Programmers don’t become more productive when they move to Silicon Valley or Seattle or NYC.
Why do you believe this? I definitely became a better developer when I moved to NYC. How do you know that everyone else didn’t either?
“programmers” is too large and diverse a group to meaningfully discuss. For this purpose, it’s sufficient to say “_SOME_ programmers become more productive when able to interact in-person with other workers and experts, which are currently concentrated in a few cities”. And that is a pretty easy claim to make—at least xepo and I assert it to be the case for ourselves.
I’ve done a lot of hiring, engineer evaluations, and related attempts at productivity-measurement for a very large software company. I can say with certainty that there was good evidence that the enforced shift to WFH was a step-change loss in productivity, only some of which has come back. I will also say that it’s NOT evenly distributed—some teams and individuals did recover quickly (and even benefitted). The median and mean was quite negative, though. Standard caveats: measurement is based on imprecise proxies, and Goodhart may have made it even more variable: it was a visible excuse for a performance drop, rather than trying to game the metrics to look good.
“programmers” is too large and diverse a group to meaningfully discuss. For this purpose, it’s sufficient to say “_SOME_ programmers become more productive when able to interact in-person with other workers and experts, which are currently concentrated in a few cities”. And that is a pretty easy claim to make—at least xepo and I assert it to be the case for ourselves.
I’ve done a lot of hiring, engineer evaluations, and related attempts at productivity-measurement for a very large software company. I can say with certainty that there was good evidence that the enforced shift to WFH was a step-change loss in productivity, only some of which has come back. I will also say that it’s NOT evenly distributed—some teams and individuals did recover quickly (and even benefitted). The median and mean was quite negative, though. Standard caveats: measurement is based on imprecise proxies, and Goodhart may have made it even more variable: it was a visible excuse for a performance drop, rather than trying to game the metrics to look good.