Hacker News Reader

> They would frequently say how they are not interested in "part time" work. I'd counter that this is full time work, with efficiency higher than the full time work, because people can focus on intellectually hard problems only for a short while sustainably.

Is it really true? Can you prove it? I mean you can argue that, but if you're absent 20% of the time compared to other workers, is it true that your value is still the same because you're so much more productive? Maybe yes, but can you prove it to an employer? You're asking them to take a risk on supporting unfamiliar approach - which their familiar approach probably worked for them for years and they are fine with it. What do you offer them to justify taking this risk? I mean, maybe you are so spectacular that employing you is worth any risk. But naturally most people aren't that exceptional, by definition. Their experience shows 5-day weeks works great for them, how much better would be 4-day week to justify the risks?

> Civilization as a whole would benefit as a result if there were more people putting their brains to problems they think matter.

Is there any proof that there's significant marginal increase compared to thousands of existing research institutions that have tens of thousands of very smart people already spending years attacking practically every important problem? Would amateurs spending one day a week on side projects significantly change the picture here - and offset the above-mentioned professionals not spending one day a week on their area of expertise (instead doing their hobbies in turn)? I am not sure this is that obvious.

Pretty good questions, thanks.

It's generally accepted that longer work hours decrease cognitive performance (see eg. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2727184/), but the base is usually 40h a week.

It's hard for me to prove the first point in relation to me personally: basically, I've done an experiment where I've worked mostly 6h days (still 5 days a week) and have been consistently rated higher than my coworkers at a software company with pretty good developers overall. OTOH, when I worked usual 8h days, I mostly compared as average. But that's me comparing myself to others and in distinct time periods with different projects and opportunities, and different life circumstances. For an actual proof, I'd need to compare myself with shorter work week to myself with a longer work week, working on the same projects at roughly the same time and similar external circumstances.

So, it is certainly anecdotal, and I am aware of that.

However, in a similar vein, they don't have proof that 5-day weeks work well for them. They just "work" for them because that's the law in most places and they never tried anything else. Actual proof would require a number of long-running experiments, so it's pretty much out of the question for any one company.

As for the second point, my experience and opinion of research institutions is not as high as yours. What you get in academia are ~5% of very smart people very much interested in the subject, another ~15% of very smart people not very much interested in the subject matter, and the rest who simply went with the flow. I've tried to find data for "best in class postdoc student retention", but nothing turned up, so I can't back up my out-of-my-behind numbers with anything. Other than adding that if you look at most research papers, a lot of it is simply bull (i.e. not research at all or just one thing repeated ad-nauseam with slight modifications, for the purpose of getting appropriate academic points to keep grants and funding going).

You might also note that I've argued for a shorter work day as well, thus allowing for significantly more non-work work to be done by amateurs than just 1 day per week.

If the premise of increased productivity for shorter work weeks applies to research work as well (can't find any study on this in particular), then we'd see no drop in their productivity, and only an increase on whatever we get from amateur work.

But, for now, these are mostly "thought experiments", and surely not "science". It's mostly a simplistic, idealized view of where the civilization should go, and it's hard to prove benefits either way.

> It's generally accepted that longer work hours decrease cognitive performance

Of course, it is more or less obvious that there's a point where working more is not going to produce marginal improvement anymore. Physiology is a harsh mistress. But claiming 4 days per week is that point is pretty bold.

> You might also note that I've argued for a shorter work day as well, thus allowing for significantly more non-work work to be done by amateurs than just 1 day per week.

To me it looks like replacing workhours done by professionals at designated workplaces by workhours done by enthusiastic amateurs using their own resources. It may be that the latter model is more efficient, but that is a rather bold claim that needs some proof, and not at all intuitive. What amateurs win with anthusiasm, they can easily lose with poor resource base, organization and cooperation issues. Some things, to be sure, lend itself to crowdsourcing easily, but some - and a lot of fundamental science seems to be in that area - require very specific large organizational effort and significant resources. If, say, an average pharma worker works 1 day less to pursue their hobby of writing gcc patches, and an average software engineer works 1 day less to pursue their hobby as a amateur chemist, I am not sure we'd get cure for common cold or order of magnitude better compiler. Maybe it will happen, but it's not obvious to me that it should.

> If the premise of increased productivity for shorter work weeks applies to research work as well

That, again, is a big if. Especially the point where the return starts to be negative.

I've done a 4-day work-week, and I was at least as productive, and definitely happier. I found myself screwing around less at work, because I had less time and there were things I needed to get done. And then, when I was really engaged in the project I had at work, I had enough slack that I was able to 'over'-produce for a short while (4-6 weeks), which was not sustainable but I didn't care because I loved it. When I'm working 5 days a week, it doesn't leave room for that kind of love; I spend my too-short weekends recharging and begrudging that I don't have enough time for the things I really want to do.

> I've done a 4-day work-week, and I was at least as productive, and definitely happier.

I believe you were happier. But how did you measure the productivity? Are you sure it was objectively the same? I don't mean it as criticism for you, but people are notoriously unreliable at evaluating their own productivity (something 85% people thinking they are above average, and so on :) and productivity in general is not the easiest thing to measure. So, how did you ensure that and how you ensured your case is not "my productivity was less but turns out that was enough too"? I would be glad to have definite proof of this, but self-reporting is not a very reliable productivity indicator, unfortunately...

What if the proof is in metrics like lower rates of burn out, less sick days taken, higher morale, etc.?

Definitely possible. However, not on the data set of size 1. If we get a study with sufficient statistical power and proper controls, sure, I am ready to believe it when I see it. Several studies, of course, would be better, since retraction and non-confirmation numbers are high enough now so that a single study does not constitute a definite proof, but it would certainly be a serious support to the idea.

