are still very small, but represent a new level of PB engagement that had previously been unattainable. The level of ease and time-saving that digital PB afforded citizens was another commonly cited positive aspect of this method of PB. Whereas in-person PB represented a significant time commitment, digital was seen as a way of engaging individuals with busy schedules in the process, and therefore capturing a broader range of voices.

In contrast to the claims about digital PB making it easier for programmes to be diverse and inclusive in the composition of participants, several PB experts raised concerns about the potential for digital PB to essentially gentrify the process. Participants noted that those most likely to engage digitally via a website or app would be the most affluent, most educated, older, white (or dominant ethnic class) and residents of more affluent areas. These assertions are in keeping with previous research conducted into tech-for-good and Civic Tech, which show that users are predominantly older, male, affluent and comparatively well educated.43 Whilst the majority of digital PB programmes included a small offline element, in practice, very few citizens in the cities examined for this research chose to engage in an offline capacity, and this was suggested to be because of the lack of direct engagement with disadvantaged citizens. One of the key aspects in the success of PB programmes has historically been in the educative aspect of the programmes, in which target groups of citizens without a good working knowledge of the civic space receive contextual information and education about the running and powers of local government though the PB process. Historically delivered face to face, this civic education aspect of PB programmes is somewhat lost in digital migration, and therefore citizens lacking an existing understanding of the local municipality rules, responsibilities and regulations are shying away from engaging. This gentrification of PB represents a serious threat to its legitimacy, as even pre-distribution phases taken by the organising municipality may not be able to offset the bias that may emerge in the submission and award of projects from predominantly one section of the population.

Another issue that was raised by participants in this research concerning digital PB was the quality of digital portals being used to deliver the programmes. As mentioned earlier in this report, PB is a somewhat isolated participatory community that does not interact with linked international networks as much as might be beneficial. As such, the PB community has not been exposed to the Open Data or Civic Tech communities that hold significant expertise in digital engagement and participation. Digital PB portals in use in Paris and Madrid, whilst very well funded, demonstrate a poor understanding of ‘user experience' and are confusing, clunky and complex. Again, while ‘digital natives' and individuals familiar with official process will make the effort to navigate these user-unfriendly portals, individuals who lack digital or civic expertise, or simply individuals who are short on time, will simply choose not to engage. Poor quality digital platforms may also accommodate subversion and corruption more easily than in face-to-face programmes, as citizens are unable to see who is voting or how many people are engaged. A digital system can be ignored or gamed to produce the politically desired outcome, and citizens may never be able to prove that the process was corrupt. This issue has been observed in Mexico City; however, attempts to reduce corruption by using passwords for accounts and installing other measures has increased the effort required to participate with a corresponding drop in participation.

It was suggested by one participant that technology might be better used for the monitoring aspects of PB (as can be seen in Mexico City's PB). Whereas the use of digital may prejudice project submission and voting, citizen monitoring could potentially be easily crowdsourced via digital platforms. Once projects have been chosen and awarded, interested citizens could provide progress and budgeting updates, and on this side of the process diversity would be less of an issue.

Digital PB has the potential to be significantly more engaging and inclusive than it currently is. While there are moves towards more user-friendly portals, such as developments by Empatia44, these are piecemeal and short-term, and these development projects take a very top-down approach due to the nature of their funding and the composition of the projects, which in the case of Empatia, relies on short term EU funding and comprises a consortium project dominated by universities. One participant noted that, in particular with digital methods, PB programmes needed to professionalise in order for those digital platforms to work effectively. As noted previously in this report, short-termism in networking, funding and development does not benefit the evolution of the PB community or contribute to the compilation of knowledge and best practice, and this applies to the digital aspects of PB and its evolution. PB programmes using digital methods could benefit significantly from exposure to, and collaboration with, the tech-for-good community, and with the corresponding academic community that hold a growing body of knowledge on best practice in digital engagement.

