Beyond the Code: Sustainability in Open Source

A project to study the sustainability of open source beyond technical spaces

#participativetech #opensource #democracy

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 744957

Table of content

What is Decide Madrid How this page is organised Timeline of Decide Madrid (Pre-2011 to 2011) The origins of Decide Madrid (Pause) The intellectual origins of Decide: Citizen participation and open source (2011 to 2015) LaboDemo: The ideas behind Decide are moulded (2015) Decide Madrid is implemented (2017) The roller-coaster of participation (2017-2019) Bringing Decide to the city

What is Decide Madrid

In particular, I study Decide Madrid, a Free and open source software on direct participative democracy. It is one of the most ambitious participative projects taken up by governments in recent times and so it's worth talking and reading about it.

Here are a few points why you should care about Decide Madrid:

  1. If you are interested in open source information systems for social innovation;
  2. If you are interested in future models of democracy that will likely become a staple of societies in the decades to come;
  3. If you're a public official implementing participative technologies in City Halls or government institutions.

You will find resources here talking about each of the above points, resources that I have found interesting in my years studying Decide Madrid. You will find, for example, an overview of some of the initiatives that particularly marked Decide Madrid and why they mattered most than others. There are many online resources that pay testimony to the complexity and ambitions of Decide Madrid, but these are often disparate. I try to tie them some together here. If you think there's something that should be in here, drop me a line.

By going through this page, my hope is that you can have a quick idea of the extremely ambitious objectives behind Decide Madrid, what problems they had to solve, and how they tried to solve it.

How this page is organised

A quick word about organisation. I'm really tired of multi-page, deep depth webpages, so this webpage is it. It's a summary of resources, so it will help me control the content creep that might take place.

Navigate through the table of content at the top of the page and the links at the beginning and end of each section to go back up to the table of content.


Timeline of Decide Madrid

I go through some of the most important moments of Decide Madrid.

(Pre-2011 to 2011) The origins of Decide Madrid

If you ask people when Decide Madrid was created, they will probably say in September 2015, which is the date the web site was launched.

* This is a good summary of many elements that went into the thinking when designing Decide Madrid. There are two other parts that continue to look at Decide Madrid's inception.But Decide Madrid wasn't born in a vacuum, the idea comes from somewhere, specifically the 15M.* The 15M was a testing ground for some of the participative technology that was being created at the time and became in the minds of some participants a key (if not the only) way to effect the democratic change they sought. Beyond urgent social demands, many participants thought that the cause of the problem was a lack of democratic implication on key political and societal questions. The reason for a lack of participation was historic: without information systems, it is difficult to scale recurrent citizen participation. This was not the case any more and indeed, many systems were developed to facilitate coordination and collaboration between the different assemblies. The 15M, thus, was the realisation that procedures could help right Spain's situation and prevent future such situations.

Decide Madrid is infused by another imaginary: that of open source. Some of its most prominent members were highly visible hackers before the creating Decide Madrid. Some had spent time in student groups disussing issues concerning open source. Others, had used open source technologies (e.g. Ruby language). Pablo Soto, one of the main proponents, had been taken to court by music majors for developing a p2p software programme. Between individual experience, through work, or within the 15M, many had felt and experienced the fruitful intersection between open source and society beyond its intellectual promise.

* I realise that (1) this trend may not be linear (e.g. past and current privacy violations that may encourage the exact opposite trend with rampant co-optation); (2) but I still think that, at least for the two last decades, open government principles have made some headway into the institutions are governed; (3) there is still a (much) long(er) way we can go in opening up organisations and institutions; (4) the participation of 'open' as an organisational vector propelled by fablabs and makerspaces during COVID-19 is a good recent example of its growing importance and capacity to participate.It is no coincidence that after the 15M, an important initiative grouped many Free Software and open source hackers with political theorists and sociologists. The purpose was to think about governance systems in an age where open principles are increasingly espoused.* This project, sponsored by Ecuador's plan of Good Living, had the suggestive name of Free/Libre Open Knowledge (FLOK) society. In Spanish, the project had an even more suggestive name: the “Good Knowing”. They organised and met to study how such a future society based on notions of commons and open knowledge (with open software) would look like. Suffice to say, for our purposes, that 'open' was more than just values (i.e. a preference for certain behaviours), it had a vision, processes to implement, and resources to create.

