Work-Sharing and Reduced Working Time: What Each Position Is Protecting

In 1930, John Maynard Keynes published an after-dinner essay called "Economic Possibilities for our Grandchildren." He predicted that within a century, rising productivity would allow people to work as little as fifteen hours a week. He was not a utopian dreamer. He was the most influential economist of his era, and he got the productivity forecast roughly right — living standards in rich countries are approximately four to eight times higher than they were in 1930. What he got wrong was the assumption that people would translate wealth into time rather than more consumption. The fifteen-hour week has not arrived. American workers now average roughly 1,800 hours per year — about the same as in the 1930s, and nearly 460 hours more than German workers, the equivalent of eleven additional forty-hour weeks annually.

The question of why this happened — and whether it should be reversed — is at the center of a debate that has been running since the early industrial era and has recently grown urgent again. The arrival of AI-driven automation, the post-pandemic reassessment of remote and flexible work, and a string of large-scale trials of four-day work weeks have brought reduced working time back into mainstream policy discussion. In March 2024, Senator Bernie Sanders introduced the Thirty-Two Hour Work Week Act, which would reduce the overtime threshold from forty to thirty-two hours over four years without a pay cut. The legislation did not pass, but its introduction reflected a shift: reducing the working week is no longer only a union demand or a Nordic peculiarity. It has become a policy proposal.

The debate has four distinct positions. They do not disagree merely about whether shorter hours are good or bad. They disagree about what the problem with work actually is, and therefore about what a solution would need to address.

What work-time reduction advocates are protecting

The proposition that the forty-hour week was a political settlement, not a natural law — and that productivity gains since the 1930s should have been taken partly as time, not purely as income. This is the oldest and most sweeping version of the argument, and its contemporary form is most closely associated with Juliet Schor, an economist and sociologist at Boston College whose 1992 book The Overworked American documented that from 1969 to 1987, the average employed American added the equivalent of an extra month of work per year. Her later work — Plenitude (2010) and her ongoing research with 4 Day Week Global — argues that reduced working time is inseparable from questions of ecological sustainability and genuine wellbeing.

The empirical case for the four-day week has been substantially strengthened by two large trials. Iceland ran a series of shorter working week trials from 2015 to 2019, covering approximately 2,500 public sector workers — over one percent of Iceland's entire labor force — across sixty-six workplaces including offices, preschools, social service agencies, and hospitals. Workers moved from forty to thirty-five or thirty-six hours per week at full pay. The results, documented in a joint report by the Autonomy Institute and Iceland's Association for Sustainability and Democracy (ALDA), found that productivity and service quality remained the same or improved at the majority of workplaces, while workers reported substantially lower stress, fewer sick days, and better work-life balance. After the trials, eighty-six percent of Iceland's workforce either moved to shorter hours or gained the right to negotiate them.

In 2022, 4 Day Week Global ran what became the world's largest four-day week trial: sixty-one companies and approximately 2,900 workers in the United Kingdom, over six months, using the 100:80:100 model — one hundred percent of pay for eighty percent of the time, with a commitment to maintain one hundred percent of output. The academic evaluation, conducted by researchers at Boston College, Cambridge University, and the Autonomy Institute, found that revenue rose an average of 1.4 percent over the trial period and 35 percent higher than the same period in prior years; staff resignations fell 57 percent; sick days fell 65 percent. Ninety-two percent of companies continued the four-day week after the trial ended. One year later, 89 percent were still on a four-day schedule.

A more recent and institutionally different test is unfolding in South Korea. In April 2025, Gyeonggi Province launched a 4.5-day workweek pilot for local firms, explicitly aimed at small and medium-sized enterprises rather than only large knowledge-sector employers. By December 31, 2025, 106 companies and one public institution had joined. Provincial results released in March 2026 reported labor productivity per worker up 2.1 percent year over year, hiring competition ratios up 72 percent, turnover down from 22.8 percent to 17.4 percent, and lower worker stress. This does not settle the debate about generalizability, but it matters because it tests shorter-hour policy in exactly the kind of firms skeptics often assume cannot make it work.

