Residents proposed it. Did the city build it?
Cities from Gdańsk to Barcelona ask residents to decide, through citizens' panels, participatory budgets and platforms like Decidim. Far fewer check what happened next. Warsaw is one that did: the official in charge of its climate panel expected to have delivered about 63% of the recommendations by the end of 2025, and almost nobody knew. What makes a city follow through, and who turns up in the first place.
Warsaw delivered, and nobody knew
At a forum for participation practitioners in Łódź, an official from Warsaw said that his city was carrying out what its residents’ climate panel had asked for. The room was lost for words. Among the people who run citizens’ panels in Poland, the reporter explained, the received wisdom is that panels are very good at finding out which changes people will accept, and that afterwards nobody acts on them.
The official was Jarosław Holwek, chief specialist in Warsaw’s Office of Air Protection and Climate Policy, whose team is responsible for putting the panel’s recommendations into practice. Asked by OKO.press in July 2025, he put a number on it.
To be precise: by the end of this year we will have delivered about 63% of the Panel’s recommendations, if everything goes to plan. It is hard to say whether that is a little or a lot.The panel met online in late 2020, in the middle of the pandemic, on one deliberately narrow question: how to raise the city’s energy efficiency and its share of renewable energy. Residents drawn by lot to mirror Warsaw by age, district, education and sex sat through two sessions with experts and two of debate. Their first list ran to 90 recommendations; more debate and voting cut it to 49.
The chart sorts that final list the way the city printed it. Look at the bottom two bars: most of what the panel asked for was new work.
The final list as OKO.press printed it, in two headed parts. The first-list figure is the official’s recollection; a NetZeroCities account of the same panel gives 93. Source: OKO.press, interview with Jarosław Holwek (22 Jul 2025)
Some recommendations proved hard. Two asked for green public procurement, which Holwek calls a revolution in a system whose law favours the lowest price. Some of the 21 were already moving: new air-quality sensors around the city came from another process, he says, and the panel simply gave it more weight.
When the interview took place, delivery stood at about 55%. Some of it you can see: solar panels spreading over the roofs of city buildings, including a tram depot on Woronicza Street, and free energy advisers for anyone renovating a block. One of the panel’s biggest results, Holwek says, is the Warsaw Green Building Standard, binding on new municipal buildings from 1 July 2025. “When citizens told us they wanted this, we had an obligation to deliver,” the office’s deputy director, Jacek Kisiel, told NetZeroCities.
And almost nobody knew. A panellist at the Łódź debate had not known that her work had come to anything. The city publishes implementation reports on its website and posts on social media, Holwek said, but it could not write to the panellists themselves: they had handed over their personal data for the panel only, and data-protection rules stopped the city using it afterwards. “Maybe that was a mistake.”
A city delivering more than half of a residents’ list, and telling only those who went looking, should not be news. Why is it? The answer starts where the Polish model began: in Gdańsk, after a flood.
Binding on a promise
On 14 and 15 July 2016 a cloudburst hit Gdańsk. The city’s drainage company estimated that some 40 million cubic metres of rain fell on the city in 12 hours. Water tore through the Wrzeszcz district, and two residents died.
By the autumn the city had convened Poland’s first citizens’ panel, on how to prepare Gdańsk for heavy rain. The idea came from outside City Hall. Sixty-three residents drawn by lot met over three Saturdays and a Sunday, heard experts and voted on proposals. The rule that made their votes matter took one line of the city’s report: recommendations were binding “on the basis of the mayor’s informal declaration”, as with the participatory budget, provided 80% of panellists agreed with them strongly or somewhat.
The votes show how the line worked. Most proposals cleared it easily; the reservoirs the city’s own drainage company had proposed did not.