Further research and recommendations

This study, in addition to the work carried out for this project by Wampler, McNulty and Touchton, identified a number of areas where the information deficit concerning PB was significant. The first two recommendations made below concern further research in this area that the authors judge would be beneficial to the William and Flora Hewlett Foundation and the Omidyar Network in informing their respective future decision-making.

The third and fourth recommendations concern practical interventions that have been identified as a clear need in the current PB landscape. These are supported by the information gathered from this study and from the existing literature, and are aimed at supporting the global PB community to professionalise and integrate with the wider global network on open governance and participation.

Recommendation 1

Fund better targeted and comparative research into PB in areas of interest

There are a number of potential complexities in attempting to produce comparative quantitative research into PB that make large-scale studies difficult to conceive. Whereas it could be possible to compare PB in Madrid, Paris and Barcelona due to the mostly homogenous nature of their structures and stability of their governments, it would be difficult to conduct a large N comparison of different existing PB exercises within Kenyan counties due to their programmatic differences, or compare PB exercises in differing Latin American countries because of their differing political and administrative structures. PB is a slippery concept that mutates as it evolves and disperses, and therefore presents significant difficulties to those wishing to compare and contrast. Qualitative studies could, however, surmount this issue, using interview, ethnography and/or documentary evidence to analyse impact in a more deep and meaningful way. Such studies would be expensive, but would likely provide the most useful understanding of long-term impact. The findings of this report support the evident need for more research as recommended by Wampler, McNulty and Touchner in the literature review, concerning conducting larger, comparative and cross-regional research into the operation and impacts of PB; however, only as a very targeted approach into areas of specific interest. It is not expected that the Hewlett Foundation nor the Omidyar Network wish to become the world's primary funders for PB research, and large, cross regional and academically rigorous studies are expensive and time-consuming to run. Therefore, it may be more beneficial for slightly smaller, targeted research to be conducted on specific themes or areas of interest, such as qualitative comparative studies of counties within Kenya, studies into the potential gentrification effect of using digital for PB, or studies comparing deliberation vs voting. The recommendations provided in phase 1 of this project provide a range of interesting targeted approaches, and so are not duplicated in full here.

Recommendation 2

Fund specific research into citizen trust and attitudes towards governing bodies in areas where PB has failed or been withdrawn

In addition to examining the specific impacts of PB processes upon citizens, there are important intellectual questions that should be considered at a wider level. With regard to citizens, examining the effects of non-linear and unsuccessful PB programmes on their lives and attitudes would provide valuable insights. If PB is generally implemented to give citizens a better voice, to leverage greater benefits into communities and to build trust between citizens and institutions, then what happens when PB fails or is halted? This is a pertinent question, given the recent withdrawal of PB in Porto Alegre in particular, and also the issues and failures of PB in places such as Peru and Madagascar. Firstly, such research could provide an understanding into how embedded and important the programmes became in citizen relationships with governing institutions, and secondly, could demonstrate if and how reduction or cessation has impacted on their attitudes and behaviours. If the loss of PB programmes has no discernible impact upon citizens' relationships with governing institutions, then this may demonstrate a lack of interest or value in PB to citizens. If citizens are angry or disappointed at the withdrawal of the programmes, this could demonstrate a belief that the programmes were working well and were achieving the original goals of PB during their operation. And, where PB programmes had been withdrawn, how did this manifest in citizen attitudes and beliefs about governing institutions? The study of failure can often illuminate how things actually work and prevent similar mistakes happening again, and yet this research found that very little effort has thus far been expended in this pursuit.

