France’s Great Debate: Civic Washing or Real Engagement?
Two weeks ago Emmanuel Macron, the French president, announced the results of the “Grand Débat” or “Great Debate,” a two-month process of intensive public consultation that garnered nearly two million comments online along with 10,000 local meetings and a series of citizen assemblies. The initiative was in response to France’s “Yellow Vests” crisis: since November last year, thousands of people have taken to the streets, every Saturday, to protest against the government’s policies and more broadly against social and democratic inequalities. Macron’s announcement focused on addressing the public’s complaints about France’s tax system, and whether the “Great Debate” will have provided an adequate response to this social movement and led to changes in policy is still quite uncertain. However, it is already clear that the process has had a tremendous impact on the French civic tech ecosystem.
What happened during the “Great Debate”?
On January 15, Emmanuel Macron announced the launch of a national participatory democracy process, comprised of both a digital platform and local meetings. Local governments, which had mobilized in 2018 to start addressing citizen requests, had already opened “grievance books” where citizens could write their concerns. Macron’s initiative was built in that same spirit. His administration invited citizens, elected officials, NGOs and companies to organize local meetings to discuss the four themes of the Great Debate: ecological transition, tax policy and public spending, democracy and citizenship, and the organization of the state and public services. Meanwhile, to drive attention to the process, Macron traveled the country meeting with local mayors.
These meetings were announced on an official Grand Débat platform, where citizens could also contribute directly, either by responding to “quick questions” or through the submission of a “contribution,” their views on the four themes. In addition, NGOs, professional associations and unions submitted independent contributions. Finally, 18 regional citizen conferences were organized, along with a national youth conference.
After two months of discussions, the platform boasted almost 2 million online contributions. At the same time there were more than 10,000 local meetings, 16,000 local “grievance books” collected and over 27 000 letters and e-mails received. These numbers make it the largest participatory democracy experiment conducted in France: it is estimated that around 500,000 people directly participated in the process, which cost between 12 and 15 million euros.
What have we learned?
Although it is still quite difficult to fully evaluate the process, since a participatory exercise of this size may have additional impacts over the next months or years, it is already possible to highlight some of the lessons learned from the experiment when it comes to digital democracy.
Civil society actors and researchers raised concerns on the design and the duration of the consultation as soon as the process was launched, but their voices was rarely heard. The French National Public Debate Commission (CNDP), responsible for organizing most public debates (mainly on major infrastructure and development projects), was originally tasked with organizing the debate. However, after working on designing the process, the commission announced it relinquished its role, stating that the government refused to truly guarantee the neutrality, transparency, and deliberative quality of the debate. In addition, the government did not commit to using the contributions or explain how the results of the consultation were to support policy making.
The consultation was therefore organized directly by the Macron government, through its Information Service, and the implementation was outsourced to private firms. For the digital platform, the government contracted with a govtech company named Cap Collectif, currently the leader on the French digital participation market. Other companies were later tasked with the analysis of contributions, including a public survey company and consultancies specialized in strategy and communication.
Although five public personalities from the academic and civil society spheres were appointed to “guarantee” the process, they arrived after it has been designed and launched, and had had little say on how the contributions were to be collected, processed, and used by the government. They supported some of the civil society claims for transparency and open data, but could not truly reshape a process planned in less than two months.
Cap Collectif’s engagement platform usually includes several functionalities for citizen interaction, including comments, voting, and the ability to work collectively on proposals. However, the format chosen for the Great Debate was far more restrictive. Citizens could either respond to “quick questions”, through Yes/No answers, or provide “contributions.” This second category could have been more open, but was also organized as a questionnaire: most questions were closed, and the open questions were themselves often quite leading.
As an example, in the category concerning taxes and public spending, one of the open questions was: “To reduce the public deficit of France, which spends more than it makes, do you think we should first reduce public spending or increase taxes?” It was closely followed by the question “To reduce taxes and the public debt, which public expenditures should be reduced as a priority”? The need to reduce public spending was nowhere to be discussed…
In fact, this format, combined with the restriction in themes that could be addressed, severely framed and constrained contributions, leading to concerns of bias. It is reasonable to conclude that the framers of the Great Debate only sought citizen contributions that would ratify answers that the government was already looking for, in order to support already established policy choices, rather than truly contribute to innovative policies and new solutions based on collective intelligence.
The process suffered, from its inception, from a complete lack of transparency. It took intense lobbying from civil society actors, namely NGOs specialized in environmental or democratic participation issues, for the government to agree that all the contributions should be publicly available. Open data was indeed a prerequisite for citizens, civil society organizations or researchers to be able to analyze the results independently.
The publication of the data proved a challenge: as it had not been planned for, the platform provider did not offer a functioning API until relatively late in the process. In addition, all the offline contributions (meeting summaries, grievance books, letters, etc…) had to be digitized. This represented over 500,000 pages of text, which were sent to the French National Library (BNF). This considerable work could however not be completed in the two weeks before the debate closed, and the contractors responsible for data analysis never managed to complete a full analysis. The regional conferences, which were supposed to serve as deliberative add-ons to the process, had to start from scratch as no data was available to them.
The analysis of the contributions was itself a great concern in terms of transparency. The Grand Debate’s managers never gave the public a clear picture of how contributions would be processed or taken into account for policy making, even though political scientists repeatedly stated how important this assurance of impact was for citizens to contribute. When the platform used to collect citizen contributions relies on proprietary software, when the analysis is conducted through opaque software relying on artificial intelligence, and when most of the contributions are not even read, what is the message sent to citizens on the importance of their participation?
