We find … better managers tend to match with worse workers, & vice versa. This stands in contrast to our estimates of the production technology, which reveal that if the firm were to positively sort, productivity would increase by 1 to 4%" https://www.nber.org/papers/w27006
-
-
Replying to @robinhanson
How likely is it that this has something to do with Berkson's Paradox? (Bad workers+bad manager -> somebody gets fired, so only good+good, good+bad, or bad+good exist)
2 replies 1 retweet 12 likes -
Replying to @imhinesmi @robinhanson
Good workers plus good managers probably cost more than the productivity bonus, if that's only 4%.
1 reply 0 retweets 0 likes -
Replying to @argyriou15 @imhinesmi
The suggestion is just to resort the same people, not hire new people.
1 reply 0 retweets 1 like -
Why would that work? Surely at the level of the firm you have same problem: mixing groups with low prod with other groups with high prod on same project should reduce overall output. The worst performing group can hold up whole project. You really need to sort into diff firms
1 reply 0 retweets 1 like -
worst performing group wont hold up everything, because the other groups will be forced to do their job for them
1 reply 0 retweets 0 likes -
Yes, at a cost of productivity. Same thing is true in individual teams. Point is that it's not clear sorting into high and low prod teams is a net benefit if it only means that the high prod teams have to step outside their institutional expertise to do work of other teams.
1 reply 0 retweets 0 likes -
the paper cites 1-4% improvement for sorting its paywalled so i cant look at their methods, but presumably the authors were aware that making people work outside their expertise in inefficient
2 replies 0 retweets 0 likes -
Yah, but the question is if they did this by comparing companies which kept the low productivity workers and managers working together at the company. I don't see how you could even hope to do this and if so then why assume it can all be done by sorting *inside* company.
1 reply 0 retweets 0 likes
i dunno, buy or pirate the paper if you want to look at their methods it seems plausible enough that total productivity is proportionate to team quality times manager quality and 2*2+1*1>2*1+1*2
-
-
And I’m saying that all the reasons for thinking that is plausible are reasons to think that total productivity on a project is given similarly by the product of team productivity by teams on the project.
1 reply 0 retweets 0 likes -
the paper was talking about clothing factories, so "different team" probably means a different production line different production lines usually dont interact, so their efforts are additive managers cause each worker to be x% more productive, so theyre multiplicative
1 reply 0 retweets 2 likes - 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.