Recommendation 3

Establish (either through support of an existing organisation, or creation of a new one) a dedicated PB organising body

The PB practitioner community has been shown by both phases of this research to be somewhat dispersed and poorly connected. While PB instances are being implemented in almost every corner of the world, the relationships between them are weak and often linked by only a small number of consultant experts. Even within overarching institutions that have systematically promoted PB, such as the World Bank, PB instances are mapped poorly and operate in an unconnected fashion, with multiple staff across multiple departments involved in PB in one way or another, but quite isolated from each other. This presents a risk to employing effective PB that is informed and shaped by the availability and consumption of information on both best practice and failure. Few networking opportunities exist to bring together PB practitioners, and PB has not been absorbed fully into the more global open governance, transparency and accountability, or participatory movements. It stands alone as something that predates many of these communities, and suffers from its existence at the sub-state level, making it much more difficult to co-ordinate, and relying on the interest and enthusiasm of public officials to engage with the wider community. Whereas some of these barriers may be broken down with the advent of OGP sub-national, in which 15 pilot areas are currently experimenting with PB alongside other linked open government practices, there is still significant benefit that could be created with a more coherent global organising body to bring this community together. There are some nascent networks, such as Participedia and Empatia, but these are limited in scope and resource. IOPD is possibly the closest organisation to resembling a global PB network, however as a resource is again limited in scope and capacity. An organisation devoted to solidifying the practice, expertise and community aspects of the global PB landscape, and drawing it into the wider governance and accountability space, may enable it to grow beyond the isolated practitioner phase of activity into using PB more strategically to achieve specific outcomes.

Recommendation 4

Establish a senior expert PB forum/committee comprised of global PB stakeholders to attempt to build consensus on improving PB implementation and its outcomes through institutional change

The two phases of this study both identified the significant influence organisations such as the World Bank, UN-HABITAT, the EU and other supranational development organisations have had on the spread of PB. The study also identified the lack of specific goals and objectives driving the implementation of PB, other than the belief that participation is a good thing, and may reduce corruption. Participants spoke of individuals and institutions finding creative ways to circumnavigate PB in order to engage in corrupt practices, and several participants suggested that PB that was imposed by ‘higher powers' tended to be somewhat watered down from its Brazilian origins and not embraced by communities. Issues such as this are likely to continue as long as supranational agencies continue to promote PB without a more strategic understanding of what outcomes it is implemented to achieve, and whether it is the most suitable mechanism. It is likely beyond the ability of the Hewlett Foundation and the Omidyar Network to change the working practices of behemoth institutions such as the World Bank or UN, however, it may be fruitful to assemble a semi-formal committee or similar body from these and other expert organisations to work on addressing the incoherence, waste, duplication and dispersed nature of the global PB landscape. Senior and influential individuals engaged in the work of this committee could therefore be best placed to change institutional behaviour from within, to improve PB for citizens all over the world.


The global proliferation of PB as a participatory mechanism has picked up speed over the last ten years, and with this rapid expansion has come significant divergence from the traditional model developed in Porto Alegre in the 1980s. While divergence and expansion are no bad thing, the implementation of PB in many areas has lacked vision and strategy in the race to implement mechanisms that on the surface should theoretically achieve greater citizen participation and reduce corruption, but that in practice may be vulnerable to politicking, apathy and subversion. The weaknesses in global PB may in part be attributed to the very weak links between practitioners and overarching global programmes to achieve greater open governance, accountability and participation, but also stem in part from a lack of coherent strategic vision in implementation, and the diversity with which PB programmes have developed away from their roots. And there are many questions left to answer. Should PB be purely participatory or fundamentally redistributive? Should PB be based upon voting or can deliberative methods produce more sound results? Do the gentrification risks of digital PB devalue it? These are important and urgent questions given the volume with which governing institutions are embracing PB.

This study has not provided an exhaustive or even entirely comprehensive overview of PB operating globally, but has examined themes and issues at the meta level to understand how PB may be supported in its development to become as influential and beneficial to citizens as its potential suggests it might. It is hoped that the findings and recommendations contained within this report, coupled with the phase 1 project work by Wampler, McNulty and Touchton, will inform the Hewlett Foundation and the Omidyar Network in their decision-making regarding the funding of PB in future.

About mySociety

mySociety is an international not-for-profit social enterprise based in the UK, where we run a number of projects designed to give people the power to get things changed. We invent and popularise digital tools that enable citizens to exert power over institutions and decision makers, and work internationally to support partners who deploy our technology in countries around the world. As one of the first Civic Technology organisations in the world, we are committed to building the Civic Technology community and undertaking rigorous research that tests our actions, assumptions and impacts.


