It's probably not a full 20% loss, but I think your point still stands. To expand a bit:
It seems to me that the only way they could work a day less and not affect the companies productivity is if those employees did nothing during that day anyway. Say that in a work week an employee produced 100 units of work (feel free to substitute this with your favorite productivity microbenchmark). Thus, with the 5 day work-week each day the employee must produce around 20 units of work. Say you reduce that to 4 days, so now each day the worker must produce 25 units of work to be as productive. It follows that in order to work one less day per week the worker must work 25% more per day or overall productivity will suffer.
Imagine that your company implements something like this, suddenly per-day expectations go way up. Can you complete that meeting at 125% speed? Will your builds finish 25% faster? Can you increase your typing speed by 25%? Can you dig that ditch in 3 hours instead of 4? Can you clear tables 25% faster? I don't think anyone wants the consequences of this.
The only way I can remotely see this working out is for the employers of salaried info-workers in high profit-margin industries. For someone employed in a manual or skilled labor position it may be impossible or outright dangerous to attempt to complete tasks at 125% speed. I could however see this being a neat benefit that employers like the Silicon Valley types offer to outbid the competition, but I don't think this is practical for the economy in general.
>It seems to me that the only way they could work a day less and not affect the companies productivity is if those employees did nothing during that day anyway.
An alternative would be that the 4-day schedule boosted employee efficiency by enough to make up for missing day 5.
Stress often degrades performance, particularly when it's chronic. I don't find it incredible that giving employees an extra day for their non-work duties/joys might increase their efficiency 20%.
>I don't find it incredible that giving employees an extra day for their non-work duties/joys might increase their efficiency 20%.
Note that the employees must improve their efficiency by more than 25% for this to benefit the employer, and that they must do this every single day.
I can easily imagine that Monday mornings would be much more pleasant with a 4-day work week, but beyond that first morning I think it would increase stress. If every day I came to work knowing that my employer expected 125% out of me that day I think I would burnout much faster.
I can run 1 mile much faster than I can run 3 miles. And that holds true day after day.
Another way to look at things is to use the 4-day week as the baseline. Add another day of work and people can only consistently perform at 80% without burning out.
Now, I'm not saying that I believe this is the case--that most companies can cut down to a 4-day week without hurting productivity long-term. Just that your argument isn't intuitive or obvious on its face, and I could see the data going both ways, or simply being a mixed bag.
>I can run 1 mile much faster than I can run 3 miles.
This ignores the fact that employers don't pay you for how fast you run, but actually by how far you run in a day. Only the employee cares about how fast they can run.
Work is not divided into even units completed at a constant linear rate over the course of the day. It's a wide variety of different actions, and the time taken to complete each will vary wildly depending on tiredness, stress and morale.
Some tasks may take the same amount of time regardless, but some can drastically reduce. A well rested developer may realise the solution to a bug in a few minutes that would take them hours to solve on less sleep. Or indeed, avoid causing the bug in the first place.
>Work is not divided into even units completed at a constant linear rate over the course of the day
You are correct, work is not distributed evenly throughout the day. But this doesn't matter to the employer. Your employer could care less how fast you complete a "days worth of work", all that matters is that each day you complete a "days worth of work". For employers of mostly hourly workers, this will change to an "hours worth of work", but the concept remains the same: the rate at which work gets done is unimportant next to the amount of work that gets done.
The criteria you give as affecting the productivity of workers is tiredness, stress, and morale, all of which I would assert decrease throughout the day. Wouldn't it then be more productive then for the employer to simply send the employee home earlier once they have lost their productivity? I think so, but employers would start to wonder why they ever paid for those non-productive hours in the first place, and that might have more sinister consequences for the worker.
It's not obvious to me that taking a day out of the work week is the best way to approach this. I think just talking with your employees to see how they're doing then adjusting their hours accordingly would be vastly more effective.
Havent RTFA but I heard an interview with the CEO where he said he read that the average employee is only productive for 2-3 hours per day. Thats why he decided to give this a try.
>...the average employee is only productive for 2-3 hours per day
This sounds true to me, so let's assume that it is. If the average employee was unproductive for most of their day it stands to reason that employers would have a great incentive to only employ workers during their productive hours. Assuming the average employee is only productive for three hours a day, the employer is overpaying their employees by about 60%.
The problem seems to be how to eliminate the unproductive hours in a way that also benefits the worker. The way I see things working out is employers asking more of their employees for a smaller amount of compensation. Great for the employer, but I really don't think this is the kind of worker empowerment people are hoping for.
"It seems to me that the only way they could work a day less and not affect the companies productivity is if those employees did nothing during that day anyway."
OR, if work-disrupting absences (appointments, etc) were all scheduled for the weekday off, rather than being distributed through the week and interfering with meetings, collaboration, etc.
It seems to me that the only way they could work a day less and not affect the companies productivity is if those employees did nothing during that day anyway. Say that in a work week an employee produced 100 units of work (feel free to substitute this with your favorite productivity microbenchmark). Thus, with the 5 day work-week each day the employee must produce around 20 units of work. Say you reduce that to 4 days, so now each day the worker must produce 25 units of work to be as productive. It follows that in order to work one less day per week the worker must work 25% more per day or overall productivity will suffer.
Imagine that your company implements something like this, suddenly per-day expectations go way up. Can you complete that meeting at 125% speed? Will your builds finish 25% faster? Can you increase your typing speed by 25%? Can you dig that ditch in 3 hours instead of 4? Can you clear tables 25% faster? I don't think anyone wants the consequences of this.
The only way I can remotely see this working out is for the employers of salaried info-workers in high profit-margin industries. For someone employed in a manual or skilled labor position it may be impossible or outright dangerous to attempt to complete tasks at 125% speed. I could however see this being a neat benefit that employers like the Silicon Valley types offer to outbid the competition, but I don't think this is practical for the economy in general.