I recall seeing at some point, but alas cannot recall where, a study that looked at average productivity per hour vs. hours worked each day/week. If memory serv

I recall seeing at some point, but alas cannot recall where, a study that looked at average productivity per hour vs. hours worked each day/week. If memory serves me, they found that productivity/hr reliably dropped at the beginning and end of the work day, mostly independent of how many hours were worked, and also dropped steadily with increasing hours per week beyond some threshold, up to a (surprisingly low) point where the marginal productivity of an additional hour was actually negative. The end result basically said that 4-hr workdays are less than half as productive as 8-hr workdays, and that diminishing returns make long work weeks quickly useless. Interestingly, per hour productivity peaked fairly low--something like 30hrs/week, I think. So, all else equal, if you're paying people hourly, you get best value for money having more people, working fewer hours in fewer days. On the other hand, if you're paying people a salary, you technically always get more value up to the point of burnout (i.e., negative marginal productivity), but for whatever work week length, you get best value for money having people work a few very long days. That is, 12hr days M-R would be more productive than 8hr days M-S. Unfortunately, I've been unable to find the study in question again; so take the above with a grain of salt, as my memory is certainly fallible.