a/work· 31 October 2026 · 7 min read

The machines did the work. Who kept the time?

Photo: A. C. Applebee

On 1 December 1930, with the Depression deepening, the Kellogg cereal plant in Battle Creek, Michigan, cut its working day from eight hours to six. It made room for hundreds of unemployed people. Two years later government interviewers asked the women on those shifts what they thought, and most of them reached for the same word: freedom. So where did it go?

Two extra hours of this, every day. That was the deal at Kellogg's in 1930.

A promise made in Madrid

That same year the economist John Maynard Keynes gave a lecture in Madrid about his grandchildren. Technical progress, he said, would solve the problem of scarcity within a century. The hard part would be what to do with the freedom. "Three-hour shifts or a fifteen-hour week may put off the problem for a great while" (Keynes, 1930/2008).

His grandchildren are at work now. Most full-time Australians still work up to the 38-hour maximum the law sets. In the United States, the 40-hour week written into law in 1938 still stands. And Kellogg's? Management chipped away at the six-hour day for decades, and the last of it ended in 1985. By then almost the only people still protesting were the women who'd kept it going (Hunnicutt, 1996).

He was right about the machines

The US Bureau of Labor Statistics counts what the business sector makes for each hour worked. By 2024 that was five times its 1948 level (×5.00) and 2.33 times its 1979 level (U.S. Bureau of Labor Statistics, 2026).

The Economic Policy Institute tracks the same story on a stricter measure, net productivity, which subtracts the wear on machines and software. From 1979 to 2023 it rose 64.7%. Typical workers' hourly compensation rose 14.1% (Bivens & Mishel, 2023). Between 1948 and 1973 the two moved together: net productivity rose 96.7% and typical hourly compensation 91.3%. Then the lines split.

Shared as time instead of kept as profit, the gains since 1979 would turn a 40-hour week into about 24 hours (40 / 1.647). Share everything gained since 1948 and the week is about 15 hours: Keynes's number. In Australia, the output of a 1994 38-hour week now takes 25.7 hours (OECD, 2026b).

Where the time went

A gain in output per hour can go three ways: to owners as profit, to workers as pay, or to workers as time. From 1979 to 2024, gross output per hour in the US rose 133.0%. Real hourly compensation averaged across all workers, executives included, rose 53.8%. The real hourly earnings of production and nonsupervisory workers, who do not supervise anyone, rose 9.9% (U.S. Bureau of Labor Statistics, 2026).

Time barely featured. Across 37 OECD members with twenty or more years of records, the median country's output per hour rose 89.5% while hours per worker fell 8.9%. Over five decades, only 7% to 16% of each gain in output per hour came back as shorter hours (OECD, 2026a, 2026b). The fall in hours that began in the 1800s slowed almost to a stop after the 1970s, as union power fell, work blurred into the rest of life and long hours became a badge of virtue (Huberman & Minns, 2007).

The countries that took some back

You'd expect a country that works less to make less. In 2025, across 35 OECD members, it was the reverse. A country whose workers put in 10% more hours made, on average, 18% less in each.

CountryHours per worker, 2025Output per hour vs Australia
Germany1,33220.6% more
Denmark1,36939.4% more
Norway1,38645.3% more
Netherlands1,43621.2% more
France1,49820.4% more
Australia1,633

(Calculated from OECD, 2026a, 2026b.)

Thirteen of the 34 other members work fewer hours than Australia and make more in each hour. Germans work 301 fewer hours a year and still produce slightly more per person (US$60,997 against US$59,490 in 2024).

One reason is old. In British munitions factories in the First World War, output per hour fell once weekly hours passed about fifty, so the longest weeks made little more than moderate ones (Pencavel, 2015).

What happened when workplaces tried it

Between 2015 and 2019, Reykjavík City and Iceland's government moved about 2,500 workers from 40 hours to 35–36 at the same pay. Service held or improved and wellbeing rose. Unions then won shorter hours, or the right to them, for 86% of the workforce (Haraldsson & Kellam, 2021).

In Britain in 2022, 61 organisations tried a four-day week at full pay for six months. Fifty-six (92%) kept it. Sick days fell 65% and departures 57% (Lewis et al., 2023). A peer-reviewed study of 141 organisations in six countries, with 12 control firms, found burnout fell and health improved, and the bigger the cut in hours, the bigger the gain (Fan et al., 2025).

One trial carries a warning. In New Zealand, management intensified monitoring and performance pressure to win the extra day (Delaney & Casey, 2022). Who designs the shorter week decides who keeps its gains.

Why the week will not move by itself

  • Hours at a desk measure presence, which a manager can see. Results take judgement.
  • The gains already have owners, and they would rather keep them.
  • A mortgage sized to a full-time wage locks a household into full-time hours.
  • Long hours became a badge. Working less can feel like letting the side down.

None of that is evidence. And the cost is real: working 55 hours or more a week, against 35 to 40, goes with a 33% higher risk of stroke (RR 1.33) (Kivimäki et al., 2015). The World Health Organization and the International Labour Organization estimated that long hours caused 745,000 deaths from stroke and heart disease in 2016 (Pega et al., 2021).