Eight of the proposals the first panel voted on, autumn 2016: the share who strongly or somewhat agreed, of 46 or 49 valid votes. Below the line a proposal was advice; above it, the mayor had promised to act. Source: City of Gdańsk, Raport podsumowujący panel obywatelski (2017)
The model travelled fast. Gdańsk ran two more panels in 2017, Lublin followed in 2018, and after the climate strikes of 2019 activists pressed for panels in Poland’s largest cities. Researchers at Adam Mickiewicz University call it a short-lived but intense boom.
- 14–15 Jul 2016[source]
Gdańsk floods
About 40 million m³ of rain in 12 hours; two residents die.
- Autumn 2016[source]
Poland’s first citizens’ panel
Gdańsk asks 63 residents drawn by lot how to prepare for heavy rain. Recommendations with 80% support bind the mayor by his informal declaration.
- 2018[source]
Lublin
The first city to follow Gdańsk, asking what it takes to breathe clean air.
- Autumn 2020[source]
Wrocław and Warsaw
Wrocław on mobility and climate, Warsaw on energy efficiency and renewables; both requested by climate activists.
- Spring 2021[source]
Poznań and Kraków
Poznań on forests, green space and household coal; Kraków on cutting energy use.
- Spring 2023[source]
Kraków and Łódź again
Kraków on sustainable transport; Łódź on cutting its greenhouse-gas emissions by 55% by 2030.
Climate-related citizens’ panels in Polish cities, all commissioned by the city. Most climate panels after 2019 were requested by climate activists.
Warsaw’s mayor later made the same kind of promise: recommendations that cleared the same bar would be carried out. The 13 activists the Poznań researchers interviewed saw the threshold, which they say became the Polish standard, as the panel’s great strength: a list backed by a random cross-section of the city is hard to dismiss as the view of a self-selected group.
The disappointment came later. No law or procedure compelled any city to act, even on recommendations that cleared the bar, and for some it came as a shock that a mayor’s public declaration was not enough. Without dedicated individuals, the activists said, implementation often stalled once the panel was over. Only in Warsaw did they recall sitting on teams working through the results, and even there they credited personal commitment, not any institutional arrangement.
An informal promise lasts as long as the people who made it, and the officials below them, choose to keep it. Which leaves a practical question: how would anyone outside City Hall know whether it was kept?
How do you tell whether a panel mattered?
Nobody yet agrees what “implemented” means. Holwek describes sitting down with Warsaw’s monitoring team, officials and residents’ organisations together, to settle when a recommendation can count as done, because some were worded so loosely that “it is not really known what would count as delivering it”.
A team of researchers writing in Frontiers in Climate has tried to give cities a common yardstick. Reviewing existing approaches, they found 94 evaluation criteria and indicators and arranged their own framework in four stages: input, process, outcome and impact. The last two are where Warsaw and Gdańsk sit. Outcome indicators count the policies that refer to an assembly’s recommendations within a set time, and ask whether the references are substantive or symbolic. Impact indicators include the adoption rate of recommendations and the resources set aside for them, checked against climate plans, budgets and interviews with officials. Tested on assemblies in Catalonia, Edermünde and Riga, the framework met the same recurring problem: making sure policymakers followed up.
Read against that list, Warsaw’s figures are a start, not an audit. They are the city’s own count, given in an interview, and there is no common definition behind them. Gdańsk’s report records every vote but, written in 2017, not what came next.
What Warsaw does have is a sequence any city could copy and publish. Read the timeline as the beginnings of a delivery log.
- 2019[source]
A climate office is created
The Office of Air Protection and Climate Policy is set up; it will own the panel’s recommendations from day one.
- 2020[source]
The panel votes
Residents drawn by lot agree their list online, during the pandemic. A monitoring team of officials and residents’ organisations is set up once the panel ends.
- 2023[source]
A team of directors
Heads of the departments that matter most for the climate goals meet to unblock work stuck between departments.
- 1 Jul 2025[source]
Green building standard in force
New municipal buildings must follow the Warsaw Green Building Standard, one of the panel’s asks.
- Jul 2025[source]
About 55% delivered
The official in charge expects about 63% by the end of the year, if everything goes to plan.