Whether or not participants should be able to contribute anonymously was itself an issue of debate. In this case, participants were not asked to provide any information beyond an e-mail and a post code when registering to the platform. Without any information on who participated, it is difficult to evaluate whether all categories of the population were represented.
An exploratory analysis conducted by the Observatory of the Debate (a civil society initiative led by researchers and participation practitioners) showed that participants in local meetings were more educated and had a higher sense of political efficacy than the average French population. Although regional conferences may have been more representative (as they were based on sortition), we have no information on the majority of participants, both those who used the online platform or contributed to grievance books.
This raises questions on the legitimacy of such an exercise when it is presented by the government as a basis for its policy choices. Considering the biased nature of the debate framework, one could wonder if a representative survey would not have been more effective in providing a view of the French population’s priorities and aspirations. Outreach activities to involve diverse population groups and counter self-census mechanisms are crucial for inclusive citizen participation mechanisms. The “Great Debate” is a reminder that providing tools and scenes for participation is not enough, and that without much more transparency and openness the process is closer to civic washing than a true civic deliberation.
5. The role of civic tech
Beyond Cap Collectif, the official platform provider, govtech and civic tech developers were also solicited to support civil society actors and local governments who wished to develop alternative methods of contribution. The Association of French Rural Mayors decided to provide digital grievance books and developed an application with the company Fluicity. The govtech company Open Source Politics helped local governments, for example Lille or Nancy, to use the platform Decidim to have a local instance of the national debate, using richer functionalities. Platforms such as Decidim, noos-citoyens, and even the Cap Collectif platform (with its full range of functionalities) were also used by citizen collectives or NGOs interested in having a digital discussion scene.
In addition, the civic tech community leveraged for itself a critical role in this debate, developing actions to show what could be achieved through well-designed, digitally-supported participation processes. There has not been as much such activity in the community since the Open Government Partnership Summit in 2016, when civil rights NGOs, civic tech developers, free software and digital freedoms’ activists mobilized to expose the distance between the government’s discourse on open government and its actions restricting civil liberties.
France’s civic tech community first opposed the design of the official consultation and argued for a more transparent, open and truthful process. As the government did not respond favorably to their demands, actors decided to “do it themselves.” Developers scraped the official platform to produce open data before the government’s contractor started producing it. Working with researchers, they developed analytical tools to demonstrate how the data could be processed, and how every analysis method provided specific opportunities to explore the data, while suffering from specific biases and limitations. The argument made was that rather than having one service provider analyzing the data, different syntheses could be produced and collectively discussed to truly make use of the rich contributions provided by citizens.
One of the most successful initiatives was the “Great Annotation”, supported by the Code for France community. Developers produced a tool that allowed citizens to tag contributions according to crowdsourced categories, as well as identify interesting or unique ideas. To date, over 1100 people have annotated over 250,000 responses provided by citizens to open questions. The tool additionally allows citizens to compare their annotation with the result which would be provided through an artificial intelligence, showing the limits of such an analysis method (for instance in terms of recognizing nuances or irony).
Although the official “Great Debate” was lacking in neutrality, transparency, deliberative quality and representation of diverse population groups, it had an unquestioned effect on the mobilization of civil society actors on democratic issues. The demand for more democratic mechanisms, including the citizen initiative referendum demanded by the Yellow Vests, led to citizen participation processes being introduced and debated widely in the public arena.
This has also been a moment of discussions and reconfigurations in France’s civic tech ecosystem. In France, the most visible actors are the govtech players that have created a “market” of citizen participation tools, while activists and citizen-led civic tech initiatives are less recognized. Although the national government refused to take into account contributions and analyses submitted through parallel platforms, the mobilization of civic tech organizations has led to the establishment of links with non-digital activists and grassroots organizations.
For example, French “Yellow Vests” have developed links with govtech companies and civic tech NGOs and participated in coding events to develop or improve existing platforms. At the Nuit du Code Citoyen (“Civic coding night”) hackathon, organized by civic tech and tech for good organizations Les Bricodeurs, Code for France, and Latitudes, representatives from environmental and social NGOs and developers worked together to find solutions to improve citizen information, namely through tools to analyze citizen contributions on digital platforms.
At the closure of the online phase of the Great Debate, French MP Paula Forteza invited civic tech developers, activists, researchers and NGOs focused on digital and civic rights to present solutions to better analyze the Great Debate contributions. Issues discussed among the civic tech community, such as the relevance of open source software or open data policies, or strategies and practices for citizen mobilization, have also been brought further into the public arena.
While the civic tech community has been instrumental in raising awareness on digital democracy processes and tools, there is still a lot to be done. There is a wealth of data that can be analyzed for its content or to study how people participate and how they respect or go around the framework provided by the consultation’s organizer. There are still questions on how to include diverse groups of population, ensure that the tools allow for deliberation and make use of collective intelligence. And of course, there is still lobbying to be done for open source, free software that can be appropriated and used by civil society actors. We can only hope the momentum of the “Great Debate” will have created enough energy for this community to consolidate its place in the public sphere, as both developers of platforms for citizen participation and activists pushing for more ambitious, open and transparent participation mechanisms.