I am unsure that this is true in all circumstances.

So the net result of these factors is that a company must work at maximum efficiency, and maximum efficiency comes when all developers are working 40+ hours a week.

I know for a fact that this isn't always true - since I've repeatedly seen teams of developers doing 45+ hour weeks become more productive by every metric we had to hand by dropping their working week to 40 hours (with only about 6 hours a day of that being coding).

I talk about this in a little bit more over here before (http://news.ycombinator.com/item?id=3883362)... copy/paste below.

-----

There is something deeply broken about equating hours to productivity.

It's been my experience that folk are very good at deceiving themselves about their productivity (myself included :-)

One team I worked with had a serious problem with overtime. They were putting in stupid hours and it was showing in the quality of work going out. So I ran an experiment where we all agreed to work "normal" hours for six weeks.

I was "only" working about 45 hours a week at this point, when other people on the team were regularly working 50-60. I was relatively young, didn't have any family pressure, enjoyed my work and felt very productive doing those hours. I wasn't one of the people with a "problem" as I saw it. We were running the experiment for the other folk on the team.

In the experiment we dropped to a 40 hours week (6 hours coding per day, 2 hours for breaks, meetings & lunch). After a couple of weeks adjustment my productivity went way up. I also felt a lot better in myself - generally sharper and more on the ball.

People seem to have quite a wide bad of "this feels okay" that subsumes the much narrower "I'm performing at my best".

Also people don't jump from a 35 hour week to 60 hours a week. It creeps up a few minutes at a time as pressure increases on the team. People have enough time to adjust to it being "normal" and don't notice the drop in productivity that goes with it.

Currently I work roughly 25-30 hours a week and am just as productive by all metrics that I have available to me as when I worked 40-50.

I would strongly urge people to experiment. Pick some metrics, try working shorter hours for a month, see what happens.

(The only caveat I would add is that with folks doing silly hours - anything over 50 I would say - there is often a couple of weeks where things go to hell as the body adjusts. On the team from the story practically everybody caught a bug and felt crap for the first week or so before productivity rose again).