Matt Yglesias commented on my post last week on the amount of wages left over after paying for housing. He noted that the analysis:

Shows that among U.S. metro areas not only is income correlated with housing costs, but the association is so strong that wages minus housing costs is positively correlated with housing costs (I wish they'd looked at income rather than wages, though, since the existence of adequate labor demand is a relevant issue here I think).

My colleague Charlotta Mellander re-ran the analysis using the per median household income (via the American Community Survey) of metros in place of wages and salaries. Income is a broader measure of overall wealth, including capital gains, rents, and transfers in addition to wages and salaries, and this provides a better measure of the overall effective demand for housing, though wage and salaries remain a better measure of underlying regional productivity and value-creation.

The variation across metros is again huge. The metro at the top of the list has more than $5,000 ($5,364) per month or more than $64,000 per year left over after paying for housing per capita. Each of the top ten metros have more than $52,000 left over after paying for housing – more than the U.S. median household income of $51,369. At the opposite end of the spectrum, the 16 metros at the bottom of the list have less than $2,500 left over per month or $30,000 per year; the metros at the very bottom of the list have less than $2,000 ($1,961) left over per month.  Across the U.S., the average income per household left over after paying for housing is $3,313 a month or $39,765 a year. 

The overall pattern is quite similar to that of wages, as the scatter-graph above illustrates. As before, the metros above the line (hover over the dots to see their names) — have more income left over after housing is paid for. Those below the line have relatively little left over once housing costs are taken into account.

Mellander's correlation analysis also turns up similar patterns (as always, note that correlation does not mean causation). The amount of income left over after paying for housing is again associated with the share of adults that are college grads (.55 vs .53 for wages) as well as for the percent of the workforce in knowledge, professional and creative occupations (.51) though this is lower than for wages (.75). Metros where people have more income left over after paying for housing again have higher concentrations of high-tech industry (.56, the same as for wages) and a higher rate of innovation (.52 vs. .45 for wages). There remains a significant correlation to metro size (.41 vs. .43 for wages). And again, we find close correlations between income left over and more overall affluence (with a correlation of .65 to economic output per capita compared to .63 for wages).

While the overall pattern is similar to the one we described in our earlier post, the specific ordering of metros changes somewhat when income rather than wages are considered. (See the top 20 table below). Washington, D.C. now displaces San Jose for first place. Bridgeport-Stamford Norwalk is third, followed by Anchorage and then San Francisco. Boston falls slightly from 7th to 9th place, while Boulder moves down to 18th from 9th. Baltimore (13th) and Minneapolis (19th) enter the top 20. New York drops out of the top 20 (it was 8th on wages) to 24th, as does Philadelphia, which falls from 20th to 27th. Chicago is 36th, L.A. 76th. Rustbelt metros also fall further down the list with Pittsburgh, Detroit, Cleveland and Buffalo each having about $3,000 left over after housing, falling behind Las Vegas and Atlantic City.

Photo credit: Fred Prouser/Reuters