🔴 The French State's Three Broken Digital Promises

Fifteen years on, the verdict is damning: none of these promises has been kept. Here is why.

Up until Sarkozy, you could still get away with grand speeches about digital technology. Hollande's five-year term marks the moment when it became necessary to act. But the team that came to power had no real vision, and it was a whole generation of activists that found itself at the controls of the state's digital policy.

With Emmanuel Macron's election, the state's digital projects carried on as they were, without being questioned, because he was mostly interested in startups. To catch his eye, many government departments went on to "startup-ify" the way they operate, forgetting that the state's digital transformation implicitly carries three promises that have to be kept.

1. A promise of lower costs

When people talked about computing in the seventies and eighties (something that has vanished from consultants' speeches today), computing promised lower costs.

When you digitize, and when the state digitizes its services, it should by its very nature cost less to operate, thanks to easier access to data. But that is a theory, for when things are done well and work properly.

But since France fired its computer engineers back in the 1990s, it no longer knows how to do things itself. These days, behind the big projects of the DINUM (the government's interministerial digital directorate) you often find freelancers recruited on Malt.

The biggest mistake of the 1990s was not the deindustrialization of France's technological crown jewels, nor their sale to the Chinese or the Americans (that is the second one), but explaining that engineers and computer scientists no longer had a place within the state, and pushing them toward consulting firms or, for the best of them, to Silicon Valley to work for Big Tech.

Today, it is the consulting firms that build the state's digital infrastructure, at costs obviously far higher than if the resources were in-house. And it is Big Tech's cloud infrastructure, also built in part by French engineers, that weighs heavily on public finances.

The provocative question one might ask is this: if we hired an elite of top-flight developers to fix the state's digital services, would it be more effective and more cost-efficient than continuing to use consulting firms scattered across poorly scoped projects, especially in the age of AI?

That is one of the questions that should be asked for 2027.

2. We have not settled the state's time debt to its citizens

The second promise of digital technology is saving time, and therefore improving public services.

That is what Elon Musk claims to deliver with the DOGE project.

In France, we do not have DOGE, we have the "platform state." A concept seized upon by a young generation of ideologues of a certain kind of digital. Their takeover of public infrastructure did not have the promised effects, except on their careers, which got a serious boost.

Indeed, the platform state has not solved the question of time debt, often more insidious than financial debt: those hundreds of millions of hours lost to inefficient processes, to essential websites like that of URSSAF (the agency that collects social security contributions) that malfunction, forcing users to make call after call or sink into Kafkaesque procedures.

One could have imagined, had the will to simplify been real, a complete overhaul of the digital infrastructure, rather than copying the existing complexity. A generation born with the smartphone could have enjoyed a simplified administrative environment, mobile-first, allowing them to handle all their paperwork with the efficiency of a well-designed system. But sadly, that is not the case.

3. Digital sovereignty, forgotten

The third broken promise is that of digital sovereignty. That is, the state's ability to operate independently of geopolitical logics, platform wars, and today's digital dependencies.

That would have meant developing skills in-house, supporting the open-source startup ecosystem, and fostering credible alternatives to Big Tech.

But after €40 billion poured into French Tech, and a dependence on Big Tech that, according to Nicolas Dufourcq (on C Politique), costs us more than €100 billion, the verdict is clear: we have no digital sovereignty.

Digital technology is first and foremost an instrument of power

In Ukraine, the minister of Digital Transformation holds the rank of deputy prime minister.

In France, digital technology has never been seen as a tool of emancipation, but as a lever of control, notably through the use of the nudge, very much in vogue on the American left and used with glee in France for some years now.

Fifteen years ago, when François Hollande came to power, there was nonetheless a real opportunity to change things. Those who lived through that period and are no longer part of it today are left with a deep sense of bitterness at the current state of affairs.

Why didn't it work?

In state digital projects, you have to start not with a use case, but with a target user group.

Start with the early adopters, what I call the "state testers."

Otherwise, this kind of experiment systematically ends up as a bureaucratic monstrosity, as in healthcare, where they try to force every special case into a standard mold.

But for that, at some point, the civil servants appointed to these posts would have had to step out of their hubris and accept that they do not know everything.

For contractors, the state remains a cash cow to be milked as fast as possible, with no penalties and no real measurement of performance. This leads to absurd situations in which products are not even delivered, including for critical police projects.

When you read the DINUM's implementing decree, written in another time, for another Internet, you have to wonder about its obsolescence. And ask whether everything should not be rethought for the age of AI and the new digital challenges.

For my part, I am in favor of a complete reorganization of public digital services in 2027, one that puts a new generation of people at the controls.

In 2027, what should we take away from the state?

Between those who advocate a French-style DOGE and those who cling to an obsolete vision of the state's digital services, there is a middle ground.

The essential debate for 2027 is not "what should be added to the state?" but "what should be taken away from it?" Not in terms of money or staff, but of processes. In a pre-war, post-truth world, the question of digital sovereignty can no longer be handled by ideological weathervanes, but by professionals.

There are three ministries that, for now, have a strategy that seems to work: Defense, Interior and Bercy.

But for the rest, the state's digital strategy, like private investment, is not up to the task. The renewal of ideologies, people and visions is now necessary. Let us hope it will be at the heart of the 2027 question.

The rest is reserved for members of the Digital Resilience Council (Conseil de la Résilience Numérique). Subscribe!