You may have missed the stories, coming as they did over the Labor Day holiday, but most news organizations spent their weekend writing up new statistics released by the Census Bureau on the distribution of income across the nation. The Post had plenty to write about, as the Washington area featured prominently on many of the highlighted lists, good and bad.

I don’t know if most people realize this or not, but the Washington metropolitan area has the highest median household income in the United States. It’s higher than that of Bay Area, significantly higher than the median for Greater Boston and Metro New York. It’s a full $20,000 higher than the Sunbelt boomtowns. And it’s twice as high–remember, this is the median–as the metro areas of the Deep South and the Gulf Coast. This never fails to amaze me. Perhaps I’m wrong, but I think we generally assume that incomes within the US don’t vary all that much by region, particularly between places that one might consider economically similar. That’s not at all the case. My hometown of Raleigh, NC has grown into a new Silicon Valley, chock full of high tech enterprises and a quality of life that Forbes Magazine seems to constantly be declaring enviable. But sure enough, Raleigh’s median income is barely two-thirds that of Greater Washington.

An economist would look at this situation and suggest that, given free labor mobility, people should gravitate towards the highest income locales and away from the lower income areas in a process that should smooth out income levels across cities. This does not actually seem to be happening. Though first in income, Washington is 26th in population growth, enjoying a healthy but unimpressive 8 percent rate of growth since 2000. By contrast, Atlanta has grown at a 20 percent clip since the last Census, even though its median income is $15,000 below Greater Washington’s. Dallas and Houston both grew about 15 percent since 2000, adding more people each during that time than the Washington area, even though their median incomes are over $20,000 lower than the capital region. Daytona Beach, Florida and Brownsville, Texas both grew faster than the Washington area, even though Daytona Beach’s median income is barely half Washington’s, and Brownsville’s is barely one-third. What sense does this make?

Recent research suggests that while population may wish to move toward high income areas, it is inhibited from doing so by constraints on housing supply. Over the past two decades, places like Houston and Atlanta have issued far more new housing permits–in absolute terms, never mind per capita–than places like New York, Washington, and Chicago. The metro areas of Boston and San Francisco aren’t even close. As a result, housing in the Sunbelt is incredibly cheap and plentiful, while housing in the Big Old Cities is limited and expensive.

An interesting look at the dynamics of these differences can be had in our local area. During the first part of this decade, growth in the region was dominated by the Northern Virginia suburbs, and especially Fairfax County. From 2004 to 2005, Fairfax’s population increase by about 6,000 people. Interestingly, during 2004, Fairfax also added about 6,000 new housing units. From 2005-2006, however, population growth was only around 500 while about 4,000 new housing permits were issued. In 2006, Fairfax only authorized 2,000 new housing units.

The dynamics of the housing bubble obscure the underlying facts somewhat, but it seems clear that demand for Fairfax has not fallen as significantly as the number of new housing permits issued. Home prices in the county have been fairly stable. Rather, residents of Fairfax County, unhappy with the pressures of new growth, have expressed a desire for slower growth, and local politicians have responded. But Fairfax is one of the wealthiest counties in the nation. By shutting off the incoming population stream, Fairfaxians are essentially pulling up the ladder after themselves, denying others access to life in Fairfax.

Because population growth in metro areas is constrained by uneven supply limits, labor markets can’t clear, and median income differences show real persistence. What’s more, it seems probable that the truncation in new growth takes place at the margins of a population, where families can barely afford a home in an expensive metro area; households earning well above the median income aren’t at all deterred. Lower and middle income families, in other words, are the first to move from Washington to Atlanta. If income is correlated with skills, however, this suggests an increasing concentration of skill, or human capital, in a shrinking pool of cities–a concentration other metro areas are unable to match. As such, the long run distribution of income across cities should tend toward dispersion rather than conversion. Only the most talented can afford to live in a handful of urban bright spots; those bright spots thereby grow brighter, increasing the premium on housing there and deterring still more marginal residents while attracting the cream from other places (for whom the return to skill is greatest in the bright spots).

If that is the case, I see two policy paths available to address the long run disparity. First, one could figure out what it is about those bright spots that makes them so bright and try to duplicate it elsewhere. This is not an easy proposition. The Raleigh area’s Research Triangle Park was one of the few successful attempts to capture some of the magic of Silicon Valley, and after decades of tech growth there, the city continues to have median incomes well below the real thing. The other potential solution is to increase the population capacity of bright spots, by boosting density and transportation connections.

This ties in to the point I made last week regarding disparities in highway funding. Highway subsidies facilitate the growth of booming exurbs in the Sunbelt, while better transit links in large cities allow those cities to develop more densely and effectively. Subsidizing highways has meant favoring, through federal policy, the abnormal growth of lower income cities. This has, in all likelihood, exacerbated the growth of inequality in the United States.