Andrew Barnes, the New Zealand entrepreneur who pioneered the model at his firm Perpetual Guardian in 2018 and co-founded 4 Day Week Global in 2019, frames the argument in terms of a basic unfairness: since the Fair Labor Standards Act of 1938 established the forty-hour threshold for overtime, worker productivity has roughly tripled. None of that productivity gain has been taken as time. It has been taken as profit. The 100:80:100 model is, in his framing, a way of redistributing that accumulated productivity dividend without requiring wage increases or tax increases — simply by returning time to workers.

What this position ultimately protects is a claim about what prosperity should feel like: that the point of economic growth is not more stuff but more life, and that a society in which people work more hours than their grandparents while owning far more than their grandparents has made a choice — not a neutral one, but a culturally and politically constructed one — that is worth examining and reversing. The forty-hour week was won through struggle against enormous resistance. So was the eight-hour day before it. The argument is that the campaign for thirty-two hours is the same fight in a different century.

What work-sharing advocates are protecting

The proposition that the real crisis is not too many hours but too concentrated displacement — and that the right response is to spread work across more people rather than letting it pool unevenly while others are left out entirely. This is the logic behind Germany's Kurzarbeit (short-time work) scheme, one of the most successful labor market interventions in the postwar era, and behind the work-sharing programs embedded in the unemployment insurance systems of twenty-six American states and Washington D.C.

The Kurzarbeit model is simple: when a firm faces a demand shock and needs to cut labor costs, instead of laying off workers, it reduces hours across the workforce, and the government subsidizes the lost wages at sixty percent (sixty-seven percent for workers with children). Workers keep their jobs, their skills, their benefits, and their attachment to the labor market. The firm keeps its trained workforce, avoiding the cost of rehiring and retraining when demand recovers. The government spends less on unemployment insurance than it would pay out in layoffs.

The results speak for themselves in two crises. During the 2008-2009 financial crisis, Germany was the only G7 country that did not see a fall in employment despite a GDP contraction of roughly six percentage points. German unemployment claims rose just 0.2 percent; France's rose 1.6 percent over the same period. During the 2020 COVID-19 shock, applications were filed for more than ten million workers; at peak in April 2020, approximately six million workers — over twenty percent of all socially insured workers — were on the scheme. A study by the Institute for Macroeconomics and Business Cycle Research estimated that Kurzarbeit saved 2.2 million jobs. German unemployment rose from 3.2 to 3.8 percent — a modest increase, compared to the catastrophic spike in American unemployment.

France's thirty-five-hour work week, introduced through the Aubry laws of 1998 and 2000, was a different kind of work-sharing intervention: a statutory reduction in the standard week as a way of spreading employment during a period of twelve percent unemployment. The results were contested — independent estimates of job creation range from 80,000 to 350,000, and firm-level studies found that total factor productivity fell in participating companies even as employment rose. The thirty-five-hour week now functions primarily as an overtime threshold rather than a hard cap; managers can impose up to 220 hours of overtime annually. The lesson drawn by work-sharing advocates is not that the policy failed but that it was eroded: the original logic of spreading work was sound, but political resistance from employers and the gradual accumulation of flexibility provisions hollowed it out.

The United States has had work-sharing built into its unemployment insurance system since the 1970s in some states, and the 2012 Layoffs Prevention Act expanded federal support for such programs. But uptake has been persistently low. During the 2020 pandemic, short-time compensation claims peaked at only 1.6 percent of all unemployment insurance claims in July 2020, while regular UI claims surged to historic levels. The reasons are structural: studies find that only 28 to 54 percent of employers in states with these programs are even aware they exist; the administrative processes are complex; and the system was never built with work-sharing as a primary mechanism. Germany's Kurzarbeit works because it is universal and well-understood. America's work-sharing programs work in theory but fail in reach.

What work-sharing advocates are protecting is the employment relationship itself as a social good — the recognition that a worker who is laid off loses not just income but identity, skills, health, social connection, and labor market attachment. The distributional logic is compelling: concentrating displacement in full job losses imposes catastrophic costs on specific people while leaving everyone else untouched. Spreading displacement as reduced hours shares the cost widely while keeping the harm manageable for each individual. When automation eliminates demand for labor, the first question should be how to share the remaining work, not who to discard.

What productivity-first skeptics are protecting

The proposition that mandated hour reductions sacrifice output without delivering the promised gains, burden sectors that cannot automate their way to efficiency, and may harm the most vulnerable workers most — and that what workers actually want is higher wages, not fewer hours. This is not a single unified position; it draws on standard neoclassical economics, sector-specific analysis, and concerns about global competitiveness and labor market inequality.

