Learning from Decide Madrid

Originally published in The GovLab blog on March 28, 2022

In 2021, the Madrid City Council commissioned The GovLab to conduct a data-driven analysis of Decide Madrid (DM), the City’s main public engagement platform. DM allows individuals to submit legislative proposals to the City Council, participate in online discussions, contribute to consultations posed by the City, vote directly on policies, decisions or actions of the City government, and participate in Madrid’s annual participatory budgeting program, to which the City allocates €50–100 million annually.

While the DM platform originally attracted a great deal of attention, with nearly 400,000 registered users in the first three years, only one of the 28,000 proposals submitted by residents through the legislative proposals feature has been implemented by the city since the platform’s launch in 2015. Critically, the lack of outcomes from residents’ proposals may discourage individuals from engaging with the platform.

This blog post explains our recommendations for how Madrid can improve the platform, strengthen the diversity of who participates, and create new opportunities for engagement, which also serves as guidance to other cities looking to improve their own engagement platforms and processes.

The City Council of Madrid should:

Example of a resident’s proposal calling for a ban on advertising on public infrastructure. Source:https://decide.madrid.es/proposals/29132-sancionar-la-publicidad-no-permitida-en-la-via-publica

Achieving representative engagement is a common challenge for participatory democracy initiatives, particularly those that allow participants to self-select. Our analysis of DM user data found that younger people (under 30), older people (over 65), and people who identify as female are underrepresented compared to their respective shares of the city’s population. Having robust evidence to understand these disparities can help indicate how they can be overcome.

Madrid, and other cities, should conduct surveys and interviews with individuals who are registered on their public engagement platforms, former participants who have dropped out, and people who do not participate to understand what motivates, as well as what discourages, people from engaging. In the interest of “meeting people where they are” we need to go beyond data about registration to understand differences in howdifferent groups experience, navigate, and benefit from the engagement process.

Engagement processes which require public servants to process large amounts of text, such as public comments or legislative proposals, can create redundancies and inefficiencies that make it difficult for public servants to keep up with the flow of information. On the DM platform, users often submit legislative proposals that duplicate existing proposals made by others. For example, there are numerous proposals submitted within the last few years calling for stricter enforcement of fines for people who do not clean up their pets’ droppings. The presence of multiple similar proposals may confuse users or cause them to split their support among these proposals, making it difficult for any one idea to garner the support it needs to advance through the process.

Cities should experiment with machine learning tools to more efficiently group, analyze, and draw insights from public proposals and comments and deliver feedback. For instance, the Belgian company CitizenLab created a text analysis tool that clusters comments and proposals submitted by citizens into common topic areas, making it easier for officials to quickly understand the public’s ideas and priorities. A similar approach could help Madrid and other cities with similar platforms to make use of public comments and proposals and enhance the relevance and usefulness of the platform for policymaking and governance. For example, the large volume of proposals about pet droppings could indicate that better enforcement and disposal of animal waste is a priority for some Madrid residents, which could in turn spark the creation of new policies or programs to address the problem.

Example of data visualization for topic clustering on CitizenLab’s natural language processing tool. Source:https://www.nesta.org.uk/feature/ai-and-collective-intelligence-case-studies/citizenlab/

Even the most enthusiastic city residents, generally, are not experts in proposing and writing legislation. Legislative proposal platforms need to guide residents through the process of identifying solutions that are within the jurisdiction and scope of what the city government can reasonably implement, or they risk wasting residents’ time and energy crafting proposals that have no chance of being implemented, and civil servants’ resources processing these proposals. In Madrid, the city government is not permitted to provide feedback on legislative proposals, resulting in many poorly formulated and unimplementable submissions, including proposals outside of the city’s jurisdiction (like constructing a new airport in the municipality of Torrejón de Ardoz).

Cities should create programs where lawyers, paralegals, and law students can lend their time to provide legal analysis and feedback of proposals. For example, in Brazil, residents use a platform called Mudamos to propose laws and policies, whereby a volunteer legal team guides users on whether their proposed law meets the constitutional requirements for a citizens’ initiative bill and whether it can be published on the platform. For Madrid and other cities, providing similar legal analysis could help to weed out legally impossible proposals and to strengthen the quality of legally feasible proposals.

Like many cities, Madrid primarily collects input from businesses, non-governmental organizations, and other groups through mechanisms separate from its public engagement platform, such as stakeholder meetings. With the exception of the participatory budgeting program, the DM platform is designed as a way for individuals, rather than businesses or organizations, to contribute their expertise and opinions.

The City should create a Groups feature that would allow individuals to create and join groups around their shared interests. Similar to Facebook Groups or the “Groups” feature on Athens’s citizen engagement platform, synAthina, teams could create a page with a description of their purpose and work and optional links to their website and social media pages. Individuals could search for teams by keyword and filter teams by topic area and district, offering people with similar interests a chance to collaborate with one another. For example, the individuals who have created separate proposals calling for more parking along Calle de Argumosa and in the Puerta del Angel area might decide to team up to create and promote a single proposal calling for more public parking across the city, or to advocate for such action offline. A Groups feature would also create an incentive for more organizations to participate, as the platform could provide them with an opportunity to connect with individuals who are interested in participating in actions they care about.

