A tale of two suburbs

I recently rode a commuter train through San Mateo County, California, which is just south of San Francisco. San Mateo is one of the most expensive places to live in all of America, and the view from the train made it easy to see why.

Zoning restrictions in the Bay Area are almost unbelievably counterproductive. Looking outside the train window, you see lots of valuable land right next to the major commuter rail line that is used for unproductive purposes. While there are a few newer apartment buildings, mostly you see lots of ugly low rise buildings, including run down ranch houses, one story warehouses, strip malls, etc. It’s a bleak sight, and a tremendous waste of valuable real estate.

While riding in the train, I thought about the Orange Line subway through Arlington, Virginia, which is the county just south of Washington DC. It’s a nice comparison, because each country is a beacon for high paid professionals working in some of our most dynamic labor markets.

But there’s a big difference between San Mateo and Arlington counties. While the latter does contain lots of low-rise residential neighborhoods, it also allows dense high-rise development within walking distance of the subway line. According to a study by Emily Hamilton, this has led to a surge in apartment construction:

Development along the Rosslyn-Ballston Corridor is visible (most strikingly where I live, in Ballston) as is development along the other transit corridors. In the 1970 Census, ahead of the Orange Line’s inauguration, the county’s housing stock included about 30,000 detached single-family houses, a number that has remained steady in the decades since. But the stock of other types of housing has more than doubled from about 41,000 units to nearly 88,000 units. This infill apartment construction has allowed the county’s population to increase by 60,000 residents in this 50-year period – without expanding the area developed at all.

This contributes to the DC area being substantially cheaper than metro areas with lots of professional jobs but tighter building restrictions.  Not surprisingly, they are showing greater population growth:

Four US metropolitan areas in particular – Los Angeles, San Francisco, New York, and Boston – have become increasingly expensive, thereby pushing out low- and middle-income families as higher-income in-migrants outbid them for a stagnant supply of housing.  Two other ‘superstar regions’ with high productivity and high average incomes – DC and Seattle – are doing a better job of accommodating new demand for housing with new housing construction. Relative to the other four superstar cities, they are losing domestic residents at much lower rates. DC, Seattle, and some other growing cities across the country are suffering from their own housing shortages, but not on the same scale as those places with the most severe impediments to housing construction.

In case you think this comparison is unfair because there’s something special about the geography of the Bay Area, or reflects the effects of Silicon Valley, keep in mind that San Diego also has tight building restrictions and extremely high housing prices.  It’s a general problem wherever you combine lots of well paid jobs and tight restrictions on building. It impacts all of coastal California, as well as New York and Boston.

And even the DC area is far from perfect, with many more building restrictions than would occur in a truly free market.  I cite Washington DC rather than a city like Houston, because DC is more similar to other affluent metro areas in the northeastern US and California.

Nor has all this apartment construction led to deterioration in the schools:

Contrary to some of the received wisdom on high-density residential construction in the U.S, Arlington has a highly-rated school district (one school ranking organization ranks it second in Virginia, behind only the city of Falls Church, Arlington’s neighbor to the west with fewer than 15,000 people)

Scott Alexander recently argued that increased density would push up property prices.  While there is undoubtedly a positive correlation between real estate prices and density, the causal implications depend on why density has increased.  If density rises because of a booming job market in a specific location, then housing prices will rise.  If density rises because regulatory changes allow for more apartment construction, then prices will usually decline, or at least rise more slowly than otherwise.  Here’s Hamilton:

Between 2012 and 2018, rents in DC actually rose slower than inflation. Virginia Beach, VA, was the only other large metropolitan area in the country for which this was true.

The Hamilton study is full of interesting information, and is well worth reading.  Here’s another example discussing northern Virginia’s Fairfax County:

Many of Fairfax County’s peers in California, Massachusetts, and New York are permitting virtually no apartment construction. For example, between 2000 and 2020, the stock of apartments in Marin County, just north of San Francisco, didn’t increase at all. Long Island didn’t do much better, with its supply of housing other than detached single-family housing increasing by about seven percent in those two decades. By contrast, Fairfax County’s increased by about one-quarter.

If you don’t build it, they won’t come.

PS.  Hamilton provides a map showing how the recent rise in density is concentrated along the Orange Line subway (yellow dots):