(Pause) The intellectual origins of Decide: Citizen participation and open source

Certain aspects of Decide, its political design of self-emergence or auto-gestión (self-government), are implicitly deeply rooted in Spanish associative movements, especially those active found during Spain's democratic turn in the previous century. These associations reflected on forms of organisation, social justice, and equity (Villasante 1999, Villasante 2014). How best to influence institutional agenda? How to make politics more democratic? What way of organising is just? How can we form collectives of equals, of peers? Principally, the ambition is not constrained to voting on ideas or proposals, but to shape what kind of questions to vote on. Paraphrasing Frank Herbert's Dune, who controls the questions controls the answers.

Let me give you an example. In one of the hackathons I followed, one of the discussions was about how to encourage proposals from citizens without leading them on unnecessarily (as you can probably foretell, inciting participation is going to play a key role):

In the case of libraries, it already sets the standard on what the library is. It would be necessary to open it, what is the library? Even the very term of the library is still marked by the term. If you open the library as the problem, maybe you don't need a library. What you still need, is a much smaller place that is a book loan hanger, which the call redefines the problem. Get away not only from the project, but rethink it.

To treat citizens as peers is the key to change democracy. Lefebvre, one of the most original thinkers of the urban in the past century banked quite a lot on the idea of the value of use (Purcell, 2013). In 1967, Lefebvre argued for citizens to have a 'right to the city'. This 'right' questioned and challenged the way cities had been built historically by technical experts. These experts defined the requirements of the cities: where residential and business zones should be, the amplitude of streets to accommodate traffic, and so one. For Lefebvre, this was a 'transactional view' of the city where citizens are shuffled along the city from their homes to their place of work and back. What matters is the city's capacity to respond to economic surpluses: Where is labour? How to employ that labour? How to create growth? For example, Paris' famous boulevards were a source of employment for a growing supply of unemployed people. Greatly increasing the width of the streets also allowed the army to manoeuvrer into the city centre to quell rebellions if it needed to. This view has little space for other considerations such as leisure, greenery, and so forth. It is the domain of certain technical experts who apply their skills according to how their profession understand the city (e.g., economic flows, military tactics, and so on).

Underscoring Lefebvre's proposal of the right to the city is the idea that the people who use the city are best placed to know about its needs. In software engineering speak, we would talk about requirements. Urban studies and open source intersect here with some long-standing problems in software engineering. The entire participative design branch of computer science and information systems revolved about how to get user input in a way that mattered. One of the pioneers, Enid Mumford, created an entire methodology (aptly called ETHICS) that focused on creating ethical systems that catered to the needs of employees and users (Mumford 1983). She reasoned that systems should foster change that increase democracy, flatten hierarchies, and implicated people (especially employees) in the design and use of systems. To do this, their worldview needs to be accounted methodologically by themselves. Accounting for all stakeholder views by giving them power over the design would also benefit organisational leaders since it would increase their commitment and productivity. This is because they possess expert local knowledge. [A later work of hers ponders whether such an incremental change is at all possible, wondering if a socio-technical revolution is not needed (Stahl, 2007).]

Open source doesn't come from the participative design school of thought, but it shares common vibes nonetheless. For example, requirements come from the personal experience of developers who create functionalities to cover their needs. (Howison and Crowston, 2014). Idealistically, even, open source could derive its functionality from non-technical users, but this is not necessarily the case (Nuvolari, 2009). Notwithstanding this caveat, development of systems in open source often originates from use scenarios, which echoes Lefebvre's suggestion to move away from development based on transactional values.

