The Economic Innovation Group recently published the Distressed Communities Index. As the most distressed large city in the United States, Cleveland dove into a Rust Belt shame spiral. The EIG managed to re-light the river fire. The Mistake on the Lake burns again. That's too bad, because the DCI is deeply flawed.
While the DCI does not measure population change, such data can help explain why places where manufacturing once blossomed seem to rule the bottom of every list. Such cities tend to be small in terms of geographic area. Wealth suburbanized relatively early in steel powerhouse Pittsburgh, effectively shutting the door on the expansion of the municipal boundaries. Worse, the more successful the manufacturing economy, the greater the sprawl of prosperity.
The DCI's focus on place obscures the experience of people. Cincinnati could have two distressed zip codes with entirely different demographic compositions and migration dynamics.
Ben Schulman and Xiaoran Li attempted to overcome this geographic unit of analysis problem, normalizing city size. For the 2010 Census, the City of Pittsburgh's population more than triples by considering communities that would be included in the average-sized American city. Boston gains over one million residents. Annex-happy Houston loses over 400,000. By comparing cities, the EIG has an apple-to-oranges problem. Tellingly, the DCI does an outstanding job of identifying Rust Belt geographies and reinforcing the pejorative stereotype.
The high variance in city area lends itself to some quirky results when surveying the sortable data. The EIG posts the numbers for overall distress score, distress core ranking for the top 100 cities in terms of population, percent of the population living in zip codes that qualify as distressed, and an inequality rank. Colorado Springs, Pittsburgh, and Nashville all have 15 percent of their population residing in a distressed zip code. The lower the better.
Colorado Springs has a distressed score of 41.4. Nashville is a little worse at 48.7. At 68.3, Pittsburgh appears, well, Rust Belt. What explains this discrepancy? Without the full data set available, I can't be sure. Perhaps Pittsburgh's "distressed" are worse off than those in Colorado Springs or Nashville. Furthermore, many distressed Pittsburghers may live in non-distressed zip codes.
And therein lies the real problem. The DCI's focus on place obscures the experience of people. Cincinnati could have two distressed zip codes with entirely different demographic compositions and migration dynamics. In the Appalachian white ghetto, few people move in or out of this rather inert community. Thus, the distress score accurately reflects the socio-economic status of both place and people. In the Hispanic neighborhood, immigrants from rural Mexico move in and thrive. But once out of distress, they move outside the city and into a suburb where schools are better. Over time, this zip code looks as distressed as the Appalachian zip code. The experience of residents couldn't be more different.
A Rust Belt-like community might be isolate or escalator. In an isolate neighborhood, poverty is inter-generational. People are stuck in a distressed place. In an escalator neighborhood, migrants move in, up, and then out. In this case, the distressed zip code is a launchpad to upward mobility. The EIG does not distinguish between the two and policies emanating from such data would be misplaced.