1:Sheely, R. (2015). Mobilisation, Participatory Planning Institutions, and Elite Capture: Evidence from a Field Experiment in Rural Kenya. World Development, 67, 251–266.

2:Fahuma, Facilitating Participatory Budgeting in Kenya,, p . 32

3:Carothers Flores via Harbers, I. (2007). Democratic deepening in third wave democracies: Experiments with participation in Mexico City. Political Studies, 55(1), 38–58, p. 46

4:Harbers, Democratic deepening in third wave democracies, p. 47

5:Harbers, Democratic deepening in third wave democracies, p. 48

6:Harbers, Democratic deepening in third wave democracies, p. 48

7:Harbers, Democratic deepening in third wave democracies, p. 52

8:Seelee, Ziccardi via Fox, J. A. (2007). Accountability Politics: Power and Voice in Rural Mexico. Oxford: Oxford University Press, p. 184

9:Sánchez, L. M. (2017). Participación ciudadana en la Ciudad de México a través del presupuesto participativo. In L. M. Sánchez & M. M. Plata (Eds.), La participación ciudadana, esencia de la gobernonza urbana. Un analisi desde el presupuesto participativo en la Ciudad de México. (pp. 25–48). Mexico City: Universidad Autónoma de la Ciudad de México.

10:IEDF. (2017). Resultados del Reconocimiento de Buenas Prácticas del Presupuesto Participativo 2017. Retrieved November 8, 2017, from

11:Plata, M. M. (2017). El régimen jurídico del presupuesto participativo en México: reflexiones desde la democracia y la administración pública. In L. M. Sánchez & M. M. Plata (Eds.), La participación ciudadana, esencia de la gobernonza urbana. Un analisi desde el presupuesto participativo en la Ciudad de México. (pp. 49–76). Mexico City: Universidad Autónoma de la Ciudad de México., p. 63

12:Plata, M. M. (2017). El régimen jurídico del presupuesto participativo en México: reflexiones desde la democracia y la administración pública. In L. M. Sánchez & M. M. Plata (Eds.), La participación ciudadana, esencia de la gobernonza urbana. Un analisi desde el presupuesto participativo en la Ciudad de México. (pp. 49–76). Mexico City: Universidad Autónoma de la Ciudad de México.,

13:Wiemann, J., & Fuchs, M. (2016). Participatory Budgeting in Mexico City: Pre-structuring of participation by territories and topics. Geographisch Zeitschrift, 104(2), 92–111. + interview (2017).

14:Fox, Accountability Politics: Power and Voice in Rural Mexico, p. 184

15:Calles, M. A. (2017). El presupuesto participativo en la Ciudad de México y los problemas en su ejercicio. In L. M. Sánchez & M. M. Plata (Eds.), La participación ciudadana, esencia de la gobernanza urbana. Un analisi desde el presupuesto participativo en la Ciudad de México. (pp. 50–77). Mexico City: Universidad Autónoma de la Ciudad de México., p. 84

16:Calles, M. A. (2017). El presupuesto participativo en la Ciudad de México y los problemas en su ejercicio. In L. M. Sánchez & M. M. Plata (Eds.), La participación ciudadana, esencia de la gobernanza urbana. Un analisi desde el presupuesto participativo en la Ciudad de México. (pp. 50–77). Mexico City: Universidad Autónoma de la Ciudad de México., p. 90

17:Grillos, T. (2017). Participatory Budgeting and the Poor: Tracing Bias in a Multi-Staged Process in Solo, Indonesia. World Development, 96, 343–358.

18:Vicente, P. C. (2014). Is Vote Buying Effective? Evidence from a Field Experiment in West Africa. The Economic Journal, 124(74), F356–F387.

19:Turid, H. (2010). Elections in the Context of everyday practices: Views from a pueblo originario in Mexico City 2006. Forum for Development Studies, 37(2), 217–241.

