Decide Madrid: citizen proposals, a city vote and a short-lived sortition body

Decide Madrid is the City of Madrid’s online participation platform, launched in 2015, through which residents can submit and support citizen proposals, debate, and vote in participatory budgets and city consultations; its open-source software, later called CONSUL and now Consul Democracy, has been adopted by institutions in many countries.⁠^1,2 In February 2017 two citizen proposals that had gathered enough support went to Madrid’s first citywide citizen vote, in which more than 212,000 people took part.⁠³ In 2019 the city added a City Observatory of 49 residents chosen by lottery, which the next city government repealed in February 2020.⁠^4,5

What happened

Decide Madrid began with a change of city government and ended its first phase with another.

Launch (2015). Royo, Pina and Garcia-Rayado, in a peer-reviewed study in Sustainability (2020), describe Decide Madrid as the “award-winning e-participation initiative of the city council of Madrid,” launched in 2015.⁠¹ A new city government took office after the May 2019 municipal elections.⁠⁵

First citywide vote (February 2017). From February 13 to 19, 2017, the city held its “first citizen vote”.⁠³ It combined two citizen proposals that had reached the support threshold, “Madrid 100% sostenible” (submitted by the Alianza por el Clima platform) and “Billete único para el transporte público” (a single public transport ticket, submitted by an individual resident), with city questions on the redesign of the Plaza de España and on the Gran Vía, and district questions in six districts.⁠^3,6

Recognition and spread (2018 onward). Royo and colleagues report that Decide Madrid was one of the winners of the 2018 United Nations Public Service Award, and that its software “has been adopted or it is in the process of being implemented in more than 100 institutions in 33 countries”.⁠¹

The City Observatory (2019). On January 29, 2019, the city plenary approved the regulation of a City Observatory, published in the regional official bulletin on February 1, 2019.⁠⁴ In February 2019 the city announced it would send invitations to 30,000 randomly chosen households, out of 1.29 million in the city.⁠⁷ Forty-nine members and 147 substitutes were selected by lottery between February and March 2019, and the Observatory held three working sessions between March and October 2019.⁠⁴ At the first session on March 30, 2019, members ranged in age from 17 to 82.⁠⁸

Change of government and repeal (2019 to 2021). After the May 2019 municipal elections, a new governing coalition took office. The Open Government Partnership’s Independent Reporting Mechanism (IRM) reports that in November 2019 the new government began the process of rescinding the Observatory, “arguing that it was an inefficient model of citizen participation,” and that the council approved the repeal in February 2020.⁠⁴ On June 11, 2021, the council created a different advisory body, the Consejo Social de la Ciudad, built on representative economic, social, professional and neighborhood organizations.⁠⁴

How it worked

Decide Madrid offered several channels with different powers, and the City Observatory sat on top of one of them.

Citizen proposals. Any resident could post a proposal and collect support online. In 2019 the city explained that a proposal needed “27,662 signatures, 1% of the voter census,” within one year to trigger a citizen consultation, and described the mechanism as similar to a petition in which “signatures are gathered digitally and activate a process that ends in a citizen consultation” (translation from Spanish).⁠⁶

Citizen votes. The February 2017 vote was open to residents through three channels: 54% voted by post, 35% on the Decide Madrid website and 11% at in-person ballot boxes distributed across districts.⁠³ The city stated that the vote “did not require a quorum,” that is, no minimum number of voters.⁠³ Postal and ballot-box votes were counted in a public count at the Matadero Madrid venue.⁠³ The city presented the outcome as decisions of the public (“La ciudadanía ha decidido,” “the citizens have decided”);⁠³ the sources consulted do not state the results’ legal force.

Participatory budgets and debates. Decide Madrid also hosted participatory budgets and open debates; Consul Democracy lists proposals, debates, participatory budgeting and voting as its core functions.⁠² The budget amounts and participation in each Madrid budget cycle were not verified for this page and are listed under unresolved questions.

The City Observatory. Membership was open to people aged 16 or over registered as residents of Madrid, excluding paid elected officials, city employees and people convicted of intentional crimes; one person per invited household could reply, and a second lottery selected members and substitutes as a sample reflecting age, sex and geographic distribution.⁠⁷ Members were to meet at least eight Saturdays a year, received €65 per meeting, and could serve only one one-year term.⁠⁷ Its mandate, as the IRM summarizes it, was “to monitor municipal actions and policies, as well as analyze, amend and recommend action on citizen proposals” from Decide Madrid, with “strong (although not legally binding)” decision-making powers over those proposals.⁠⁴ The city also said the Observatory could propose calling citizen consultations.⁠⁷

Measured results

Decide Madrid’s measurable results include the 2017 vote, the threshold record and the Observatory’s brief activity.

The February 2017 vote

The city reported that more than 212,000 participants cast more than 950,000 votes across questions, with about 1,600 postal votes still to be checked.⁠³

Source: City of Madrid press release, February 27, 2017.⁠³ More than 83,100 people answered the four Gran Vía questions.⁠³

The city’s press release did not state a turnout percentage. As an interpretation only: if the census of residents aged 16 and over was roughly 2.77 million, which is what the 2019 threshold of 27,662 signatures at 1% implies,⁠⁶ then 212,000 participants would be under 8% of eligible residents. The 2017 census figure itself was not verified.

