I would be sceptical of any account that suggests that the life of a pre-modern peasant was anywhere a happy, leisurely life, and that this idyll was shattered by ruinous industrialisation. But more plausible is a world where the amount of work done is constrained by the agricultural seasons, by limited markets for both selling and buying marketed products, and by nutritional standards not so far from subsistence. The changes of early modernity (new goods, better transport, stronger and more centralised states) may well have started to change that picture, leading to an increase in working hours.

I think there is general agreement that hours of work rose leading up to and during the Industrial Revolution, at least in Northern Europe. And it is well-known that work hours have fallen dramatically throughout the 20th century. But the quality of the data prior to the 18th century is extremely poor. For one thing, it is very difficult to get data on working hours for people who managed their own time and effort, especially if those people were seldom literate. There could be very large variations in working hours across both time and space, and we would have no way to know for sure.

There is much discussion of an "industrious revolution" in economic history, a period in the 18th century when workers began working more hours per year. Jan de Vries build on the foundation of Gary Becker's work on household division of labour to provide an explanation how it was possible to simultaneously have a "consumer revolution" leading to industrialisation, and yet also stagnating wages. His explanation is that households reallocated their time away from leisure and non-marketable work (like subsistence farming or household chores) and towards marketed work. With the advent of new goods, the value of monetary income increased, relative to both leisure and non-monetary income. Thus, working hours increased, but they did so in response to the availability of goods that households wanted, and the need to earn money incomes to buy them. (This is, perhaps not surprisingly, a very Dutch story about the transition to modern economic growth.)

Joachim Voth wrote an excellent book, Time and Work in England, 1750–1830, checking this hypothesis of increasing working hours. He checks this by gathering data from witness accounts in English criminal trials, wherein people report (on penalty of perjury and without any consistent reason to lie) what they were doing and when during their testimony. Voth finds that working hours increased enormously leading into the 19th century, to a high of around 3500 (!!!) annual hours for full-time working men, which far higher both than in 1750, and of course than today. Much of the extra working time comes from working extra days, rather than the work day getting longer. Though, as Greg Clark points out in his review of the book, Voth has a predominantly urban data set with very few agricultural workers in his sample, and thus his conclusions for that sector rest on quite thin grounds.

In the 20th century, meanwhile, working hours have fallen and fallen, and our data is obviously much better for this period. Huberman and Minns' "The times they are not changin’" (2007) charted this change across 1870-2000, and the overall picture is a fall from 3000+ hours in 1870 (3484 in Belgium, 67 hours a week, 52 weeks a year!) down to less than 1800 in most countries by 2000. (Belgium down to 1547, substantially less than half!) That's a pretty dramatic shift.