Housing construction in the United States has failed to match population growth for decades. Growth in housing availability depends on developer interest in responding to local real-estate market demand, but local governments also influence housing production through land-use regulations. Is the location of additional housing supply aligned with what we might expect given developers’ interest in investing in more expensive, in-demand communities? Or do certain local characteristics undermine the production of new housing?

Using data from 2000 to 2020, I find that municipalities with lower home values and residents with lower incomes, less educational attainment, and more moderate ideological views had less housing growth. These results confirm an uneven distribution of housing construction within US metropolitan areas, paralleling the inequitable distribution of resources between communities. But outcomes for the most expensive municipalities vary. Most communities with the highest demand for development added less than their fair share of metropolitan housing units. They are hoarding resources by blocking housing, likely through land-use rules. The expensive municipalities that added the least are mostly midsize suburbs. All are characterized by higher levels of educational attainment than their respective metropolitan areas, and few have many affordable housing units.