Hacker News Reader

Comment by getoj | Hacker News Reader

Posting to a top-level comment: The productivity claim has been removed from the Japanese report[0]. An erratum at the bottom of the page dated November 8, 2019 says this:

"In the announcement dated October 31, one of the listed "improvements" from the 2019 Summer Work-Life Choice Challenge was an increase of 39.9% in labor productivity (sales revenue per employee) in August 2019 compared to August 2018, with a graph below.

"While this number is factual, it is not solely the result of this challenge, and was achieved due to a number of different factors.

"To avoid misunderstanding, we have removed that claim from the above summary of the direct effects of the challenge."

[0] https://news.microsoft.com/ja-jp/2019/10/31/191031-published...

View on HN

> In addition to reducing working hours, managers urged staff to cut down on the time they spent in meetings and responding to emails.

> They suggested that meetings should last no longer than 30 minutes. Employees were also encouraged to cut down on meetings altogether by using an online messaging app (Microsoft’s, of course).

If I am told I am part of an experiment, I would do my best to make it a success and make my 3-day weekend permanent :)

I've wondered about this. In some limited areas, it seems like change helps keep things fresh and people off of autopilot. From that perspective, it's possible that if prolonged the results would level out.

And, yes, I say this as someone who would rather have a four day workweek.

Autopilot is nothing bad. For repeative tasks that cannot be improved with reasonable amount of ressources, you are better off in autopilot. It saves energy for critical times.

There's even a name for this: Hawthorne Effect From Wikipedia: "The Hawthorne effect refers to a type of reactivity in which individuals modify an aspect of their behavior in response to their awareness of being observed."

https://en.wikipedia.org/wiki/Hawthorne_effect

(Yes, the Hawthorne Effect is more general than the specific case here where people want to influence the outcome in a particular way, but it's still cool that this is known to be a thing)

That's not the Hawthorne effect, that's people responding to incentives.

> was achieved due to a number of different factors.

How surprising that something is used as a single variable explanation by a PR department. I for sure have never seen that before. Glad to know they backtracked on that, but it's too late since the news has been out there already for a long time and this CNN article perpetuates it.

I would not blame solely company PR departments for this, often the press exaggerates claims for clicks.

I remember working on a non-profit that provided philanthropy advice to millionaires. Once we hosted a small conference and we ran a very informal poll about the said millionaires satisfaction with their own philanthropy.

The result was something like 32 out of 40 wanted to donate more money annually than what they were currently donating (as opposed with “satisfied” with the amount or wanted to “donate less”).

We mentioned the result to a journalist covering the event and of course the newspaper headline was ”80% of Brazilian millionaires want to donate more”. No mention that it was a poll during the event with only 40 people that were highly selected to people wanting to donate more (the purpose of the event was teach how to donate well). On the contrary, someone reading would have the impression that our organization funded a well-done proper research covering a significant sample of all millionaires in Brazil.

That was not our PR, that misleading headline was solely the journalist creation. The event itself was pretty boring, so they went with that headline to justify their time investment in covering it I think.

> I would not blame solely company PR departments for this, often the press exaggerates claims for clicks.

PR specifically knows how the press reacts and tailors their messages accordingly. Also, there's a lot of cross-pollination between people working for PR and people who work in the press.

True, that said, I think it's a good data point towards a 4 day workweek none the less. I feel most people would have predicted the outcome to be negative or hurt by the four day workweek. So we don't know if it's sufficient to increase productivity, but it clearly isn't sufficient to decrease it either. And that's already pretty promising data.

I don't think that you can make that conclusion. How do you know that a 5-day week with the other interventions still in place isn't better than a 4-day week? Or how do you know that keeping everything the same and moving to a 4-day week wouldn't have a negative/neutral effect?

> How do you know that a 5-day week with the other interventions still in place isn't better than a 4-day week?

That we don't know.

> how do you know that keeping everything the same and moving to a 4-day week wouldn't have a negative/neutral effect?

I don't know why we'd want to keep everything the same. What we do know here at least is that a four day work week implemented as it was in the experiment did not result in negative or neutral effects. And that's already quite the positive outcome which means more experimentation is mandated as it seems there's hope for four day work week being practical and realistically implemented by companies without hurting their output.

> What we do know here at least is that a four day work week implemented as it was in the experiment did not result in negative or neutral effects.

We don't know that, though. Maybe sales would've increase 10x with a 5-day work week and the 4-day week hurt significantly.

I'm saying negative or neutral to absolute growth, not to opportunity.

Basically, what would you have said prior to knowing about this experiment if I asked you: "What do you think will happen if we do this experiment to productivity?"

I think a lot of people would have said probably you'll see a decrease. But instead we saw a 40% increase.

Now I'm assuming Microsoft has YoY data, so I don't know if 40% yearly increase is the norm, above the norm or below and that be interesting to know.

But still, this 40% increase is surprising to what I think I would have guessed and assume a lot others would have too, that you'd see a decrease or maybe neutral.

So I'm not saying it tells us 4 day work week is better then 5 or 6 or 7. But it seems like a viable model that still allows for quite a bit of productive growth. And that's already a good start for it, one I'm a little surprised about.

Cause it means that possibly time worked has minimal impact on productivity, at least within a 1 day margin. It seems other factors might be more important, maybe simply strategic decisions, cutting out of time wastes like useless meetings or discussions, etc.

A 40% improvement doesn't become negative just because you compare it to something else!

The contribution of the 4-day week on its own could be negative for productivity. It’s possible that without it you would have higher improvement to productivity than 40% due to the other changes in the experiment.

You don’t know whether you got that extra 40% by getting 20% from the change to a 4-day week and 20% from the other changes. Or -20% from the 4-day week and +60% from other changes. Or any other combination.

> on its own

Which is why the parent comment explicitly said "implemented as it was in the experiment".

We don't know what it does on its own, but we know that this combo can be applied to similar jobs.

(Well not 'know' for sure after a single experiment but it's useful evidence.)

> I don't know why we'd want to keep everything the same.

To actually find out whether a 4-day week is better than a 5-day week.

Well, that's something we could find out. But to me the more interesting thing to find out is: is there a work process that is as good or better then the current work process in output produced yet allows me to only work 4 days.

Now sure you could say, all productivity boost were due to other process changes and the benefits of this boost should not go to employees, but if you wanted to be more attractive to talent, that be a good benefit, and knowing that your business wouldn't be affected compared to how it's doing currently is the first step to consider this benefit as a viable one.

> I feel most people would have predicted the outcome to be negative

With a company like Microsoft that sits on a monopoly? They could fire all their staff and still sell licenses just fine.

Reply on news.ycombinator.com