Beth Noveck, director of Open Government with President Obama, gave a lecture in January in Santiago, Chile. He spoke of dozens of cases of citizen participation around the world. The Decide Madrid portal, he said, was one of the most prominent. But it is not yet a success: "If they do not fix that platform, people will stop participating," said Noveck.
"I said that? It is recorded? Well, at least they are still talking to me, "explains Noveck in Madrid, where he has just signed an agreement of 150,000 euros for the center he runs, GovLab, to have its first European headquarters in the capital of Spain, Noveck has been speaking for over a year. with the delegate of Participation and Transparency of the City Council, Pablo Soto, on how to improve the tool Decide Madrid. "Here there is a real rigidity with the model. We need to try more things, "explains Noveck in Madrid during the visit in which he signed the agreement and participated in an act of the Aspen Institute.
Noveck is one of the great experts in participatory democracy, which consists in improving the ways for citizens to intervene more in political decisions. Soto understands Noveck's criticism: "The way for these innovations to find their place and future and to consolidate is to investigate and investigate, it's an innovation process," says the Transparency delegate.
For Noveck, Madrid is not just any model: "It's easy to criticize the first one who moves and the one who has been brave to try," he says. Soto has arguments to defend his project: United Nations granted Decide Madrid the prize for the best public service of 2018. Consul, the open source with which the platform was created, is used today by dozens of public institutions.
The problems of Madrid
Madrid is therefore at first sight a success story. But participation is not only giving the option to do so. It is also to receive a good response from citizens. The Madrid model has two basic tools: one, proposals, to propose initiatives for the councilors to discuss, and two, participatory budgets. Madrid allocates 100 million euros from its accounts -from the almost 5,000 total- to which citizens propose and vote their preferences.
On the platform there are 400,000 registered and 20,000 proposals, according to Noveck, but only two of them They have surpassed the 27,000 votes – 1% of the census – to be debated at City Hall. "There is no mechanism for a conversation between City Council and citizens about new possible proposals," he says.
"One of the great challenges with direct democracy is that it is very good for a normal person to propose something, but governing is more difficult"
It's not the only problem. The budgets receive dozens of proposals – bike lanes, more trees, better asphalt – and with a few hundred votes, measures are approved that fit into the general policy of the City Council is uncertain: why a bike lane on one street and not on another? Who decides which parks should be improved? "One of the great challenges with direct democracy is that it is very good that a normal person can propose something, but governing is more difficult than that," says Noveck. "The challenge of these proposals is that we have the danger that nothing will go forward, but we also have the almost equal danger of what happens if something goes ahead? It can be a bad idea," he adds.
The citizen vote of budgets has at least three concrete problems, according to Noveck: one, of legitimacy for lack of participation; two, of effectiveness for lack of coordination with general policies, and three, because it is not easy to clarify which proposals go forward and which do not. "Naked ideas are promoted, without the necessary support," says Noveck.
The germ of participation
Citizen participation has not yet found a clear path: "We do not know what the future citizen participation will be," says Noveck. No one knows what voter involvement will be like with today's media, but it is probably not the same as current models.
What does Noveck propose for Madrid? Many experiments According to the example that Noveck likes the most, Google decided with tests to its users what blue tone to choose for its links among 41 options: it showed different tones to thousands of users and the one that received the most interactions won. So, today it is tremendously easy to compare which option works best: "We bored the City Council telling them why it did not work and what they needed to do experiments and test what works and what does not," says Noveck. "We've had very long discussions about the model, I've said, 'Look, I do not want to persuade you how to do it differently, but I want you to persuade yourself through investigation.'"
"The simple question of why people do not sign proposals is testable," he says. In the end it seems that Noveck has convinced the people of the Town Hall: "The people here are very open to the idea that we must find a way in which we can do better," he adds.
"We bored the City Council telling them why it did not work and what they needed to do experiments"
Noveck proposes seven experiments, which will not all be done in Madrid. It is not the same for Amazon or Uber to use their website to test recommendation algorithms or psychological mechanisms that a public body does: "Technical and probably legal changes are needed to become a platform that relates better to citizens," says Noveck .
The priming It is one of the most obvious examples of experiment: the way you ask questions varies the answers. "You can send messages to citizens to become police, but you can tell them to do it to help their community, to improve their professional career or to shoot," says Noveck. Each message would receive a different number of responses. The same happens with the emails of Decide Madrid. You can be encouraged to participate to improve your life, that of your neighborhood, to help the neighbors, to get to know the city better.
Noveck has other examples of experiments to launch immediately: reduce the participation period to one month instead of several months, have citizens respond to a form with their interests and send them the proposals that are more similar, using processors of natural language to unravel the central messages of thousands of comments or messages, offer rewards to users who achieve their proposals receive more votes or, finally, if it is better to encourage citizens to draft their initiatives or amend existing ones. None of this is science of last generation, but it has never been applied to a field dominated by activism.