From Papers in Regional Science, a new paper shows the connection between lost labor productivity and sprawl. From the abstract:

This paper draws on urban agglomeration theories to empirically investigate the relationship between the economic performance of US metropolitan areas and their respective amounts of sprawl. To measure urban sprawl, we construct a distinctive measure that captures the distribution of population density and land-use within metropolitan areas. Using both ordinary least squares (OLS) and instrumental variables (IVs) approaches, we find that higher levels of urban sprawl are negatively associated with average labour productivity. This pattern holds even within given industries or within given occupational classifications.

link: Urban sprawl and productivity: Evidence from US metropolitan areas – Fallah – 2010 – Papers in Regional Science – Wiley Online Library

Here’s their sprawl measure:

It’s a relative measure: L% is the percentage of the regional population living in block groups with population densities lower than the US median, H % is the percentage of the metropolitan area living in block groups above the US median. The number is going to range between 0 and 1, with higher numbers meaning greater sprawl.

So the least sprawled regions wind up being the big metro areas, with some surprises:

Miami-Fort Lauderdale-Pompano-Beach, FL — 0.3405
Stockton (!!!) — 0.3394
CA Chicago-Naperville-Joliet, IL-IN-WI — 0.3329
El Paso, TX — 0.3315
San Diego-Carlsbad-San Marcos, CA –0.3176
Honolulu, HI — 0.3170
New Orleans-Metairie-Kenner, LA — 0.3104
Laredo, TX — 0.2620
New York-Northern New Jersey-Long-Island, NY-NJ-PA — 0.2479
San Francisco-Oakland-Fremont, CA — 0.2313
San Jose-Sunnyvale-Santa Clara, CA — 0.2058
Los Angeles-Long Beach-Santa Ana, CA–0.1630

According to their measure, Los Angeles is the least sprawled metro region in the country, beating out New York and San Francisco.

This creates problems. Because Los Angeles IS what people worry about when they worry about sprawl.

The most sprawled:

Barnstable Town, MA–0.9497
Sebastian-Vero Beach, FL –0.9232
Punta Gorda, FL –0.9041
Panama City-Lynn Haven, FL –0.8868
Spartanburg, SC –0.8767
Pensacola-Ferry Pass-Brent, FL –0.8722
Burlington, NC–0.8657
Fayetteville, NC — 0.8560
Lakeland, FL — 0.8538
Chattanooga, TN-GA — 0.8478
Hickory-Lenoir-Morganton, NC — 0.8404
Fort Walton Beach-Crestview-Destin, FL — 0.8402

They find that the connection to labor productivity comes from the share of population who has a college degree. That is, the self-selection of college graduates in the denser regions. It’s hard telling what these findings really show; their data year is 1990, and the sprawl measure is blunt.

They play with different sprawl measures, and their findings are pretty robust.

So if they are right, what does this mean? I’m a wee bit worried about this finding, actually. It suggests that the connection between sprawl and labor productivity derives from residential self-selection into particular metropolitan regions–the big ones–rather than from anything related to physical aspects of urban form. They aren’t, for example, capturing any real inherent disadvantages from commutes on labor productivity, or loss of connectivity in the exchange of ideas. Instead, it’s a matter of correlation rather than causation: high-productivity laborers like urban amenities–Richard Florida’s arguments.

If that’s the case, the labor productivity nexus can change a lot if those taste preferences change.

Go read the full article:

Fallah, Belal, Mark Partridge, and M. Rose Olfert. “Urban Sprawl and Productivity:Evidence From US Metropolitan Areas.” Papers in Regional Science (2010).

Add to: Facebook | Digg | Del.icio.us | Stumbleupon | Reddit | Blinklist | Twitter | Technorati | Yahoo Buzz | Newsvine