Porto Alegre participatory budgeting: the original model, its peak and its long decline

Porto Alegre participatory budgeting (Orçamento Participativo, or OP) is the process, begun in 1989 in the southern Brazilian city of Porto Alegre, through which residents meet in regional and thematic assemblies to set investment priorities and elect delegates and councillors who negotiate the city’s public works with the administration.⁠¹ It became the world’s reference model for participatory budgeting, drew a record of more than 20,000 people to its assemblies in 2015, stopped holding assemblies from 2019 to 2022, and resumed afterward.⁠^1,2 Its record is the clearest evidence available that participation tracks delivery: when approved projects stopped being built, residents stopped believing they decided anything.

What happened

Porto Alegre’s OP began with a change of city government. The city’s own description dates its implementation to 1989;⁠¹ the national study discussed in the comparative overview dates the name and assembly format to 1990, so both dates appear in the literature.

Growth (1990 to 2002). A 2013 study by the Observatório das Metrópoles and the city’s ObservaPOA, led by Luciano Fedozzi, reports “almost continuous growth” in assembly attendance from the start until 2002, when it peaked at 17,241 people in the year the city changed the cycle to a single round of assemblies.⁠³

Decline and partial recovery (2002 to 2012). After 2002 attendance fell until 2006, recovered from 2007, and reached 16,721 in 2012, still below 2002.⁠³ Fedozzi and colleagues attribute the fall in part to “the feeling that participation was not effective” as a way to get prioritized demands met, because from 2000 the city faced a financial crisis that led to investment plans not being carried out.⁠³ The 2004 change of party in city government, analyzed in the comparative overview, sits in the middle of this period.

Record attendance and backlog (2015 to 2017). The Jornal do Comércio reports, from city figures, that the attendance record was set in 2015, with more than 20,000 people.⁠² At the same time, unbuilt projects had accumulated. In October 2019 the city’s news service reported a city official’s statement that “in 2017, the OP had 2,395 unexecuted demands, which generated a liability of R$1.5 billion in unfulfilled promises” (translation from Portuguese).⁠⁴

No new money and a pause (2018 to 2022). A fact-check by a journalism project of the Federal University of Rio Grande do Sul (UFRGS) found that “in 2018 and 2019, not a cent was budgeted for the OP” (translation), which the city attributed to lack of funds, and that between 2013 and 2017, under the previous administration, R$494.3 million was budgeted for the OP but only R$4.2 million, or 0.8%, was actually spent.⁠⁵ The city reported in August 2022 that regional and thematic assemblies “had not been held since 2019” and would resume on October 31, 2022, with 23 meetings: 17 regional and 6 thematic.⁠⁶ The COVID-19 pandemic fell within this period.

Restart (2022 to 2025). Assemblies resumed in late 2022.⁠⁶ The Jornal do Comércio reports that assemblies were canceled in 2024 because of the floods that struck the state, that 2025 was a year of resumption, and that by July 29, 2025, with 14 of 23 assemblies held, more than 9,000 people had taken part.⁠² The same report gives an average attendance of 11,918 for 1989 to 2022.⁠²

How it worked

Porto Alegre’s OP is an annual cycle of assemblies and elected bodies, run by the city executive rather than created by a national law.

Geography and themes. The city organizes the OP in 17 regions and 6 thematic areas: education, sport and leisure; transit and mobility; housing and urban and environmental planning; economic development, taxation, tourism and work; culture and youth; and health and social assistance.⁠¹

Assemblies. In 23 regional and thematic assemblies, ordinarily held in July and August, residents set investment priorities, elect councillors, and determine the number of delegates for their forums.⁠¹ City departments attend to explain the criteria that guide the process and whether demands are feasible.⁠¹ The 2025 assemblies, as reported, elected priority themes, delegates and councillors.⁠²

Forums and council. Delegates and councillors meet every 15 days in regional and thematic forums, and a body of 92 councillors meets every two weeks to deliberate on regional, thematic and citywide matters.⁠¹ The detailed allocation formula is set in the OP’s internal regulations, which the city publishes;⁠¹ this page does not summarize them because the text was not reviewed for this case.

Authority. The OP shapes the investment plan proposed by the executive, but the city’s budget is enacted through the ordinary municipal budget process, and execution depends on the executive. The UFRGS fact-check’s contrast between amounts budgeted and amounts spent.⁠⁵ shows the gap that this arrangement leaves open.

Digital participation. By 2019 the city also ran an online channel; its news service reported 15,418 votes and 2,599 registered users on the “OP Digital” platform in that cycle⁠⁴

Measured results

Porto Alegre’s OP has been measured on attendance, on who attends, on participants’ views of their own power, and, in broader Brazilian studies, on outcomes.