Open source have also a very specific added-value which shares some common ground with emergent organisation and auto-organisation. Open source is often organised around flatter hierarchies, based on meritocracy (with the discussable consequence of social inclusion: the best ideas are implemented regardless of their origin). There are also values of reciprocity which guide methodologies and principles of participation. The open source world is replete with example that started from scratch and organised bottom-up by individuals around which communities form: Linux came from a Finnish student, LaTeX was a pet project from a computer scientist, and so on.

Open source has both methodological values (the best ideas win, grassroot organising) and emancipative values (freedom to remix and improve on others' work, transparency of systems), that cater well to operationalise Lefebvre's conception of another kind of city. It's not surprising, then, that Decide's development was spearheaded by people who were convinced of the important role of open source in opening a city and in changing its development from one of transaction to one of use. For citizens to become some form of peers, a community growth model inspired by open source was a safe bet to make.

(2011—2015) LaboDemo: The ideas behind Decide are moulded

The Spanish Occupy movement provided a fertile soil to test ideas of direct democracy and voting on proposals. These ideas were then taken up and further elaborated in a laboratory lead by three people who would have a major voice in how Decide would be shaped in the future: Pablo Soto, Miguel Arana, and Yago Bermejo. This lab called LaboDemo (short for Laboratorio Democratico in Spanish, or Democratic Laboratory) would play a central role in developing citizen participation within a growing political party that had made public its desire to improve democracy and political transparency.

Despite the initial openness of the newly formed party to more direct democracy mechanisms, the people from LaboDemo who had been asked to help, soon found that it was the political system is mistrustful of participation. In an interview by Marco Deseriis, Arana showed how difficult it is to play party politics and be open to the party base:

In the beginning Podemos liked to use this kind of tools, but very suddenly the leaders of Podemos feared that if they really opened decision channels from the bottom up they would lose control of the party and that was not gonna let the party win the elections and produce changes. They were really afraid of this kind of new mechanism. [original emphasis] Deseriis (2018)

There is a strong tension between party politics and how representative politics work, and politics which account for and knows how to work with citizens. This is a tension that will remain visible in Decide throughout its implementation. The tension will take various skins: debates about offline/online democracy, the marketisation of politics, and the individualism of citizen democracy. Some of these tensions, such as the latter, are so deeply embedded to the way our society works that proponents will suggest that participatory tools have to become normalised into the culture for them to work. How to make such cultural change is the name of the game.

(2015) Decide Madrid is implemented

When Ahora Madrid wins the local Madrid elections on a civic movement ticket of participative democracy, the doors are opened to another large-scale experiment on democracy and the transformational force of open source. This time, the experiment is within (and against) the way institutions are currently imagined. The shared feeling that there are deep, inset flaws in Spanish democracy and the experiments conducted in the 15M will shape Decide and how it is thought and evolved in meaningful ways.

Decide becomes implemented in just about six weeks with a minimum viable product launching in September. The initial uptake is massive with more than 700000 votes are submitted in the first few months ( The ambition of Decide to open source the city is seemingly being realised. In the words of one who is close to the project:

A society’s operating system would therefore be a series of common practices and human relationships, not just a set of online platforms. Gutiérrez (2016)

The city becomes something people can interact with and transform through collective practices. The individual must recede back to allow for the city to be shared again and understood collectively. In other words, the city can only be truly open if such collective relations and common practices exist. This will become the crux of the difficulty of Decide Madrid in the future. In particular, one way to evaluate such city is based on the concept of participation, and in particular, full participation. In Purcell's (2013) words:

Through this kind of participation, inhabitants experience an awakening. They come into consciousness of themselves as inhabitants, as embedded in a web of social connections, as dependent on and stewards of “the urban.” As they become conscious in this way, they recognize the need to struggle against the industrial capitalist city and for the urban. They come to see participation not as speaking at a public hearing or serving on a citizens’ panel, but as the living struggle for a city that is controlled by its inhabitants. Purcell (2014)

Proponents of open sourcing the city voice a similar wish based on “full participation” (though without the marxist undertones). Full participation is a recurring topic in participation studies and political science. For Jason Hibbets (2013), for example, “full participation” is about being “embedded in a web of social connection”, but also about “full participation in government decision-making”, which involves relations between city administrations, organisations, and citizens may hold an equal footing, changing the roles of each and bringing us close to Lefebvre's vision of open cities. If such a re-shaping is possible, then citizens can become an organisational force.