20:Translated - original in Plata, M. M. (2017). El régimen jurídico del presupuesto participativo en México: reflexiones desde la democracia y la administración pública. In L. M. Sánchez & M. M. Plata (Eds.), La participación ciudadana, esencia de la gobernanza urbana. Un analisi desde el presupuesto participativo en la Ciudad de México. (pp. 49–76). Mexico City: Universidad Autónoma de la Ciudad de México.,p. 66


22:Sánchez, L. M. (2017). Participación ciudadana en la Ciudad de México a través del presupuesto participativo. In L. M. Sánchez & M. M. Plata (Eds.), La participación ciudadana, esencia de la gobernanza urbana. Un analisi desde el presupuesto participativo en la Ciudad de México. (pp. 25–48). Mexico City: Universidad Autónoma de la Ciudad de México.

23:Standing, A. (2014). Ghana's extractive industries and community benefit sharing: The case for cash transfers. Resources Policy, 41(1), 74–82.

24:Williams, E., Denny, E. S., & Bristow, D. (2017). Participatory Budgeting : An Evidence Review Participatory Budgeting : An Evidence Review,, p.1

25:Justice, J. (2011). Review: Why Bother with Budget Work? Public Administration Review, 71(1), 126–128., p. 126

26:Fung, A. (2015). Putting the Public Back into Governance: The Challenges of Citizen Participation and Its Future. Public Administration Review, xx, 1–10., p.519

27:Goldfrank, B. (2011). Participatory budgeting and urban sustainability: Reviewing lessons from Latin America. In H. A. Mieg & K. Töpfer (Eds.), Institutional and Social Innovation for Sustainable Urban Development (pp. 570–71). Oxon: Routledge, p. 66

28:Ganuza, E., & Baiocchi, G. (2012). The Power of Ambiguity : How Participatory Budgeting Travels the Globe. Journal of Public Deliberation, 8(2), 1–12. p, 6

29:Ganuza, E., & Baiocchi, G. (2012). The Power of Ambiguity : How Participatory Budgeting Travels the Globe. Journal of Public Deliberation, 8(2), 1–12. p. 7

30:Interview with IEDF official (2017)

31:Peck, J., & Theodore, N. (2015). Fast Policy: Experimental Statecraft at the Thresholds of Neoliberalism. Minneapolis: Univ Of Minnesota Press, p. 211

32:Porto de Oliverira, O. (2017). International Policy Diffusion and Participatory Budgeting: Ambassadors of Participation, International Institutions and Transnational Networks. Palgrave Macmillan.

33:Fernandes via Porto de Oliverira,International Policy Diffusion and Participatory Budgeting, p. 74

34:Fernandes via Porto de Oliverira,International Policy Diffusion and Participatory Budgeting, p. 74

35:Fernandes via Porto de Oliverira,International Policy Diffusion and Participatory Budgeting, p. 74

36:Fernandes via Porto de Oliverira,International Policy Diffusion and Participatory Budgeting, p. 119

37:Fernandes via Porto de Oliverira,International Policy Diffusion and Participatory Budgeting, p. 74

38:Goldfrank, B. (2012). The World Bank and the Globalisation of Participatory Budgeting. Journal of Public Deliberation, 8(2), 20. Retrieved from, p. 2

39:Cammack via Goldfrank, The World Bank and the Globalisation of Participatory Budgeting, p. 6

40:Goldfrank, B. (2012). The World Bank and the Globalisation of Participatory Budgeting. Journal of Public Deliberation, 8(2), 20. Retrieved from, p. 7

41:Goldfrank, B. (2012). The World Bank and the Globalisation of Participatory Budgeting. Journal of Public Deliberation, 8(2), 20. Retrieved from, p. 6

42:World Bank. (2017). Inclusive and Effective Citizen Engagement: Participatory Budgeting - Makueni and West Pokot Counties., p. 40

43:Rumbul, R. (2015). Who Benefits From Civic Technology?. [online] mySociety Research. Available at: [Accessed 15 Nov 2017].