Thanks for the Braess paradox link. I think you could potentially generalize the Braess paradox to talk about where people live. The intuition is that we'd be

Thanks for the Braess paradox link. I think you could potentially generalize the Braess paradox to talk about where people live. The intuition is that we'd be best off if the distribution of demand matched supply, then housing costs from town to town would be identically distributed and all markets would be well utilized and none over-congested. But many people selfishly prefer to live in dense areas because network effects make it more valuable to live in those places, more choice in jobs, more people to interact with, etc. Not only does this cause under-utilization in less desirable locations, it huts affordability-- which people consider a terminal value ('cause housing is necessary for life) independent of the overall utility or efficiency. Though we're OKAY with certain houses or even neighborhoods being out of reach of some, we consider it a problem when whole cities or regions are broadly unaffordable. [It's not clear to me that we should, any more than we consider it a problem that not everyone can afford to live in SwankyHeights. But it's a complicated situation, since there is a reason people want to live in dense areas, like wanting access to jobs or services, making "so go live elsewhere" not a great answer.] Okay, so we add housing. Now the incentives change around in terms of who wants to live where, and there isn't a particular guarantee that the equilibrium prices that result will be lower-- they could well be higher because we just shifted the balance to attract more people from elsewhere. It's a handwave, but over the time span of years, I'd argue that demand for dense cities is more like your transit-to-driving example than your "everyone-is-in-cars" example. People outside of the dense areas are "taking transit", but if you changed the economics they'd prefer to take a car. > More trips usually mean that you actually improved people's lives, which is a good thing. Right but that also assumes overall improvement is the thing people care about. So in a hypothetical where increasing the number of units in a city made every unit more valuable and so made the city less affordable, it would still create an aggregate improvement in utility and so it arguably should be done. But the issue being complaint about wasn't cities doing the most good possible, it was about cities being affordable to the most people. Kind of like complaining about roads because they take up space that could benefit mankind in other ways, vs complaining about roads because they facilitate cars.