The concept of open cities is not an abstract concept. It will be deployed as a way to evaluate whether Decide will be successful. for Decide to be sustainable, it must thus not only be sustainable in the code, but also in the objective it was (is) supposed to accomplish: it must create different social relations, a different politics, and a different city.

(2017) The roller-coaster of participation

When Decide is implemented, there is a rush of participation. In the first months, more than 700000 votes will be cast on proposals. Staunching a vacuum, the platform clearly responds to a felt need. As time goes by, however, participation dwindles. February 2017 will mark a particularly bittersweet moment. It will be both historical: two citizen proposals from the platform are accepted by the citizenry through voting. The downside is that a large part of that voting came from traditional paper ballots sent to citizens by post for a cost of more than one million euros (including publicity of the event). Though the platform has successfully passed two proposals, the platform is not yet seen by the citizenry as the go-to-way to vote. Social relations within the city, then, are likely to take much more time to alter.

(2017-2019) Bringing Decide to the city

The bitter-sweet February event will launch a number of initiatives specially focused on evaluating the sustainability of participation of the platform. That is, when the tool does not sustain by virtue of its code its democratic ambitions, it's to the city that the digital activists go back to, the very same one that launched this project in the first place.

It's on this aspect that the research has focused and you can find in lessons learned the findings that build on the incredible history of this platform. You can also find resources for further study. I will try to keep that up-to-date.

This webpage serves as the open repository of public resources for the Marie Curie project 'Beyond the Code: Sustainability in Open Source'. I investigate the sustainability of open source beyond the technical spaces where open source projects are created.

What happens when open source leaves the technical domain and comes into contact with the 'wild'? What forces does it stir? What questions does it ask of sustainability?

This web contains posts and work-in-progress on research the sustainability of open source beyond technical environments.

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 744957


This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 744957

The purpose of the project was to understand how open source sustainability is managed beyond technical production environments. Previous studies have focused on the sustainability of open source within the development realm, but have paid scant attention to how sustainability is articulated once deployed into the world. Since open source software is increasingly common and no longer the sole remit of software developers, it is important to understand how it is sustained in more open social contexts. Are there specific barriers to sustain such projects? What mechanisms are put in place to sustain such systems?

This page is a summary of the lessons learned from the project. It has been under active construction the past year, but is now in a much more succinct form, resting on the findings of this project. Around these four core ideas, a number of papers have been derived, which will be linked from here.

Sustainability in open source commons


This part of the project critically looked at the adoption of Ostrom's work on ecological commons directly into open source and digital commons. After a literature review, we propose that open source has multiple types of sustainability.

Democracy and the organisation of passions and interests


Participation captures the zeitgeist of contemporary society with the language of participation being used today as a surrogate for processes of democratization, empowerment, and political plurality.

We suggest that participation is an organisational matter that hinges on the material and logistical forms lent to the relation between passions and interests. We draw inspiration from Hirschman’s intellectual history of the development of capitalism as the historical replacement of the language of passionate governance for self-interested economic advantage. In particular, we argue that ‘passionate participation’ and ‘interested participation’ modulate the organisation of participation.

Our argument is based on our longitudinal research of Decide Madrid. Born of the wave of protests in 2011, and in particular the Spanish Occupy protests (locally known as 15M), Decide was developed with the aim of empowering citizens and responding to democracy’s persistent disenchantment.

When participation falters, Decide Madrid tries to organise participation by re-stoking the collective passions of the 15M or to consider ways to create communities aligned to interests rather than fuelled by passions. Choices regarding the mobilization of passionate and interested participation calls for different resources, spaces, and material requirements.