DM could be used to run a City Challenge as an additional public participation opportunity for Madrid residents. In a City Challenge, cities work with residents to codesign and collaboratively implement solutions to urban challenges, such as climate change and sustainability, public health and equity, congestion and pollution, structural racism and civic engagement. The GovLab has organized past City Challenges in partnership with the cities of San Pedro Garza García, Mexico; Barcelona, Spain; Oakland, California; and in multiple cities simultaneously across Mexico and Africa, providing training for participants and public servants to strengthen their capacity to implement impactful and sustainable projects. A City Challenge would give Madrid the opportunity to more rapidly develop and implement solutions to priority problems identified by residents, while also building capacity among civil servants to support and sustain those projects.

Successfully engaging youth is a common challenge for public engagement projects. Too often, cities design engagement initiatives without regard for existing channels that youth use to make their opinions heard (such as through online activism), neglect to create outreach materials that appeal to young people, or design processes whose rules exclude young adults altogether. In Madrid’s case, for example, individuals under age 16 are not eligible to participate in the City’s participatory budgeting initiative.

Madrid, and other cities with participatory budgeting programs, should provide a portion of this annual funding to projects proposed and voted on by youth, following the lead of other cities such as Vienna’s pilot €1 million participatory budgeting initiative called Dein Wien for Future (“Your Vienna for the Future”) that allocates €1 million for projects that are proposed and voted on by for people ages 5 through 20 or Helsinki’s €150,000 participatory budgeting for people ages 12 through 17 who participate online through a City-run platform built upon the Decidim framework.

Screenshot of the online platform for Vienna’s youth participatory budgeting initiative. Source:https://www.junges.wien/domain/1338

A citizens’ assembly is composed of randomly selected and representative citizens, who are convened to deliberate and make recommendations on a specific issue. Random representative sampling as well as the ability to design citizens’ assemblies to oversample for underrepresented groups, means that decisions made by the assembly are more representative of community sentiment, and a diversity of viewpoints, and thus are more legitimate.

While Madrid briefly experimented with a citizen assembly and created a new participatory body for stakeholder and interest groups in 2020, a well-designed citizens’ assembly could serve as a more effective, diverse, and legitimate mechanism for participatory lawmaking.

There are a number of existing citizens’ assembly models in place at the city, state, and regional level throughout Europe that Madrid and other cities could build upon and learn from. In 2021, Barcelona created Forum Jove BCN, a citizens’ assembly of 99 young people aged 16–29 who made recommendations to the Barcelona City Council regarding policies that impact young people in the city. Belgium has also implemented a citizens’ assembly of 45 randomly selected members of the public who work with members of parliament to design new policies around topics including the regulation of 5G infrastructure, the re-accommodation of homeless people, and the role of citizens in times of crisis. The City of Paris inaugurated its permanent citizens’ assembly in November of 2021, composed of 100 Parisian citizens that deliberate and issue recommendations to the City Council and draft laws, among other powers.

As described in greater detail in our analysis report, the use of the DM platform to submit legislative proposals is decreasing each year. There is an urgent need to radically rethink the design of the process, using lessons learned from impactful legislative proposal platforms from around the world.

A number of countries — including Finland, Estonia, Latvia, Iceland and Brazil — have implemented online platforms at the national level that enable citizens to submit legislative proposals and sign-on to proposals submitted by others. Several of these platforms have achieved higher and more sustained participation that the DM proposal feature currently exhibits. Finland’s Ministry of Justice maintains an online platform where citizens can initiate and sign petitions under the country’s citizens’ initiative law. The country’s parliament must consider any initiative which receives at least 50,000 signatures. Since the citizens’ initiative law went into effect in 2012, more than 1,000 petitions have been proposed, 56 of which have collected at least 50,000 signatures, including a proposal to legalize same-sex marriage, which parliament passed into law in 2017. Estonia’s government runs a similar platform for its residents to propose and support policy ideas at the national or city level. More than 50 proposals have reached the signature threshold necessary for consideration (1,000 signatures at the national level or 1% of the population for city-level proposals) in just the last two years. Studying these and other existing examples could provide additional lessons for how the Proposals feature of DM could be improved in order to expand engagement and impact.

The recommendations and examples above illustrate how cities are using new platforms and processes to enable more effective, equitable, and impactful collaboration with residents. Madrid and other cities should continue to experiment with such new forms of engagement that can provide the infrastructure needed to innovate and to solve emerging problems in new ways.

To read these recommendations in their entirety, visit:http://files.thegovlab.org/madrid-report-english.pdf

For more information about the design and development of Decide Madrid, read ourcase studyand previousblog series.

To learn more about how cities are using new technologies to enable better collaboration with the public in lawmaking, visit:https://congress.crowd.law/