Picture the worker

Picture someone who puts in fifty hours to pay a mortgage on a house they barely see, to feed children they barely hold. They aren't lazy and they aren't stupid. They're doing everything the arrangement asks, and it pays them in money for gains that could've been paid in time. It pays them less of the money too.

A 22-hour week sits inside the range productivity already pays for: about 24 hours from the gains since 1979, about 15 from the gains since 1948. It leaves room for jobs such as nursing and childcare, where output per hour can't rise as fast. The path runs in steps: a 32-hour, four-day week at current pay, then 28, then 22.

Where this could be wrong

Comparing countries can't show which way cause runs. Part of the pattern could be part-time work pulling average hours down, though holding the part-time share constant barely moves it. The trials, not the country comparison, carry the causal claim, and several of those figures come from company or consultancy reports, not peer-reviewed studies. The 24-hour and 15-hour weeks are arithmetic, not forecasts. And New Zealand's lesson stands: a shorter week designed by management alone can turn into a harder one.

What you can do

Nobody hands this back. Every cut in the working week we've ever had, someone asked for. Iceland's came through public-sector trials, independent evaluation, then unions using the evidence. Britain's came from employers who tried it and mostly kept it. America's 40 hours came through law.

So pick the door nearest you. Show your boss the British trial and ask for six months. Join your union and ask what it's bargaining for besides pay, because almost every claim we make is for money and hardly any is for time. Next time your MP talks about productivity, write and ask who gets it. Then, if you can, take the time you already have: go home when the work is done, and don't apologise for it.

A woman on a Kellogg's line in 1932 could've told you what it's for. The machines did the work. The gains belong to the people who produced them. That question was never economic.

The machines kept working. Who got the evening?

Sources

  1. Bivens, J., & Mishel, L. (2023). The productivity–pay gap. Economic Policy Institute. https://www.epi.org/productivity-pay-gap/
  2. Delaney, H., & Casey, C. (2022). The promise of a four-day week? A critical appraisal of a management-led initiative. Employee Relations, 44(1), 176–190. https://doi.org/10.1108/ER-02-2021-0056
  3. Fan, W., Schor, J. B., Kelly, O., & Gu, G. (2025). Work time reduction via a 4-day workweek finds improvements in workers' well-being. Nature Human Behaviour, 9(10), 2153–2168. https://doi.org/10.1038/s41562-025-02259-6
  4. Haraldsson, G. D., & Kellam, J. (2021). Going public: Iceland's journey to a shorter working week. Alda and Autonomy. https://autonomy.work/portfolio/icelandsww/
  5. Huberman, M., & Minns, C. (2007). The times they are not changin': Days and hours of work in Old and New Worlds, 1870–2000. Explorations in Economic History, 44(4), 538–567. https://doi.org/10.1016/j.eeh.2007.03.002
  6. Hunnicutt, B. K. (1996). Kellogg's six-hour day. Temple University Press.
  7. Keynes, J. M. (2008). Economic possibilities for our grandchildren (1930). In L. Pecchi & G. Piga (Eds.), Revisiting Keynes: Economic possibilities for our grandchildren. MIT Press. (Original work published 1930) https://doi.org/10.7551/mitpress/9780262162494.003.0002
  8. Kivimäki, M., Jokela, M., Nyberg, S. T., et al. (2015). Long working hours and risk of coronary heart disease and stroke. The Lancet, 386(10005), 1739–1746. https://doi.org/10.1016/S0140-6736(15)60295-1
  9. Knight, K. W., Rosa, E. A., & Schor, J. B. (2013). Could working less reduce pressures on the environment? A cross-national panel analysis of OECD countries, 1970–2007. Global Environmental Change, 23(4), 691–700. https://doi.org/10.1016/j.gloenvcha.2013.02.017
  10. Lewis, K., Stronge, W., Kellam, J., et al. (2023). The results are in: The UK's four-day week pilot. Autonomy. https://autonomy.work/portfolio/uk4dwpilotresults/
  11. OECD. (2026a). Average annual hours actually worked per worker [Data set]. OECD Data Explorer. Retrieved October 13, 2026.
  12. OECD. (2026b). Productivity database [Data set]. OECD Data Explorer. Retrieved October 13, 2026.
  13. Pega, F., Náfrádi, B., Momen, N. C., et al. (2021). Global, regional, and national burdens of ischemic heart disease and stroke attributable to exposure to long working hours for 194 countries, 2000–2016. Environment International, 154, 106595. https://doi.org/10.1016/j.envint.2021.106595
  14. Pencavel, J. (2015). The productivity of working hours. The Economic Journal, 125(589), 2052–2076. https://doi.org/10.1111/ecoj.12166
  15. U.S. Bureau of Labor Statistics. (2026). Labor productivity, real hourly compensation and average hourly earnings [Data sets]. Federal Reserve Bank of St. Louis, FRED. Retrieved October 13, 2026. https://fred.stlouisfed.org/

Keep reading