Case study

Participatory Budgeting in Porto Alegre

Citizens directly allocated a share of the city's investment budget through open neighbourhood assemblies, and the model was later copied in more than a thousand cities worldwide.

BrazilMunicipalityParticipation & deliberation1989

01 - What happened

In 1989, Porto Alegre’s newly elected city government handed a portion of its municipal investment budget to open assemblies held in every district of the city. Residents debated and voted on which streets got paved, which neighbourhoods got sanitation, and which schools got built first. The administration kept the technical and legal authority to execute projects - but the priorities were no longer set behind closed doors.

The process ran in yearly cycles: regional assemblies elected delegates, delegates negotiated a citywide list of priorities against a technical needs assessment, and the following year’s budget followed that list. It was slow, repetitive, and by most accounts genuinely boring to sit through. It also worked: infrastructure spending shifted measurably toward the poorest districts, and independent researchers credited the process with improvements in sanitation coverage and school enrolment through the 1990s.

What made it travel - first to over a hundred other Brazilian municipalities, later to cities across Latin America, Europe and Africa - was that it was legible. It didn’t require a new law, a new ministry, or new technology. It required a mayor’s office willing to give up some of its own discretion, and a population willing to show up.

02 - Preconditions

What had to be true first.

A newly elected municipal government (the Workers' Party, 1988) willing to spend its first term on process, not just visible projects.
An existing tradition of neighbourhood associations and community organising to build the assemblies on, rather than inventing civic infrastructure from zero.
A city small enough, and a civil service motivated enough, that a multi-year rollout across dozens of districts was administratively possible.
Real budgetary authority delegated to the process - assemblies decided actual investment spending, not an advisory shortlist.

03 - What went wrong

Told straight, not as a footnote.

Participation was never representative: poorer, more organised districts turned out consistently, while the middle class largely opted out, skewing whose priorities got funded.
Turnout plateaued and then declined through the 2000s as the initial novelty wore off and meetings became routine.
When the Workers' Party lost the mayoralty in 2004, the incoming administration scaled the process back, showing how dependent it was on sustained political will rather than being self-sustaining.
The model demanded a serious ongoing time commitment from participants, which quietly excluded people with the least flexible schedules - often the people the process was meant to reach.

04 - Transferability

Would this work where you are?

Rated 0–3 on each dimension. A low score is not a disqualification - it is a list of what to check before you try.

Context dependence

Resources required

Causal simplicity

Institutional substitutability

Value consensus

Population similarity

Talk to someone who was there

This case has a practitioner willing to talk.

Reach out with what you are trying to build where you live, and we will make the introduction.

Get in touch