Segment 1 — The Cold Open

In the spring of 2000, managers at a mid-sized automotive supplier near Lyon sat down with union representatives to implement the new statutory 35-hour week. Instead of hiring additional permanent staff, the company expanded its use of temporary contracts and reclassified many hours above 35 as overtime under newly negotiated annualised schedules. The policy had been written to reduce unemployment by sharing existing work; within a few years the same plant employed roughly the same number of total hours but with a noticeably larger share of contingent workers whose wages and protections lagged behind those of the permanent core. The arithmetic was straightforward on paper: a 39-hour baseline reduced to 35 should have required roughly one additional worker for every nine already employed if output stayed constant, yet the firm’s records showed no such addition to the permanent roster. Instead, the extra capacity came from workers whose contracts reset each assignment and whose benefits did not accumulate across assignments.

Segment 2 — The Good Intention

The measure originated with the government of Lionel Jospin, elected in 1997 on a platform that included fighting France’s stubbornly high unemployment rate, which stood near 12 percent. Labour Minister Martine Aubry championed the reduction from the long-standing 39-hour statutory week, drawing on long-standing French debates about work-sharing that dated back to the 1980s and to earlier experiments in the Mitterrand years. Policymakers believed that capping individual hours would force employers to spread the same volume of work across more people, lowering joblessness without requiring large fiscal outlays. At the time, productivity growth remained solid, and several neighbouring countries were also exploring shorter hours; the French approach therefore appeared both economically feasible and socially progressive. The government explicitly framed the change as a modernisation of working life rather than a blunt restriction. One objection that surfaced even then was that unemployment might be driven more by skill mismatches or regional imbalances than by total hours worked; the response was that a broad cap would still create openings at the margin because firms could not simply intensify the existing workforce indefinitely without visible productivity losses.

Segment 3 — The Implementation

The first Aubry law of 13 June 1998 set the legal framework and offered financial incentives for firms that negotiated earlier reductions. The second law of 19 January 2000 made the 35-hour ceiling compulsory for firms with more than 20 employees, with smaller firms following in 2002. Early adopters in the public sector and in large manufacturing companies often paired the hour reduction with modest wage increases financed by productivity gains and state subsidies. Proponents pointed to initial surveys showing thousands of new hires in firms that had concluded collective agreements. Skeptics, including some employer organisations, warned that rigid hour caps would raise labour costs and encourage firms to shift work to overtime or to external contractors rather than expand permanent payrolls. The laws included explicit provisions for annualised time accounts precisely because negotiators recognised that seasonal or weekly variation in demand would otherwise make the cap unworkable; those same accounts later became the primary vehicle for preserving total hours. The subsidies attached to early agreements were calibrated to offset roughly the cost of the extra hires that the policy hoped to trigger, yet the formula assumed firms would treat the new permanent payroll as the default response rather than one option among several.

Segment 4 — The Unintended Consequences

Once the laws took full effect, many companies responded by redesigning compensation and staffing structures rather than expanding headcount. Annualised working-time accounts allowed firms to schedule longer weeks during peak periods and shorter ones during slack periods while still averaging close to the old total hours. Overtime exemptions were widened through branch-level agreements, and the use of fixed-term contracts (CDD) and temporary agency workers rose sharply; by the mid-2000s the share of temporary employment had increased several percentage points. Because contingent workers fell outside the core 35-hour calculation for many statistical purposes, measured average hours for permanent employees declined modestly while overall labour input per firm remained relatively stable. Wage inequality widened between protected CDI staff and the growing pool of temporary workers who received lower hourly rates and fewer benefits. Second-order effects appeared in work intensity: remaining hours were often packed more densely, with reports of accelerated production rhythms in both manufacturing and services. Unemployment did not fall by the margins originally projected; estimates from INSEE and OECD reviews placed the net employment effect in the low tens of thousands rather than the hundreds of thousands once hoped for. The policy therefore redistributed hours and protections more than it expanded total employment. A natural question is whether some sectors simply absorbed the cap by raising output per hour; the data showed modest productivity gains in adopting firms, but those gains were not large enough to offset the incentive to substitute contract type for headcount. Another objection is that temporary contracts already existed before 2000; the change was that the 35-hour rule gave employers an additional reason to route marginal hours through those contracts rather than through permanent overtime or new hires.

Segment 5 — The Aftermath

Subsequent governments gradually loosened the framework. The 2007 TEPA law under Nicolas Sarkozy increased tax exemptions on overtime, explicitly encouraging hours above 35. Later reforms under Emmanuel Macron further decentralised bargaining, giving individual firms more latitude to set working-time rules. The statutory 35-hour ceiling itself was never repealed, yet its practical reach narrowed through accumulated exemptions and the expansion of “forfait jours” annualised packages for managerial staff. Today most full-time employees still reference the 35-hour benchmark in contracts, but actual average hours worked sit higher once overtime and second jobs are included. The temporary-contract share remains structurally elevated compared with the late 1990s. Each loosening step was presented as a correction for the rigidity that had already emerged, yet the cumulative effect was to embed the very contract differentiation the original law had not anticipated.

Segment 6 — The Lesson

Rigid quantitative caps on a single input—here, weekly hours—invite substitution across other margins such as contract type, overtime rules, and work intensity. When those margins are easier to adjust than the capped variable, the original objective can be met only partially or not at all. Policymakers therefore benefit from modelling the full menu of responses available to the organisations they regulate rather than assuming the capped variable will move in isolation. The French experience continues to surface whenever governments consider hard limits on working time or platform-mediated labour: the question is not only what the rule prohibits but which adjacent practices it leaves unrestricted. One forward-looking implication is that any new cap on hours or on contractor usage will be evaluated by employers against the next easiest margin—whether that is reclassification, geographic relocation, or automation—rather than against the single margin the rule directly targets.

💬 Reply to this email — Patrick reads every one.

Share: X · LinkedIn · WhatsApp

Forwarded this email? Subscribe here — it's free.