The core arithmetic challenge is real. For a four-day week to be revenue-neutral, output per hour must rise by approximately twenty-five percent — a large productivity gain that the trial evidence, however positive, does not prove is replicable across the economy. The 4 Day Week Global trials selected for self-motivated, productivity-conscious organizations; the UK trial's 35 percent revenue growth benchmark compared to prior years occurred during a period of post-pandemic economic recovery, a potentially confounding factor. Researchers at Washington University in St. Louis have noted that approximately seventy-five percent of the American workforce works with their hands or in service roles where there are no extraneous meetings to eliminate — the productivity interventions that make four-day weeks work in knowledge-work environments may simply not be available in construction, healthcare, food service, or elder care.

Japan provides a cautionary historical case. When Japan shortened its standard work week from forty-six to forty hours in 1988, the predicted productivity compensations did not materialize, and economic output in the following years was lower than it would otherwise have been. A 2022 study in the journal Economica by Nicholas Crafts reviewed the evidence for Keynes's fifteen-hour prediction and concluded that the main reason it has not arrived is not irrational consumer preferences or capitalist oppression but the simple fact that hourly productivity, while much higher than in 1930, remains well below the level required to sustain current living standards at fifteen hours of work per week.

The inequality concern is less widely discussed but deserves attention. When firms face mandated hour reductions without proportional wage cuts, one rational response is to hire more part-time workers rather than give full-time workers fewer hours. France's experience with the thirty-five-hour week included this dynamic: firms in hospitality, healthcare, and trucking responded to the reform by expanding part-time hiring, creating a workforce divided between full-time workers with employment security and protections and part-time workers without them. Wharton management professor Matthew Bidwell has argued that economy-wide work-time reduction is in tension with itself: if pay reflects value created, and workers create less value in fewer hours, then same-pay shorter-hour mandates either compress output or shift costs onto employers who will seek workarounds — often at the expense of workers with the least bargaining power.

The deepest version of the skeptic argument is not that shorter hours are bad in principle but that mandate-driven reduction is the wrong mechanism: that firms, workers, and sectors differ enormously in their capacity for time compression, and that forcing a single schedule on this heterogeneity produces winners and losers in ways the proponents of the policy do not fully reckon with. What this position ultimately protects is the concern for workers who are not in the knowledge-sector offices where four-day weeks are easiest to implement — for nurses and care workers and warehouse staff and restaurant workers for whom flexibility is either structurally impossible or already weaponized against their interests.

What worker-control advocates are protecting

The proposition that the fundamental problem is not how many hours people work but who controls those hours — and that a four-day week imposed by employers or mandated by governments does not address the underlying power asymmetry that makes work exhausting, precarious, and dehumanizing in the first place. This is not opposition to reduced working time; it is a critique of the frame.

For gig workers, caregivers, retail workers, and the growing workforce of algorithmically managed employees, the crisis is not too many hours per se but chaotic, unpredictable hours — what labor economists call "schedule volatility." A retail worker on-call for a thirty-hour week who does not know until forty-eight hours in advance which thirty hours she will be working cannot arrange childcare, take a second job, attend a night class, or plan a weekend. The harm is not hours; it is the power relationship that determines when those hours happen. Research by Susan Lambert, Arne Kalleberg, and others on precarious schedules has documented that schedule volatility independently predicts lower income, worse health outcomes, and reduced children's educational attainment — separate from total hours worked.

The European "right to disconnect" framework captures part of this argument. France was the first country to codify it in 2017, legally protecting employees' right to ignore work communications outside working hours. Belgium extended it in 2022 to public sector workers, and Ireland's 2021 Code of Practice on the Right to Disconnect gave workers a documented basis to refuse after-hours contact. Germany, the Netherlands, and Italy have followed. The argument is not that workers need fewer hours but that they need genuine sovereignty over the hours they are not working — that the colonization of personal time by employer expectations is a distinct harm from long working hours, and that a four-day week in which Thursday evenings are spent on Slack has not actually solved the problem.