> Is it really true? Can you prove it?

Let me flip this around: Employers are terrible at measuring productivity of certain kinds of knowledge workers, so they just try to squeeze as many hours out of us as possible to compensate.

True, measuring productivity is notoriously hard, especially in areas like software engineering. It's largely an unsolved problem. So we have to rely on intuition, tradition and gut feelings, unfortunately. All of those suggest "let's work less by 20%" is probably not the most obvious way of improving productivity. I mean, of course, there's a place where working more does not yield more output. But 4 days a week doesn't seem to be that point - if fact, it seems pretty luxurious point to be in, and, again, intuitively, it looks like working a bit more would still produce marginal increases. It very well may be that the intuition is wrong - but to claim it's the case one need some serious proof. "I feel happy not working" doesn't really cut it - I'd probably feel happy working one day a week, for the same money, but my productivity would hardly be the same, so I don't see how I could sell such arrangement to my employer.

People argue it here all the time but I'm certainly skeptical. The hardest intellectual work of my life, my undergrad at MIT, I spent far more than 8 hours a day, 4 days a week on. The idea that all the undergrads there burning the midnight oil are just being foolish and don't they know they could get all their work done in less time doesn't pass the smell test for me.

Most people don't do MIT level intellectual work...

Such an effort is also quite different from salaried work. When you are taking an undergraduate course you put a lot of effort into a finite project. You tend to have the energy reserve for something bounded like that, but it does come at a cost. Usually, at the age where you would enrol in an undergraduate program, the gains outweigh the cost, and the cost is often lowered significantly by the intrinsic motivation of working on your own interests.

With salaried work the only effective limit is your retirement. Working more simply means more income, but if that extra income isn't needed, or doesn't net you enough extra benefits compared to simply having time off, the cost of being mentally engaged for such a large part of the week just doesn't weigh up to the gains.

Sure, if your work is so engaging and rewarding that work itself is a pleasure, than it might pay off. But for most of us it just doesn't work that way. Work often means doing things were mostly others set the agenda, and while you may be good at what you do and find motivation in doing it, I've found that it rarely means that you can do it with the same sustained level of energy and quality for more than four eight hour days — and even that isn't a given.

Yeah, but that is work towards a tangible goal with a fixed time to finish. It is easy to make a big push in those types of situations. In your vanilla work life, this stuff is no longer there anymore. You can't work like that forever with no payoff, delayed gratification only stetches so far.

I am VP engineering in a 100 person company with 30 engineers. Noone is asked to work more than 40 hours and <5% do actually work more. I will happily have someone work 4 days 8 hours and 30% of our devs are in such a model. But I would not go below this (especially not less than 4 days in the office). Devs aren't working in total isolation, but are heavily involved in their cross functional product teams. Reducing availability further has an above linear hit on team productivity as colleagues become more and more blocked in their own tasks.

So: just saying this is more complex than looking at increased per-hour productivity of a single person

Sure, though a solution is simple: just have everyone work the same shorter hours :)

I've actually worked in primarily remote company, and the challenges there are even greater: no offices, hand-offs and hand-overs have to be well documented or you are wasting time, timezone differences, discussing architectural documents over video and voice calls...

Basically, I wasn't asking for full flexibility in choosing one's hours: I think that's harder (for all the reasons you bring up) than just shortening the hours.

> especially not less than 4 days in the office

Why? Remote work is a thing. If your processes demand having people on premises all the time, your processes may need improving. If a programmer want to work one day from home to cut down on their commute and be able to focus on a task, why would you want them in the office unless they aren't trustworthy? For colleagues to break his concentration when they need him?

Meetings can be scheduled, work coordination can be done electronically, and you surely have something like Mattermost or Slack set up?

I am in a not too dissimilar position to yourself (role/team size). I am working 4 days one week, 5 days the next (90% hours). It work out to be a nice compromise. The cross functional does not take too much of a hit, but the boost in productivity (IMHO) is still present.

I have some push back about the arrangement from people two or three links away from me and my team output and I have not got metrics to prove that this is effective.

Do you have any push back about your devs - 4 days 8 hours and 30% ? Do you have any other evidence/metrics for the benefits ?

> "... 12h days, ..."

I recently read Sebastian Junger's Tribe, which had many memorable takeaways for me.

One, in particular, was the assertion that, historically, subsistence farmers and the like generally only needed to work, on average, 12 hours per week.

While I know modern farming practices require copious amounts of work, it is nonetheless an interesting idea.

Imagine all the remaining time to just sit in sun or socialize.

Progress, what is that?

> One, in particular, was the assertion that, historically, subsistence farmers and the like generally only needed to work, on average, 12 hours per week.

I’ve seen this claim before, and it typically revolves around careful definition of the word “work” rather than any fundamental difference in how we structure activities and priorities. iirc altering the definition to “any work that provides utility” puts the time at 30-40 hours a week. Still very much for a 30 hour/ 4 day work week, but I’ve found that particular argument to be very weak.

Granted, the west never seems to count reproductive labor when talking about work, which is itself a problem. Most Americans spend the vast majority of their waking hours in some form of labor.

> the extra time I get would not be spent in pyjamas watching netflix: it would be quality time with my family

why would it matter where the extra time goes. you're saying that caring for your kids is more important than a single guy working at the gym or watching netflix. everyone has his priorities and none are more important than others.

You are right: I should have not been so judgemental. Watching netflix might be a net win for someone because of their improved mental state (relaxation, self satisfaction, maybe even education, hobby or as you point out, whatever...). For me it's not, but I understand people do differ.