France’s Payroll Software Abandoned Since 2014: €346 Million Spent, Yet No Civil Servant Was Paid

July 18, 2026

A software intended to pay 2.7 million civil servants never paid a single cent. It was supposed to simplify the life of the French administration. It ended up in the scrap heap after seven years of development, leaving behind a debt of €346 million and an admission of failure that remains in the annals of public management.

Key takeaways

  • A massive IT project meant to revolutionize the State’s payroll stalls and never works
  • Years of internal dysfunction: a turnover of 60% per year and chaotic governance
  • A devastating report by the Court of Auditors labels the fiasco a “calamitous counterexample”

A Supercomputer That Never Really Calculated Much

It all began in 2007. As part of the general public policy review decided by Nicolas Sarkozy to save money, a decree created the national payroll operator (ONP), tasked with designing a new payroll software for civil servants. The idea looked good on paper: replace the mosaic of ministerial systems with a single tool, the SI-Paye, capable of automating the remuneration of all state agents. The project aimed to streamline the payroll of the 2.7 million public employees, processing it in a single tool fed by HR information systems modernized within the ministries, with savings estimated at 3,800 positions.

The calculation seemed irrefutable. The stated objective was to achieve 190 million euros in annual savings by eliminating 3,800 posts, and to put an end to irregular payroll practices. One hundred fifteen people worked on this supercomputer for seven years, according to data collected by the association Contribuables Associés. But the French administrative machinery, with its multiple statutes and remuneration rules that vary by body, proved far more stubborn than expected. The processing volume represented 2.8 to 3.1 million pay slips per month, more than any other payroll system in the world. And above all, the State handles 1,500 payroll elements, where large companies manage only a hundred rules, juggling a multitude of procedures specific to each entity.

Seven Years of Drift, Governance Adrift

The project quickly bogged down in technical problems. Eight ministries were to connect to the central calculator, each with its own HR system. None succeeded. If, in the end, the program’s cost turned out to be lower than the most pessimistic forecasts, that was solely because no HRIS managed to connect to the SI-Paye given the delays and ministerial withdrawals. The tally would have been much higher if the project had worked even a little.

The Court of Auditors, in its 2015 annual report, pointed to a staggering accumulation of internal dysfunctions. The operator’s leadership was weakened by a 25-month vacancy in the position of secretary-general and by the instability of the head of the IT department, with four leadership changes in five years. Turnover among teams reached extreme levels: the program’s difficulties led to frightening situations, with the court noting that the turnover rate reached a staggering 60% per year. Eighteen positions remained vacant on average out of a theoretical staff of 133. How does one steer a nationwide IT project under these conditions? The answer is one word: badly.

Due to a lack of a central authority, responsibility dispersed among ministries and the national operator. The Prime Minister Manuel Valls himself would admit later that the absence of central accountability for the program had fallen short. In March 2014, the axe fell: the Ayrault government abruptly halted the venture. The project had been stopped; the tests showed too great a risk to the payroll chain, and the delays to fix the observed dysfunctions proved far too long. Continuing would have cost even more: according to the program’s refounding team, an additional €422 million would have been needed to connect the ministerial HRIS to the SI-Paye, not counting €290 million in extra costs to support its deployment.

Valued at the Symbolic Euro in the State’s Accounts

That is where the story shifts from a budget fiasco to an accounting case study. The 2015 Court of Auditors report delivers an unambiguous verdict: the software acquired by the State under the ONP program is for the most part useless, the SI-Paye showing zero usable value because none of the pilot ministries could connect to it. The consequence is radical and rare in French public accounting: this finding led the State to value the SI-Paye at the symbolic euro in the general accounting that tracks its assets. A public asset of €346 million, reduced to the value of a single coin. It is hard to find a clearer admission of failure.

At its core, the magistrates of Rue Cambon do not mince their words. They speak of a “patent failure of particular gravity,” of “resources spent in pure waste,” and of “the scale of administrative dysfunction.” They summarize the ONP program as a “calamitous counterexample of investment” that nonetheless cost €346 million. The expression has stuck, as it succinctly captures years of mismanagement.

The judicial aspect, for its part, faded away rather quietly. After three years of investigation, the Court of Budgetary and Financial Discipline could not press charges, due to lack of evidence, in this fiasco that had cost the State €346 million. No official therefore had to answer for this expenditure without any corresponding return. The State’s payroll system, meanwhile, continues to rely largely on the same historic applications that ONP was supposed to replace, PAY and ETR, only modestly modernized by the Directorate General of Public Finances. A timely reminder that in public IT, ambition does not guarantee success, and that sometimes the wisest course is to abandon the project rather than dig yourself deeper in.

Sindre Halvorsen

I write about space exploration, frontier science and the technologies that are quietly shaping the future. From Norway, I follow the missions, discoveries and ideas that connect life on Earth with what lies beyond it. My goal is to make complex subjects clear, useful and worth paying attention to.