[returningfory2]

There's a whole subfield of economics that studies clustering in cities ("agglomeration"): https://en.wikipedia.org/wiki/Economies_of_agglomeration There are lots of benefits.

Also, my own take is that the high rent in NYC is sort of proof that the quality of life is high. Or at least, NYC is desirable. People are willing to pay a premium to live there, which is a strong signal of their preference.

[zamadatix]

I think it's a situation akin to HOAs where there are absolutely people who prefer being in an HOA but it has this feedback loop which results in significantly more people being in an HOA than would prefer to be just because there are limited options and undoing an HOA is higher friction than new construction including one.

On one hand this is still preference. They pick to be in a city over the other options available. On the other hand the other options aren't available because enough people are already interested in centralizing life choice options into a city and so it just drives that feedback loop over and over as more people choose where the option of the day is rather than what they'd like. The only thing holding this loop back from runaway is large cities eventually seem to have population growth fall behind cost of living growth and that stops the runaway for the particular city.

Perhaps more simply: the immediate and big picture preference often don't align and this misalignment further drives a larger gap in those two preferences over time until the cost to scale the city finally becomes too high.