Hacker News Reader

Comment by jialutu | Hacker News Reader

The working hours is one thing. What I am most curious about is does more hours being put in mean more/better work output; ie does the 996 work culture actually means a higher output, even if by a slight margin?

Personally my opinion is that coding is not like the assembly line where more hours mean more things that's produced, but more lines of code can mean absolutely nothing (sometimes can even be a negative). However, looking at the recent progress in China, I am wondering what percentage of that is attributed to the 996 working culture, or is it to do with something else?

View on HN

9am-9pm really is the period of presence in office. It's always that way in countries where the stats show high "working hours".

It is also traditional in Chinese work culture to have a nap after lunch, which provides a break and helps with staying later. Dinner is also usually early (6pm) so 9am-9pm days mean a lunch break and a dinner break.

Working 6 days a week has been quite standard in Asia. I remember Korea moved to 2-day weekends only a few years ago, and it's hard to change work culture.

It's probably also compounded by the fact that in China there is no day off for shops, restaurants, banks, etc. Everything is open 7 days a week by default, which is great as a consumer but probably means pressure on staff.

So a 40 hour work week does not mean 40 productive hours. Let's say that you are only truly productive at a flat rate of 40% of your work time, then 40 hour work week means 16 productive hours per week, whereas 996, 72 hours a week means 28.8 productive hours per week.

Now my question is, is the assumption of productive hour ratio real or does it start high and end low? And if my assumption of the flat productive ratio is true, then does it actually yield any good results?

> Let's say that you are only truly productive at a flat rate of 40% of your work time

That's not the case. Hourly productivity decreases at some point.

The law of diminishing returns is at play here. Working longer hours leads to higher overall output up to a point as the marginal increase diminishes.

Then, at some point you are too tired and produces bad software that will cost to fix.

If your employer pays you a flat rate then it is profitable to make you work long hours up to the point when you start screwing up. If they pay you by the hour or if they pay you overtime, rationally they shouldn't push you beyond the point where your hourly output is too low to justify your pay.

Look, I completely understand what you are saying, because it's something I believe as well. However, what I am trying to find out more is whether this belief is correct or not and whether anyone has done any actual studies.

This is the way it is. Anyone who's worked in the industry (or any industry?) knows from experience that this is the way it is. I'm sure that there are data available.

A quick Googling for "productivity vs hours worked" led me to this:

http://bleedingheartlibertarians.com/2013/10/hours-worked-an...

Unfortunately there are quite a few problems in the report: 1. it measures productivity by GDP, which itself is not a great measure 2. it compares Greece vs Germany, and fails to mention that they rely on completely different industries; Greek - agriculture, Germany - manufacturing

But what we are talking about here specifically is the software development industry, where quantity does not always mean quality. So herein lies the problem that potentially a lot of our beliefs that "more hours worked means diminished return" could just be based on bad data analysis.

Reply on news.ycombinator.com