OverTime is the time spent working that makes you less effective in the long run.
As a person works more and more time per day, that person's output will eventually decrease.
Here is an interesting thought from PeterSenge's LearningOrganization book TheFifthDiscipline (ISBN 0385260954 ), adopted to this discussion:
What makes projects successful is the stuff that is needed to build a good team. It's everybody's PeopleSkills that make a team productive, more than any single other factor.
A person can learn some of that in the professional arena. But one also needs family and friends to develop PeopleSkills. Being a husband or a parent teaches PeopleSkills. Being active in hobbies not connected to professional life teaches PeopleSkills.
That said, it is clear that the overall productivity goes down with LongHourWeeks. Not this week's or this month's, not even this year's, but in the long, long run, the productivity decrease may start much sooner than most people expect.
Possibly even the magical 40 is above the optimum for many people. But that's a personal guess.
See also SleepToWork, DeathMarch, WhyDoYouPermitThisToBeDoneToYou, DayOff, OvertimeTheater, and, in particular, FortyHourWeek.
For a contrasting pattern, see ThirtySixHourDay.
The link doesn't have the original article but the Discovery Channel is reporting a study that shows sleep deprivation is worse than alcohol on performance in terms of reaction time.
http://exn.ca/stories/2000/09/18/52.cfm
See also PhilipGreenspunOnOvertime. --PaulChisholm
RapidDevelopment book has a chapter on Voluntary Overtime. Read it online: http://www.stevemcconnell.com/rdvolot.htm
This topic and its many linkages both in support and opposition have employed the IntensifyDownplaySchema. Failure to focus and produce during the first 10, first 40, or any additional hours is far more complicated and involves far more important issues than time on the job whether formally in the office or informally in off-job activities. The fatigue one encounters and the productivity depreciation can be extreme in those who through sleep-deprivation, or wasteful activity off the job can not give 100% to even the first 10 hours!
See MarsPolarLander for an example of an infamous system failure that has been partially attributed to too much overtime.
Moved from WhyDoYouPermitThisToBeDoneToYou
What's really interesting is that I'm morally certain (that means I have no experimental evidence, only years of experience) that long-term overtime makes people more stupid than they'd be if they rested once in a while. Therefore they don't get more done through overtime, they get less. Duh.
Hmmm... I'm glad I found out about this "less is more" approach. However, extrapolating from my current situation, I think cutting back to 40 hours a week still may not increase my productivity enough to get everything done. From now on I plan on only working 20 hours a week so I can get even more work done.
-- JeffShelby
The question is how do hours and results graph. People who work tons of hours aren't necessarily more productive. In my experience it's often quite the reverse. -- RonJeffries
My experience with our current project (which has been running for two years now) is that every time I worked late (meaning > 10 hours a day), I put in some very nasty bug that came back to haunt me. After about 10 hours of coding (sometimes a little sooner or later), my thoughts turn to gibberish. Of course this gets worse if you repeat the exercise a couple of days in a row. -- ThomasMaeder
I usually have that power of choice Jeff was talking about. But, in my case, that's not related to OverTime, but to the fact that I like the complicated, demanding stuff and am good at it. -- AndreasKrueger
I was contracting a few years back at a place where the senior manager used to come round for a 'chat' just when everyone was leaving to go home. So they had to speak to him in their time. Because I was on an hourly contract I simply added his little chat to the invoice at time-and-a-half. I wonder if he ever realised this?
How about some OvertimeProductivityStudies??