France and open data: A narrow view

Open data is in: The French association Gironde launched its open data website DataLocale, Saône-et-Loire announced their portal will open in October, and the French government’s site will be up and running in just a few short months. Other cities are considering this next step (Le Havre, Saint-Maur-des-fosses) while Montpellier has already taken the plunge. The open public data has become the flagship for cities who want to show off their transparency and democratic spirit.

A thorough report entitled For an Ambitious Open Data Policy was recently released on the subject. It was produced by four École des Ponts ParisTech students: Romain Lacombe, François Vauglin, Pierre-Henri Bertin and Alice Vieillefosse. The report synthesizes what’s at stake with open data, and provides a series of 16 recommendations for implementing open data policies. As the students just finished their Masters’ in Public Action at the prestigious École des Ponts, we can probably trust their expertise. The report was given to Éric Besson (Minister of Industry, Energy, and Digital Economy) on July 13.

Romain Lacombe, now an innovation and development agent at Etalab, explains to OWNI the report’s origin and objectives.

“Identify the economic models”

“After studying science at l’école Polytechnique, I transferred to l’École des Ponts and then finished my studies at MIT. After graduation I founded a mobile geolocalization application startup in Silicon Valley in 2008 then sold it in 2010.

When I returned to France with my expertise in both digital culture and public policy, open data seemed the best way for the French government to encourage innovation. In my mind, it was my patriotic duty to develop an open data movement in France. At the end of the summer in 2010, I pitched the idea of conducting a study on public data to the école des Ponts ParisTech.

The study focuses mostly on the economics of open data, but our complementary expertise pushed us to look at the entire spectrum of issues that open data represents.”

“The objective of our study was to identify economic strategies and models that the state could use as inspiration to ease into opening its data and to encourage the data’s reuse. So it’s a reflection on state action in favor of open data – on what motivates a public policy decision in federal governments and in other public institutions.

We started by taking inventory of international and local experiences, then we identified the central issues and open data’s principle stakeholders as well as the fundamental roadblocks to opening data – especially technical, legislative, economic, and organizational aspects. Finally, we outlined potential solutions.

The single most important goal was to identify economic models that would ensure the support for innovation and use of open data while still protecting the public finance equilibrium that produces data.”

“This prospective work and our fruitful discussions with stakeholders on the subject led to the formulation of 16 proposals for an ambitious public data policy, which was submitted to the Department of Higher Education and Research and the Ministry of Industry, Energy and Digital Economy.

The proposals are very concrete recommendations on issues such as interoperability and open formats, free rights to use, enriching the data “bottom-up,” a national strategy for all public stakeholders, and the emergence of a true ecosystem of public data.”

Ambitions

The 16 propositions that the report details are ambitious idea rarely mentioned elsewhere:

• The need for training and pedagogy for public stakeholders who integrate the production and publication of data with their work. This is the report’s first proposition. OWNI has already noted how lack of training negatively affects public data access and use.
• The question of formatting is frequently invoked. The second proposition is dedicated to the quandary:

encouraging the use of easily exploitable formats that respect interoperability norms and ‘computer readability.’

This is a crucial point – according to data from a study (published by Proxima mobile on July 26) conducted by François Bancilhon (Data Publica) and Benjamin Gans (INRIA), of the 6.2 million public data files available online, only 11% are formatted to be easily exploitable.

• Make free use rights the norm. Fees for access to certain data have become exceptions. The report makes this proposition empirically:

Our economic analyses lead us to conclude that free use rights – including free commercial reuse – favor innovation and contribute to financing the public service by generating tax revenue; it is the optimal economic model for public data.

• Multiple propositions aimed at creating a public data ecosystem, valuing an open data policy that will endure the test of time and become a tradition. Any data concerning new contracts (such as public water filtration or public transport), would be made public…

…with a desire for control

As most open data initiatives, this report is driven by the “top down” philosophy, meaning what is published online is decided directly by the organizations. Yet this isn’t the only option. Opening public data is giving citizens back the information that they themselves created.

That’s what Jean-Marc Manach explains here, and that’s the idea behind the Guardian’s “Free our data” campaign which motivated the British government to open their data portal. The report goes so far as to present the inverse alternative (“bottom up”) as dangerous. For example, the report lists three possible “state attitude scenarios” regarding open data policies: inertia, seizure and symbiosis.

Inertia (which is the historic rhythm of data policy developments) is traditionally justified by “the strong risk … of weakening the institutions which have traditionally produced data”:

Citizens’ and consumers’ frustration with data that is difficult to access could potentially motivate them to develop their own ad hoc solutions, perhaps even in a collaborative manner. Although commendable, if such solutions do arise effectively in public and private partnerships, this type of development can be risky – especially considering the loss of legitimacy for the institutions that historically produced data. The risk is that useful expertise would disappear, and the quality of service for citizens would deteriorate.

Breaking down the monopoly structure of data production won’t be a walk in the park. The “bottom-up” system (the request or initiative coming from citizens) proposed in the report clings to crowdsourcing and is relatively limited in scope.

The government should think about the possibility of moving away from a “one-way” model (public sector diffuses data to civil society) towards an ecosystem model where governments and civil societies exchange their data freely and enrich one another’s data in a collaborative way (“crowdsourcing”).

The report doesn’t make any recommendations concerning the implementation of requests for specific data sets on open data websites. Sweden is the only country to offer this service.

The report handed to a minister represents neither Etalab’s guiding principles nor the overall policy of the French government concerning open data – even if quite a few points are shared (data free of cost, for example).

Its contents reveal the current open data political philosophy and initiatives. Public data is just beginning to be released in France. But only in certain, very specific ways.