Last weekend Angie Schmitt pointed me to an article by Douglas Trattner in Fresh Water Cleveland. The author suggests Rust Belt cities, left for dead, are suddenly booming again. Angie was suspicious of some of the claims and I offered to check it out. Let's start with the article...
Daily, it seems, another cultural sociologist is writing about the current trend of reverse migration -- young creatives fleeing the Coasts in droves in favor of "decaying" industrial cities like Cleveland, Pittsburgh and Detroit. These cities, you see, are appealing because of the decay. That and ironic pleasures like bowling, pierogies, and polka.

Of course, there is enough truth and fiction in that charming narrative to choke a thesis on contemporary demographics. The truth is, young people are moving back to cities like Cleveland, Detroit and Pittsburgh -- and at rates that outpace those of posh suburban zip codes. Offering the promise of a better (cheaper) quality of life -- and yes, the ironic pleasures of bowling, pierogies, and polka -- Rust Belt cities truly have become "chic."

The pivotal point in the narrative may have occurred on May 12, 2012. That's when Salon published the article "Rust Belt Chic: Declining Midwest cities make a comeback." The sub-hed was "Gritty Rust Belt cities, once left for dead, are on the rise -- thanks to young people priced out of cooler locales."
The quote above conflates two distinct ideas. First, that young people have found new love for Rust Belt cities because the "cooler locales" in coastal metros have gotten so expensive that young people can't afford them anymore. Second, that when these people live in rust belt cities, they opt for the urban core rather than the "posh" suburbs.

I think the second point has some merit, and it's been documented that even in central Cleveland neighborhoods that are losing population, the number of young people in those neighborhoods is growing (the key to remember is that these neighborhoods are still shrinking, they're just shifting appeal to a younger crowd). Nevertheless, the first point is open to interpretation, and whether you think it's true or not depends how you define "droves" of people.

To figure out the answer to this question requires more than just anecdotes about people who moved from Brooklyn to Cleveland or San Francisco to Detroit. I dug into ACS data from 2008-2010 to answer a few key questions for the three cities mentioned in the article above:
  1. How many moved from the Rust Belt metro to a coastal metro?
  2. How many moved from a coastal metro to the Rust Belt metro? 
  3. How do the two numbers compare?
I looked at three groups who moved: everyone, people aged 20-35 (since much of the literature on this topic refers specifically to "young people") and people with college degrees (to address the "brain drain" question). Coastal metros for this analysis are those top 50 metro areas that are located in a state that touches either the Atlantic or the Pacific Ocean. A full list of them can be viewed here.

Let's start with Cleveland. Between 2008 and 2010, more people left for the coasts each year on average than came, and the results are statistically significant (the thin bars on these charts represent the calculated 95% confidence interval). However, there's no statistically significant difference between the number of young people and the number of degree holders who are in-movers and out-movers. Cleveland had the fewest number of migrants in either direction of the three Rust Belt metros in question.

(click to enlarge)

Detroit is a different story. In all three categories there are more out-movers than in-movers. Consistent with most accounts of depopulation, more people are decamping from Detroit for the coasts than vice-versa.

(click to enlarge)

Pittsburgh is an interesting case because more people arrived from the coasts than left for them. However, when you filter it down by young people and degree holders, there's no significant difference. What's interesting about Pittsburgh is that what appears to be driving the first set of columns is that there are more kids (and presumably thus families) moving to the metro area than in Cleveland or Detroit.

(click to enlarge)

In the end, these graphics show something interesting; but I don't think they're consistent with the bold claims made in the Freshwater article. You could argue that the data is old, and that the shift didn't start until 2011. Maybe that's true, but we'll have to wait to find out.

You could argue that Cleveland and Pittsburgh are actually in great shape if the number of young people and degree holders that are leaving are being replaced at essentially a 1-to-1 ratio. I really have no dog in  this fight, I just want to numbers to back up the rhetoric. Anecdotal evidence only goes so far.

When thinking about in and out migration, I think it's important to be cautious of the availability heuristic. When someone leaves a metro area, they usually do it quietly. They pack up, leave, and then people in that area rarely hear from them again. But when someone moves from another metro area, you hear about it all the time, because there are plenty of opportunities to hear about it. I suspect that people remember many more cases of people moving to their city than they remember cases of people leaving.

I'll end by saying that I understand why people want to live in Rust Belt cities. For some, the low cost of living makes the quality of life unbeatable. That might not be true for me, or for others, but I wouldn't question why someone would want to move there.


    On June 14, 2012 richey said...

    Great work Rob. This is metro to metro right? If so, I am surprised that Cleveland is at a nearly one to one ratio of in- and out-migration of young and college educated.


    Yes, it's all metro-to-metro. Exact definitions are here

    On June 15, 2012 Lynn P said...

    Metro to metro isn't always the most useful, or the most telling, comparison. There are important trends that start at the city or neighborhood level, and that often can only be seen there. If you looked at the demographics of the Washington DC metro area over the last 40 or 50 years, you get a very different picture than what was actually occurring in DC itself. The metro area-particulalry Northern VA--had tremendous, steady growth. In DC, however, there was consistent loss of population from 1950 to 2000 but an increase of 5% or 6% from 2000 to 2010. People laughed in Tony Williams' face when he said in 1998 that he wanted to increase population in DC significantly by 2010. If you looked at the numbers for specific DC neighborhoods, however, you could see that it might happen. I'm not suggesting the same pattern applies to Cleveland or Detroit or wherever, just giving an example that there are lots of different layers to the picture and that you can't embrace any one of them as telling the whole story.


    Lynn, I think Richey Piiparinen has done a pretty good job of analyzing the trends you mention at the city and tract-level. I'd recommend his work as further reading.

    On June 19, 2012 Al said...

    Interesting stuff, and it pretty much matches my anecdotal observations that Cleveland's "brain drain" is more of a slow bleed than an exodus. It doesn't seem too far-fetched to imagine it reversing in the next twenty to thirty years. The question I had is this: how do the rates of inflow/outflow from 2008-11 compare to previous three-year periods? I wonder if we're seeing a slow decline of net outmigration for the 20-35 age cohort and/or degree holders.

    On June 19, 2012 Anonymous said...

    Good work, but neither this nor Piiparinen’s study gets at the question I think you’re really after: are people moving to Cleveland (specifically downtown) because of the so-called “Rust Belt Chic”? Or are people moving downtown because they grew up in the Cleveland suburbs and are now sick of it? Or because they moved away for college or for a job and are simply returning home after a few years someplace else.

    There’s a big difference between people moving downtown because downtown Cleveland is awesome and people moving downtown because Cleveland’s suburbia is brutally bland and depressing.