Proposals and the threshold

The two proposals voted on in 2017 are the ones the city itself cites as having reached the threshold.⁠⁶ In 2019 the city described “hundreds of proposals” still waiting to gather the required support.⁠⁶ The total number of proposals submitted and the number that ever reached 1% were not verified in an official count for this page.

The City Observatory

The IRM reports that the Observatory deliberated on whether to implement the most-supported Decide Madrid proposals and voted on priority issues for working groups on the environment and sustainable transport, cleanliness and maintenance of public spaces, and inequality between districts, and that its decisions were presented to city authorities.⁠⁴ The IRM concludes that while it “led to changes to government practice for the short time it was operational,” it “did not lead to a sustainable change in practice”.⁠⁴

Failures and criticisms

Decide Madrid and the Observatory were criticized by the government that replaced them, by researchers, and implicitly by the platform’s own numbers.

A threshold few could reach. A 1% threshold in a city of millions meant tens of thousands of verified supporters.⁠⁶ The city’s 2019 campaign to promote waiting proposals.⁠⁶ is itself evidence that the channel was not producing votes on its own. Supporters of a high threshold argue that it prevents small groups from forcing citywide votes; critics argue that an unreachable threshold turns a petition channel into a suggestion box.

Early enthusiasm, later problems. Royo and colleagues found that “the high expectations of citizens explain the high levels of participation in the initial stages,” but that “the lack of transparency and poor functioning of some of its participatory activities” were “negatively affecting its performance”⁠¹ They also found that individual and organizational factors, not technology, were the most relevant success factors.⁠¹

Low turnout and no quorum. The 2017 vote had no quorum.⁠³ Its results were lopsided, which made them easy to read, but critics of advisory city votes argue that a vote without a quorum lets a mobilized minority speak for a city. Defenders answer that more than 200,000 participants is a large number for a city consultation, and that the questions were published and counted in public.⁠³

The new government’s objections to the Observatory. According to the IRM, the incoming government argued that the Observatory was “an inefficient model of citizen participation,” that it did not fit the new government’s participation model, and that its discussions were “hyper specific,” focused on Decide Madrid proposals rather than the broader government strategy.⁠⁴ Ganuza and Menéndez-Blanco, in a peer-reviewed article, describe it as “the first permanent citizen participation scheme in which members are chosen by lot,” abandoned by the new government after the May 2019 elections.⁠⁵

Public consultation ignored. The IRM reports that in the public consultation held before the repeal, “the majority of citizens that participated supported the maintenance of the Observatory,” and that the council repealed it anyway.⁠⁴ Defenders of the repeal answer that a self-selected consultation is not representative either; critics answer that this is precisely what the lottery-selected body was meant to address.

Dependence on one coalition. Every element examined here was created by one governing coalition and changed by the next.⁠^4,5 Royo and colleagues’ emphasis on organizational and political factors;⁠¹ fits this pattern.

What transferred and what did not

Decide Madrid’s software traveled much further than its institutions. Royo and colleagues report adoption in more than 100 institutions in 33 countries as of 2020⁠¹ the Consul Democracy project’s own site states that the software “is and has been used by more than 250 cities and organizations all over the world” and lists Madrid among them.⁠² Because the software is open source and free to use and modify,⁠² cities could adopt it without adopting Madrid’s thresholds, its vote design or its Observatory.

What did not transfer, even within Madrid, was institutional permanence. The city’s poster for the Observatory said it was inspired by citizens’ assemblies in Canada, Australia and Ireland;⁠⁹ see Ireland’s citizens’ assemblies. Unlike Ireland’s assemblies, created case by case by national parliamentary resolution, Madrid’s was a municipal regulation that a new majority repealed within about a year.⁠⁴ Compare Reykjavik, where a similar idea platform survived changes of government, and Paris, where a participatory budget ran on a single mayor’s program.

Lessons for My Vote Forward

The following are interpretations and design proposals.

• Software is the easy part. Royo and colleagues found that organizational and individual factors mattered most.⁠¹ My Vote Forward should not treat its demonstration as evidence that any institution would use what it produces; see evaluation.
• Thresholds must be explained with their denominators. A 1% threshold sounds low until it is 27,662 people.⁠⁶ Any threshold in the framework’s design should be published with the absolute number it implies; see measurement.
• Report results with participation, every time. The 2017 vote’s percentages were striking, but the press release did not state turnout.⁠³ A published release should always give both.
• Durability needs a constituency beyond one government. Madrid’s Observatory was repealed despite consultation support.⁠⁴ The framework’s governance should plan for changes of leadership.
• Separate agenda-setting, deliberation and voting. Madrid combined open proposals, a lottery-selected deliberative body and an advisory vote. My Vote Forward’s separation of priorities, opinions and private voting mirrors the idea that different stages need different safeguards.

Unresolved questions

• How much money did each Madrid participatory budget allocate from 2016 onward, how many people voted, and how many winning projects were completed? Official figures were not accessible for this review.
• How many citizen proposals were submitted in total, and did any besides the two voted on in 2017 reach the threshold? An official count was not verified.
• What exactly was the participation rate in the 2017 vote, with the official census as denominator? The press release did not give it.⁠³
• What did the Observatory recommend in its three sessions, and did the city act on any recommendation before the repeal? The IRM describes priorities but not outcomes.⁠⁴
• How does the Consejo Social de la Ciudad created in 2021 compare in representativeness and influence? Not assessed in sources consulted.