- Sep 2025[source]
Next report, after the monitoring team meets
The city plans to publish its next progress report once the monitoring team has discussed it.
Every step comes from the interview with the official in charge. The percentages are the city’s own estimates, not audited.
Each entry is a date, an owner or a count. Put them on one public page, updated after every monitoring meeting, and residents would not need a journalist to find out.
Holwek’s other lesson is about the list itself: keep it to 10 to 15 recommendations at most, each worded precisely and measurably, with a defined end. A citizens’ panel is the most complicated and costly tool a city has for asking residents, OKO.press notes: a representative sample, expert teaching and several rounds of debate. Counting what happened next costs a table and a meeting.
Platforms: what does the software change?
Software has changed how many people a city can ask. Decidim, the open-source platform that grew out of Barcelona, counted more than 500 installations in its 2025 activity report; why it outlived so many civic tools is told in our read on pilots that outlast their grants. Berlin runs its own, meinBerlin. An installation count says nothing about what was decided on it, though.
Germany gives the widest view of what the software is used for. In mid-2022 researchers searched the website of every German district and every city of more than 5,000 people and found 3,828 digital participation processes; 1,156 of 2,057 cities had run at least one. Almost 40% of the methods were surveys, and formats in which residents could actively shape a plan, by voting or designing, were very rare. The study counts what was offered, not who took part.
Count the ideas that reach a decision and the funnel narrows sharply. The chart follows two participatory budgets run on Decidim in Switzerland, from ideas submitted to ideas funded.
Two trial participatory budgets: one run by two civic associations with city money, one run top-down by the area’s management. Funding was not the end: Wipkingen’s winners then needed municipal approvals. Source: Suter et al., Urban Planning 9 (2024)
Funding was not the finish line. In Wipkingen, the winning projects ran into municipal approvals, for events for instance, because the city had no legal framework for fast-tracking ideas the public had chosen.
Newer tools aim at the conversation itself. In January 2023 Helsinki’s planners brought UrbanistAI, a generative-AI design tool, into two workshops for their Summer Streets project, one with a citizens’ committee of 15 and one with local business owners. Groups turned descriptions into realistic images of their street, voted on them, and the favourite ideas were built into that summer’s street design. In Rotterdam, a study of the city’s urban digital twin asks whether residents have any right to it, and examines the limits built into how they can take part.
The platform’s most useful job may be the dullest. Vendors told the authors of a 2023 report on Europe’s democracy-technology market that the rising number of projects had created a need “to keep track and learn from past participative projects”. Decidim’s own social contract asks that everything on the platform be accessible, downloadable and traceable. A platform that keeps a public record of what was proposed, chosen and built is the delivery log Warsaw lacked. The harder question is who writes into it.
Who turns up?
Louvain-la-Neuve should have been easy. The Belgian university town has advanced digital infrastructure, a highly educated population and an active civil society. Yet its first large-scale digital participatory budget drew about 4% of citizens, according to the municipal officer in charge.
Researchers spent the autumn of 2025 asking why, in 12 interviews and two focus groups. Students had the skills but little reason to care: the city, a student representative said, is “a backdrop for them, not a community they feel part of”. The town’s founding generation had the motivation but not the skills; asked to vote online with a digital identity app, “they are lost”, said the head of the municipal digital help service. The few hundred votes came overwhelmingly from older, already-engaged residents. Above all, people doubted their input counted. One had written a detailed comment on a planning consultation and watched the project approved exactly as first presented. “Why would I participate again?”
In Lucerne’s LuzernNord, a district with many migrants, the participatory budget’s flyers and website were in German only, and residents over 50 said they preferred analogue channels.
Gdańsk shows what design can do. Instead of waiting for volunteers, the city posted 8,968 invitations to people drawn at random from the voter register, and 896 signed up. Compare them with the city they came from.