The more structural version of this position pushes further: into questions of ownership and governance. Germany's codetermination (Mitbestimmungsgesetz) system requires that companies with more than 2,000 employees give workers half the seats on supervisory boards — the bodies that set strategy, approve major investments, and hire and fire senior management. This is not primarily a work-time policy, but it shapes work-time outcomes: German workers can negotiate working hours at the board level, not just at the collective bargaining table. The result is that German manufacturing workers average approximately 1,343 hours per year — roughly the same as the UK's 1,524 and substantially below the US's 1,799 — without a statutory four-day week, because the power structure of German corporations includes workers as a genuine countervailing force.

Worker ownership models — employee stock ownership plans, worker cooperatives, platform cooperatives — extend this logic further. Mondragon, the Basque cooperative federation, has adjusted working time during downturns by distributing reduced hours across its workforce as a democratic decision by worker-owners rather than as a management imposition. Platform cooperatives like Up&Go (a cleaning services cooperative) and Stocksy United (a photographer-owned stock image platform) build schedule autonomy into their governance structures from the beginning. What these models protect is the recognition that the hours question cannot be separated from the ownership question: who decides how time is allocated is inseparable from who owns the enterprise whose work fills that time.

The structural tension the debate reveals

The four positions appear to be arguing about a common object — the length of the working week — but they are actually arguing about four different things: the distribution of productivity gains, the distribution of employment risk, the efficiency of output, and the distribution of power over time. These are not four angles on the same question. They are four different questions that share vocabulary.

This explains why the debate often generates more heat than light. When a work-time reduction advocate cites the Iceland trials and a skeptic cites Japan's experience in the late 1980s, they are not actually disagreeing about the same intervention. Iceland reduced hours within a highly organized public sector; Japan mandated hour reduction economy-wide without the organizational supports that allow compression. When a work-sharing advocate defends Kurzarbeit and a worker-control advocate points to schedule volatility in American retail, they are describing real harms but they are not in conflict — Germany's Kurzarbeit addresses cyclical job loss while American schedule volatility describes chronic misuse of flexibility. Both can be true simultaneously.

The hidden agreement running beneath these four positions is more important than their surface disagreements: all four accept that the current American arrangement — 1,800-hour work years, declining union density, a safety net designed for temporary rather than structural unemployment, and a corporate governance structure that treats labor as the most adjustable cost — is not working for most workers. The disagreement is about what structural change to make. And on that question, the four positions are not mutually exclusive. A policy suite that combined work-sharing infrastructure (expanded and simplified STC programs on the German model), statutory rights to disconnect and predictable schedules, works councils in large firms, and a gradual reduction in the overtime threshold would address different aspects of the problem simultaneously.

What is not available as an option is the status quo indefinitely extended. AI-driven automation is already eliminating cognitive work at a pace and scale that the existing safety net was not designed to absorb. The question of how to share working time — rather than concentrating its losses on specific workers and its gains at the top — is not going away. The debate over mechanism is real and important. But the urgency of having any mechanism at all is the thing that all four positions, in their different ways, are trying to convey.

What sensemaking surfaces

Each position has a legitimate core concern that the others tend to talk past.

Work-time reduction advocates are right that the forty-hour week is a political artifact, not a natural optimum, and that the failure to translate three generations of productivity growth into more leisure represents a specific set of choices — about taxation, about corporate governance, about the cultural equation of long hours with seriousness and commitment — that can be made differently. The weakness of this position is that the strongest evidence (the trials) comes from self-selected, primarily knowledge-sector organizations, and the extrapolation from motivated trials to economy-wide mandates involves assumptions that the trial evidence does not validate.

Work-sharing advocates are right that Kurzarbeit is a proven, practical, and relatively cheap mechanism for distributing employment risk more fairly — and that the United States has almost entirely failed to build a comparable infrastructure despite having the legal framework for it in twenty-six states. The weakness of this position is that work-sharing schemes address cyclical displacement (demand shocks) better than structural displacement (permanent automation of job categories). Germany's Kurzarbeit saved millions of jobs during temporary downturns; it cannot preserve jobs that automation has eliminated permanently.

Productivity-first skeptics are right that the trial evidence is stronger for some sectors than others, that mandated rigidities can harm the most precarious workers through part-time substitution, and that the gap between a successful pilot in a voluntary, organized organization and an economy-wide statutory change is large and underexamined. The weakness of this position is that it tends to treat the current arrangement as a neutral baseline rather than as a set of equally contestable choices — as though the forty-hour week was determined by efficiency rather than by the outcome of a labor struggle in 1938.

