Wednesday, June 24, 2009

Give me land, lots of land

Wendell Cox reports that recent census data show continued population exurbanization -- for the period 2000-2007. That's where the land is, where permitting is (mostly) simpler, and where most people want to be. It's an old story.

Our research (with colleagues Harry Richarsdon and Soojung Kim) considered employment decentralization for the years 1969-2004 (using the BEA's REIS data file). We examine county-level data aggregated into broad categories that include the relatively recent metro-micro area designations. We look at differences, local less U.S. growth, filtering national cycles. Figure A (below) shows job growth (five-year moving averages, dark line) trends for the "large" (over one-million pop.) metro areas; the other figures show the concurrent trends for "small" (less than one-million) metro areas, micropolitan areas adjacent to large and small metros as well as non-adjacent micropolitan counties and finally adjacent as well as non-adjacent non-core (not metro, not micro) counties. Very roughly, these represent the urban-rural hierarchy.

What do we see? First, there are cycles but no discernible trends. The rise of the Internet, for example, is not observable. Second, for private employment growth, the cycles become more pronounced as we move down the urban hierarchy. Third, the larger areas consistently do better in terms or proprietor employment growth. Fourth, the two types of micropolitan areas display very similar growth paths, but with the one exception that the ones adjacent to the large metros have stronger proprietor employment growth in the late 1990s, perhaps reflecting the beginnings of the boom in tech start-up firms. Fifth, the remaining three areas show similar growth patterns with the most remote ones (non-core nonadjacent) experiencing slightly sharper fluctuations.

The pulses are the reverse mirror image of the large metros -- and they extend ever more sharply down the urban hierarchy.