Social inclusion


The project also sheds light on how open source projects can be sustained outside of technical environments and the role these have on the project in return. The particular case studied mixed open source ethos and issues related to urbanisation and direct and deliberative democracy. For example, France’s recent Convention pour le Climat, which randomly selected citizens all over the country to discuss and propose solutions to the climate emergency, were supported by a non-profit group called open source politics using a software programme such as Decide Madrid. The key insights developed about the barriers to deploy such systems are thus useful beyond the open source communities and projects, and extends to all stakeholders. Such open systems are gaining momentum given their adoption by cities (Madrid, Paris, New York, Porto Alegre), countries (France’s Convention pour le Climat), and institutions (European Commission, United Nations).

The principal idea is massive and direct participation will improve our democratic systems. Similar to the cornerstone open source adage advanced by Raymond: the more eyeballs, the less bugs, a democracy where anyone can participate should improve the social outcomes for everybody. We find that this cannot be so because it ignores asymmetries of power. Indeed, in technical environments where developers share skills, asymmetries are lesser.

In society at large, some groups need representatives to pass on their message. One example: victims of abuse are not registered in the same database that is used to verify that digital participation is valid. These people are thus involuntarily set aside since, for security reasons, their data cannot be accessed by participation platforms. As such, massive participation systems cannot, on their own, resolve the democratic disenchantment that is taking root in Europe.



One key question has been how the platform should access the city, and how the city should access the platform. Contrary to what we know, where the interface is located has been a tense frontier. For developers, the interface is located at the screen-level: it is where participants press buttons. The simplicity of the interface was thought to be self-evident which would avoid requiring guidelines on how to participation. For others, the interface was located in the streets, in the squares, in communities. Those that adhered to the screen-interface view pushed for efforts to transform direct participation by algorithmic means, effectively turning direct participation into an indirect one. Those that saw the interface as more malleable argued for different, but complementary ways to adapt the platform to their various needs. The conclusion that we find is that open systems have particularly open interfaces that should attempt to interface in different ways the complex realities within and outside the digital space.

This is repository of public resources about Decide Madrid. It serves as a central repository for anyone who wants to study the platform and its incredible trajectory. I exclude academic articles because those are easier to find on search engines (unless they are from people who were involved).

Dissemination articles from the project (new ones coming out)

Civic tech platforms that enable citizens to participate in decision-making activities in their city are becoming more common. Yet, they still face many challenge.

In this article on, I argue that opening cities require developing an infrastructure to cultivate and nurture a participatory culture, something that open source has done for decades and which contributes to its sustainability.

This article reflects on two important contributions of MediaLab Prado to the development of Decide Madrid and participative technology in general.

  • How the gig economy shapes us

I've written a short article with Roser Pujadas on the role of invisible infrastructure that Uber tries to create and how it affects our cities.

This is part of our larger work on the digital economy, of which you can find an academic article here.

  • Slides from a conference on institutionalising participation

Participation is key to understand our contemporary societies. Despite its prevalence, it remains insufficiently theorised in certain context. One such context is crowdsourcing, which has been proposed as a way to resolve so called ‘wicked’ problems. Crowdsourcing has tended to hold implicit rational and universal epistemologies: the more participation the better. Using resourcing theory, we look at a civic platform implemented with the ambitions to change social relations within the city of Madrid. We propose a processual view of participation which evolves in time taking and being given different qualities as schemas are enacted and resources created. The platform we analyse was supposed to engage citizens to become participative members of society in the definition of public policies. When this does not occur, alternative resources are designed to palliate the deficiencies of the platform. We contribute to the literature on participation and urban policies by arguing that participation, however it is designed, will necessarily favour some at the expense of others, challenging its often implicit universalist and rationalist assumptions.

Useful resources to study Decide Madrid

From people involved in thinking about participative democracy

(2021) Citizen Participation and Machine Learning for a Better Democracy, by M. Arana-Catania, F.A. Van Lier, Rob Procter, Nataliya Tkachenko, Yulan He, Arkaitz Zubiaga, Maria Liakata

Primary sources of Decide Madrid

Medialab Prado's own design work

Other interesting sources