Housing Price Inflation Impacts Different Social Classes Differently In The U.S.

People get paid more in big cities, but not enough more for less educated workers to make them more affordable than rural areas and small towns. 

This drives geographic segregation of the nation by education. 

It also transfers a lot of wealth from highly productive big city workers to landlords (even though these groups partially overlap), and reduces the economic incentives to get better educated at the margins (which is where a lot of working class people considering higher education are found). 

The high cost of housing in big cities has many reasons, but excessive land use regulation is an important one.

This analysis supports an economic and political case for YIMBY ("yes in my back yard") approaches to land use regulation.

Also, for what it is worth, it is not obvious to me why the incentive for less skilled people to live in high density v. low density areas changed when it did, although deindustrialization of the U.S. economy (i.e. the erosion of low skilled manufacturing jobs in big cities) is one plausible reason.

People are more productive in cities. As a result, people move to cities to earn higher wages but some of their productivity and wages is eaten up by land prices. 

How much? 

In a new paper Philip G. Hoxie, Daniel Shoag, and Stan Veuger show that net wages (that is wages after housing costs) used to increase in cities for all workers but since around 2000 net wages actually fall when low-wage workers move to cities. . . . 

As I wrote earlier, it used to be that poor people moved to rich places. A janitor in New York, for example, used to earn more than a janitor in Alabama even after adjusting for housing costs. As a result, janitors moved from Alabama to New York, in the process raising their standard of living and reducing income inequality. Today, however, after taking into account housing costs, janitors in New York earn less than janitors in Alabama. As a result, poor people no longer move to rich places. Indeed, there is now a slight trend for poor people to move to poor places because even though wages are lower in poor places, housing prices are lower yet.

Ideally, we want labor and other resources to move from low productivity places to high productivity places–this dynamic reallocation of resources is one of the causes of rising productivity. But for low-skill workers the opposite is happening – housing prices are driving them from high productivity places to low productivity places. Furthermore, when low-skill workers end up in low-productivity places, wages are lower so there are fewer reasons to be employed and there aren’t high-wage jobs in the area so the incentives to increase human capital are dulled. The process of poverty becomes self-reinforcing. . . .

Moreover, as I also argued earlier, even though the net wage is still positive for college-educated workers a significant share of the returns to education are actually going to land owners! Enrico Moretti (2013) estimates that 25% of the increase in the college wage premium between 1980 and 2000 was absorbed by higher housing costs. Moreover, since the big increases in housing costs have come after 2000, it’s very likely that an even larger share of the college wage premium today is being eaten by housing. 

High housing costs don’t simply redistribute wealth from workers to landowners. High housing costs reduce the return to education, reducing the incentive to invest in education. Thus higher housing costs have reduced human capital and the number of skilled workers with potentially significant effects on growth.

From here.