Who takes part

Fedozzi and colleagues compiled participant surveys from 1993 to 2012.⁠³ Their findings, in their words or translated:

• Income. In 2009, 52.5% of assembly participants had family income of up to two minimum wages, “the largest proportion of this income segment” across the surveys; 28.7% had two to four, and 17.6% four or more; the share above eight minimum wages was about 5%.⁠³
• Sex. Women’s share rose from 46.7% in 1993 to 61.1% in 2012, above the roughly 55% share of women in the city’s population aged 16 and over.⁠³
• Schooling and age. Primary education was the most common level, people with higher education were underrepresented, and young people were underrepresented.⁠³
• Race. Participants were mostly white, followed by Black participants, whose share increased over the survey years.⁠³
• Renewal. On average about 42.4% of participants in surveyed years were new, from 48.6% in 1995 to 31.7% in 2005.⁠³

Who gets elected

Representation inside the OP was less equal than attendance. In 2009, participants with family income up to two minimum wages were associated with not being elected to OP bodies, while those with four to eight minimum wages were associated with being elected as councillors; the authors conclude that income, like education, “is a factor associated with conditions of inequality in the exercise of representation functions” in the OP.⁠³

Do participants think they decide

Asked “In your opinion, does the population really decide on works and services in the OP?”, more than half of participants in surveys from 1995 to 2009 answered “always or almost always,” but that share fell in 2005 and 2009, about one third answered “sometimes” in those years, and the authors conclude that opinion about decision-making power “has become more pessimistic”.⁠³

Outcomes beyond Porto Alegre

Research on Brazil more broadly links well-designed participatory budgeting to social outcomes. Wampler and Touchton, in the European Journal of Political Research (2018), studied 114 Brazilian municipalities with PB programs from 2009 to 2016 and found that programs “are associated with lower infant mortality rates when they broaden participation,” expand deliberation, and are embedded in ongoing policy-making venues.⁠⁷ That finding concerns PB design across many cities, not Porto Alegre specifically, and is an association, not proof of causation.

Failures and criticisms

Porto Alegre’s OP has been criticized from the left, the right and the center, often with the same facts.

Promises outran money. Fedozzi and colleagues argue that “the sustainability of participation” in co-governance “depends heavily on the degree to which” collectively agreed demands are resolved, and that financial capacity and governmental political will are the key variables (translation).⁠³ When the city could not build what assemblies chose, participation and renewal fell.⁠³ Critics who favor fiscal discipline read the 2,395-demand backlog;⁠⁴ as evidence that the process generated commitments the city could not afford. Defenders read the 0.8% execution figure for 2013 to 2017⁠⁵ as evidence that administrations simply chose not to fund the OP.

Each administration blamed the other. The R$1.5 billion “liability” figure comes from an administration explaining why it reset the OP⁠⁴ the UFRGS fact-check found that the same administration budgeted nothing for the OP in 2018 and 2019.⁠⁵

Participation without equal power. Attendance skewed toward lower-income residents, which supporters see as the OP’s great achievement.⁠³ But election to OP bodies skewed toward higher-income and more educated participants,⁠³ which critics on the left see as a reproduction of inequality inside a redistributive institution.

Fragmentation. Fedozzi and colleagues note the “low or almost nonexistent integration” of the thematic assemblies with the city’s other participatory bodies, such as municipal councils on the same policy areas, and suggest the fragmentation may hinder the quality of participatory democracy (translation).⁠³ They also note that thematic assemblies increasingly worked like regional ones, centered on specific demands.⁠³

Parochialism and short-term focus. An assembly system built around neighborhood demands tends to favor visible local works over citywide or long-term investments. The 2025 reporting describes participants as mostly bringing specific infrastructure needs.⁠² Whether that is a flaw or the point of the design is disputed.

What transferred and what did not

Porto Alegre’s model spread across Brazil and the world, but usually in modified form. The national pattern of adoption and decline in Brazil is covered in the comparative overview. Outside Brazil, cities adopted PB with smaller, capped budgets and voting rather than assembly negotiation, as in New York City, Paris, Madrid and Reykjavik.

What transferred most easily was the idea of residents proposing and choosing projects. What transferred least was the political setting of the early years: a governing coalition that treated the OP as central to its program, and assemblies with a large say over a city’s investment plan. Wampler and Touchton’s finding that outcomes depend on broad participation, deliberation and institutional embedding.⁠⁷ suggests that imported versions without those features should not expect the same results.

Lessons for My Vote Forward

These are interpretations and design proposals.

• Delivery is the participation strategy. Porto Alegre’s attendance and renewal tracked whether approved works were built⁠³ Any process that publishes priorities should also publish what happened to them; see proposals and participation over time.
• Separate “chosen” from “funded” from “built.” The gap between 2,395 approved demands, budgets of zero, and 0.8% execution.⁠^4,5 is exactly the kind of status difference a public record should make visible.
• Measure who participates and who represents. Porto Alegre’s surveys found inclusive attendance and less inclusive representation⁠³ Both should be measured; see measurement and representation.
• Ask participants whether they think they decide. The declining sense of decision-making power;⁠³ is an early warning signal any evaluation should track; see evaluation.
• Priorities are not budgets. Porto Alegre allocated real money. My Vote Forward’s priorities are information, not allocations; see tradeoffs.

Unresolved questions

• How much of the 2,395 demand backlog⁠⁴ was later built, canceled or re-prioritized? The city has reported numbers in political contexts, and no independent audit was found for this page.
• Who participates in the restarted OP after 2022? No profile study after Fedozzi and colleagues’ 2012 data⁠³ was verified.
• How are funds now allocated among the 17 regions, and how much money is at stake in recent cycles? The internal regulations and recent investment plans were not reviewed.
• Did the online OP channel⁠⁴ change who took part? Not established.
• Is Porto Alegre’s own OP associated with health or other outcomes, as studies of Brazilian PB in general suggest⁠⁷? Not established for the city alone in sources consulted.