Worker-control advocates are right that schedule autonomy and governance power matter independently of total hours, and that a four-day week in a firm with no worker voice is a smaller reform than it appears. The weakness of this position is that it can function as an argument against any specific reform by always pointing to a more structural change that hasn't happened yet — and that workers in the meantime are still working fifty hours a week on unpredictable schedules.

The positions are more complementary than competitive. A policy framework that simultaneously expanded work-sharing infrastructure, codified rights to disconnect and predictable schedules, and gradually reduced the overtime threshold would be more defensible than any single intervention alone — because it would address the cyclical, structural, and power-distribution dimensions of the problem together. The obstacle is not analytical. It is the same obstacle that blocked the eight-hour day in the nineteenth century and the forty-hour week in the twentieth: organized resistance from employers who benefit from the current arrangement. The debate about mechanism is real, but it should not obscure what all four positions share — the recognition that the current default is both inefficient and unjust, and that choosing it by inaction is still a choice.

Further reading

• Juliet Schor, The Overworked American: The Unexpected Decline of Leisure (Basic Books, 1992) — the foundational modern argument that Americans work more than their productivity requires and more than their counterparts in other rich countries.
• Juliet Schor, Plenitude: The New Economics of True Wealth (Penguin, 2010) — extends the argument to connect reduced working time with ecological sustainability and genuine wellbeing.
• Andrew Barnes, The 4 Day Week: How the Flexible Work Revolution Can Increase Productivity, Profitability and Wellbeing (Piatkus, 2020) — the practical and philosophical case from the entrepreneur who pioneered the 100:80:100 model.
• Gudmundur Haraldsson and Jack Kellam, "Going Public: Iceland's Journey to a Shorter Working Week" (Autonomy/ALDA, 2021) — the full report on Iceland's 2015–2019 trials, covering methodology, results, and policy uptake.
• Juliet Schor, Wen Fan, and colleagues, UK Four-Day Week Pilot Results (4 Day Week Global/Boston College/Cambridge, 2022) — the academic evaluation of the 61-company UK trial; the most rigorous large-scale evidence available.
• Gyeonggi Province News Portal, "2025 Policies and Administrative Changes in Gyeonggi Province" (December 31, 2024) — official announcement that the province would launch a 4.5-day workweek pilot in 2025, initially covering roughly fifty firms and pairing shorter hours with support for participating employers and workers.
• "Gyeonggi Province's 4.5-Day Workweek Pilot Boosts Productivity, Cuts Turnover", Seoul Economic Daily (March 6, 2026) — reporting on provincial 2025 pilot data: 106 participating companies, modest productivity gains, lower turnover, and improved worker-reported wellbeing. Useful because it tracks a public-policy pilot in SMEs rather than a self-selected white-collar trial.
• John Maynard Keynes, "Economic Possibilities for our Grandchildren" (1930) — the original prediction and its reasoning; worth reading alongside the critiques.
• Nicholas Crafts, "The 15-Hour Week: Keynes's Prediction Revisited" (Economica, 2022) — a careful empirical revisit of why the prediction hasn't arrived and what it would require.
• OECD, Compendium of Productivity Indicators 2025 — the authoritative cross-country comparison of hours worked and GDP per hour, documenting the US/Germany/France divergence.
• Pavlina Tcherneva, The Case for a Job Guarantee (Polity, 2020) — the systematic argument for distributing work as a public good; complements the work-sharing position.
• Arlie Hochschild, The Time Bind: When Work Becomes Home and Home Becomes Work (Holt, 1997) — an ethnographic study of why workers at a progressive employer with flexible policies chose longer hours anyway; essential for understanding the cultural and organizational dynamics at play.
• Susan Lambert, "Passing the Buck: Labor Flexibility Practices that Transfer Risk onto Hourly Workers" — the empirical case for schedule predictability as distinct from and as important as total hours.
• Will Stronge and Aidan Harper, "The Shorter Working Week: A Radical and Pragmatic Proposal" (Autonomy, 2019) — the UK thinktank that has done the most systematic policy work on four-day week implementation.