Shares of the city’s adults on the voter register (October 2016) against the 896 people who registered after the random mailing. A second draw by profile then filled the 63 seats to match the city exactly. Source: City of Gdańsk, Raport podsumowujący panel obywatelski (2017)
Every age group landed within a few points of the city’s own profile, and a second draw, by district, age, sex and education, closed the rest. Panellists were paid a 600 zł allowance for the panel’s days.
UN-Habitat makes the general point in its review of digital participatory budgeting: digital technologies can reduce the barriers to taking part, but also raise them for some. The European cities it studied, among them Paris, Madrid and Lisbon, ran their platforms alongside workshops and face-to-face meetings.
Inclusion is designed, not assumed: a letter, a lottery, an allowance, a translation, someone to help with the app. The Louvain-la-Neuve researchers add the step that ties it back to delivery: show people, systematically and after every process, that what they said changed something.
Make it land
Three things, each already tried somewhere in this read, decide whether a residents’ list gets built.
- A threshold that binds. Gdańsk’s 80% rule, and the Warsaw mayor’s matching promise, gave officials an argument against inertia. When the team met resistance inside City Hall, Holwek recalls, they would say that “this is what residents decided in the panel, and the mayor promised we would deliver it, so we have to do something about it”. It worked, he says. Write it into a council resolution, not only a speech.
- A named owner. Warsaw knew from the start which office would carry the list, and later added a team of directors to unblock work that crossed departments.
- A public delivery log. A short list of measurable recommendations, a status for each, reviewed with residents’ organisations and published after every review, ideally on the platform where the ideas were gathered.
None of this is expensive next to the panel itself. The Citizen Participation & Civic Tech topic collects the platforms, budgets and assemblies behind this read, and Warsaw’s profile shows the city around its panel.
If your city has run a panel or a participatory budget and counted what followed, or tried and stopped, the Hub wants to hear about it: take a minute to tell us.
References
From our libraries16
- Warszawa. Jak się wdraża zmiany wymyślane przez mieszkańców. I dlaczego nikt o tym nie wie — Knowledge
- Warsaw Green Building Standard (WGBS) — Knowledge
- Building for the Future: How Warsaw's Green Building Standard is changing the city — Knowledge
- Raport podsumowujący panel obywatelski w Gdańsku (Urząd Miejski w Gdańsku, 2017) — Knowledge
- Climate Citizens’ Assemblies in Poland: Activists’ Nuanced Perspectives — Knowledge
- Assessing the effectiveness of climate assemblies: framework for measuring deliberative impact — Knowledge
- Decidim — Organisation
- Decidim Activity Report 2025 — Knowledge
- meinBerlin — Project
- Digital and multi-channel citizen participation in Germany: A comprehensive overview of patterns, methods and determinants — Knowledge
- Digital Rights to the City: Local Practices and Negotiations of Urban Space on Decidim — Knowledge
- Reimagining Helsinki: Participatory Urban Planning with Generative AI — Knowledge
- Right to the Digital Twin City? Citizen Participation and Limits-by-Design in Rotterdam's Urban Digital Twin — Knowledge
- Democracy Technologies in Europe: Online Participation, Deliberation and Voting — Knowledge
- Beyond access: social, institutional and cultural barriers to digital citizen participation in a well-equipped European City — Knowledge
- Innovation and digital technology to re-imagine Participatory Budgeting as a tool for building social resilience — Knowledge
Further sources3
- Jarosław Holwek, chief specialist, Office of Air Protection and Climate Policy, City of Warsaw, July 2025 — Jarosław Holwek, chief specialist, Office of Air Protection and Climate Policy, City of Warsaw, July 2025
- City of Gdańsk, Raport podsumowujący panel obywatelski (2017) — City of Gdańsk, Raport podsumowujący panel obywatelski (2017) · 14–15 Jul 2016
- Pospieszna & Zielińska, Voluntas (Cambridge University Press, 2026) — Pospieszna & Zielińska, Voluntas (Cambridge University Press, 2026) · 2018