In debates about the future of work, shortening the workweek often sounds like an appealing solution to unemployment and worker burnout. Advocates frequently suggest that if employees work fewer hours for the same pay, businesses will need to hire additional staff to keep output steady, spreading employment across a wider pool of workers. Critics counter that forcing employers to maintain full salaries for less time creates an immediate spike in labor costs, which can discourage hiring and damage competitiveness.
A study detailed in an NBER working paper examines what happened when this concept was implemented across an entire economy. Looking closely at France’s shift from a 39-hour to a 35-hour standard workweek at the turn of the century, the researchers found that the policy did not produce the expected wave of new hiring. While overall employment appeared unchanged on the surface, that flat outcome concealed two strong opposing effects: the mandatory reduction in weekly hours reduced employment by 3.8 percent, while accompanying government payroll tax cuts offset the decline by roughly the same amount.
Disentangling a Landmark Labor Reform
Pauline Carry of Princeton University and her colleagues, Claire Montialoux at Sciences Po, Elio Nimier-David at Cornell University, Alexandra Roulet at INSEAD, and Nina Roussille at the Massachusetts Institute of Technology, set out to solve a long-standing economic puzzle. In the late 1990s, the French government passed the Aubry laws, reducing the legal full-time workweek from 39 to 35 hours without cutting monthly pay. At the same time, the state introduced payroll tax reductions targeted at lower-wage workers to cushion the financial impact on firms.
Because the hours reduction and the tax cuts were rolled out together, previous evaluations struggled to isolate the impact of the shorter hours alone. To separate these forces, the researchers took advantage of how the law was introduced. The mandate applied first to firms with more than 20 employees in January 2000, while smaller firms were given until January 2002 to comply. This staggered schedule allowed the team to compare similar businesses on either side of the 20-worker threshold.
The research team assembled comprehensive administrative data covering the universe of private-sector French companies between 1994 and 2003. They matched annual employer-employee payroll filings with corporate balance sheets, tax records, and national labor surveys. By examining differences in company workforce structures, the economists could also evaluate how vulnerable each firm was to the policy. Companies with a high share of full-time staff faced a heavier burden from the hours mandate, while companies with many low-wage employees stood to gain larger payroll tax deductions.
The Missing Work-Sharing Effect
The analysis confirmed that the reform produced a substantial reduction in working time. By 2001, average weekly hours in mandated firms dropped by two hours relative to smaller, non-mandated firms. Adoption of the 35-hour schedule climbed by 35 percentage points more among mandated employers, and 86 percent of those adopting firms claimed the tax relief.
Hourly wages in mandated firms rose by 2.1 percent on average, and by 4 percent for full-time workers who stayed at their jobs. This increase was smaller than the theoretical 11.4 percent increase that a four-hour reduction at constant monthly pay would imply. The researchers found that employers moderated regular pay increases in the period between the law’s announcement and its enforcement, helping dampen the wage jump. The accompanying tax subsidies absorbed roughly 60 percent of the wage increase, leaving the average hourly labor cost to employers higher by less than 1 percent.
Despite the cut in hours, the central goal of work-sharing failed to materialize. Total employment counts at mandated firms tracked those of non-mandated firms almost identically, showing no net job creation. Instead of hiring more people, companies saw their total annual hours worked fall by nearly 6 percent.
By separating the policy components, the authors showed why employment stayed flat. Adopting the shorter workweek without any tax relief led to a 3.8 percent drop in jobs. The payroll tax cuts operated in the opposite direction, lifting employment by 3.9 percent. The two policy levers canceled each other out, making the full reform appear neutral when it was actually hiding substantial private-sector job losses that were prevented only through public subsidies.
Shifting Hours and Growing Skill Divides
Looking across the entire spectrum of working schedules revealed that the reform altered more than just full-time schedules. While positions at 39 hours dropped sharply, they were not replaced exclusively by 35-hour roles. Instead, companies expanded part-time positions, with notable job gains appearing between 25 and 34 hours per week.
Even incumbent workers who were already working 35 hours or fewer experienced a modest reduction in weekly hours and a slight boost in hourly pay. Firm-level workweek agreements from the period suggest that employers extended these adjustments partly out of perceived fairness toward part-time staff.
The reform also generated uneven outcomes across different types of workers. For managers and high-skill professionals, the law functioned much as its designers had envisioned. Their weekly hours declined, their monthly earnings remained intact, and their employment rose by an amount equal to 1 percent of firms’ pre-reform total employment, a sizable increase given that they made up only about 10 percent of the workforce. In contrast, lower-skilled workers, including administrative staff and blue-collar labor, experienced both an employment drop equal to 1 percent of pre-reform total employment and a 1.7 percent decline in average annual earnings for those who remained employed.
The authors suggest two main explanations for this divergence. France had a high national minimum wage relative to the median wage, leaving companies little room to absorb higher hourly costs for low-wage staff. At the same time, survey evidence suggests that the organizational changes firms made in response to the reform were better suited to higher-skill workers.
Productivity and Work Intensification
Even though total hours worked at mandated firms declined by 5.9 percent, company output, measured by value added on corporate balance sheets, fell by only 1 percent. This discrepancy pointed to a marked gain in labor productivity.
To examine whether this was simply a result of changing capital or skill levels, the researchers calculated total factor productivity, an economic metric that captures how efficiently a company turns combined labor and capital inputs into final output. After adjusting for the growing share of high-skill workers, total factor productivity rose by 2 percent in mandated firms.
Workplace surveys analyzed by the team offered clues as to how companies achieved these gains. Before the reform, nearly a quarter of workers reported periods of idle time on the job, and manufacturing firms reported scope to expand production without new hires. Following the hours cut, 42 percent of surveyed employees reported completing the same volume of work in less time. Many cited greater versatility demands, task reallocation, new work procedures, and greater use of information technology, which the authors interpret as firm-led reorganization rather than workers simply increasing their individual pace. These changes coincided with a noticeable uptick in reported workplace stress.
This productivity boost cushioned overall output losses, but it did not prevent them entirely. An accounting breakdown revealed that lower-skilled workers bore 70 percent of the net output decline through a smaller collective wage bill, while business owners absorbed the remaining 30 percent through lower profits.
Tradeoffs in Workweek Regulation
The study provides quantitative evidence on the mechanics of statutory workweek reductions. When governments require employers to reduce hours without trimming pay, the immediate rise in hourly labor expenses tends to decrease employment. Payroll tax subsidies can offset this reduction, but they carry a direct cost to the public purse. In France, the direct cost of the tax relief was approximately 15,000 euros per job saved, which equaled 63 percent of average annual gross earnings at the time. The authors estimate, however, that avoided unemployment benefits and preserved tax revenues more than offset this direct cost, while the reform package as a whole carried a substantial net fiscal cost.
The authors also examined longer-run effects, noting that these estimates rely on stronger assumptions. After the French government decoupled payroll tax cuts from working-time requirements in 2003, they estimate that adopting the 35-hour schedule reduced employment by 3 percent three to four years later, a result significant only at the 10 percent level. The authors say this estimate partly reflects the effect of the shorter workweek once the adoption-linked subsidies had been removed.
The researchers interpret the findings as a cautionary lesson for labor market policy. While shorter workweeks can generate productivity adjustments and provide more leisure time for certain workers, they do not inherently lead to work-sharing or expanded employment, and their burdens tend to fall hardest on the workers who are least equipped to absorb them.
