Exerpt from map published in CityLab. Purple = net in-migration. Gray = net out-migration.
Exerpt from map published in CityLab. Purple = net in-migration. Gray = net out-migration.

by James A. Bacon

Richard Florida’s latest demographic research project presents an interesting twist on the old “domestic migration” data we’ve all seen (well, we’ve all seen it if we’re regular readers of Bacon’s Rebellion!) The U.S. Census tracks “domestic migration” — the movement of American citizens from one locale to another — and commentators have made much of the fact that some metros are consistent population losers while others are consistent gainers.

While the migration data is useful, there is much it doesn’t tell you. It’s one thing if your region is seeing an influx of college grads and Ph.Ds and quite another if you’re getting flooded with high school drop-outs. Florida broke down the migration data by education level to determine which metros were enriching their human capital, which metros were diluting their human capital and which were simply exporting their human capital. You can see his presentation at CityLab.

In some ways, the data presents the same picture. Big Northeastern and Midwestern cities like New York, Chicago, Philadelphia and Detroit lost population across all education categories between 2011 and 2012. But other cities showed more complex patterns. The Washington and San Francisco metros saw a net in-migration of people with college degrees and advanced degrees and a net out-migration of the less-than-college educated. Not a bad trade — you gain the people who contribute the most to the economy and tax base and you shed people the most likely to wind up on food stamps and unemployment insurance!

Here are the details for Virginia MSAs. Take the numbers for Charlottesville, Blacksburg and Harrisonburg with a grain of salt. Their massive outflow of college grads and Ph.D.s undoubtedly reflects the outsize demographic impact of local universities — hundreds of students come in with high school degrees and leave with sheepskins.

Richmond saw the greatest net in-migration of any Virginia metro during this period, although the newcomers were a mixed bag educationally. The region saw a bigger influx of people with less-than-high-school education levels than college and post-graduate combined. Roanoke experienced what can only be called a brain drain — no, make that a brain hemorrhage — losing large numbers of educated residents while receiving equally large numbers of ill-educated citizens.

This data is fundamental to economic development. Political and civic leaders need to know not only how many people are coming and leaving but what their net education level is — whether the region is building human capital or seeing it shrink. Each region needs to track this data over time. And, if it wants to have any kind of future in the entrepreneurial knowledge economy, it needs to attract citizens with the highest level of education who are likely to contribute the most to the economy.

The prognosis from this one year’s worth of data is not good for downstate Virginia. Throwing out the figures for Bristol-Kingsport (which is mostly Tennessee) and the three college towns, I conflated results for Hampton Road, Richmond, Roanoke, Lynchburg, Danville and Winchester. Here is the composite tally:


That’s just one year but it looks like a major brain drain to me. We should be terrified. But we’re not. We should feel a sense of urgency. But we don’t.

Share this article


(comments below)


(comments below)


7 responses to “Sure Looks Like a Brain Drain”

  1. larryg Avatar

    My initial reaction is that the data is all over the map (literally) and I wonder how valid it is especially for college towns.

    and in those cases – my opinion is that data like this needs a second, independent validation… and the first thing I would do is to randomly select one or two places with questionable data and look at the last 5 years …

    The way that the Census folks go about getting Community Survey Data, I think, can lead to errors of interpretation.

    this data is just so skewed looking… without a good rhyme or reason why.

  2. Larry, I do agree with you — I question some of the numbers, too. Roanoke lost 816 people with post-graduate degrees in one year? At that rate, the metro will totally empty out of doctors, lawyers, MBAs and college professors in just a few years. The number seems extreme.

    Could it be that the Census’ American Community Survey takes samples and then extrapolates from those samples? There is probably a lot more variability in the numbers for smaller MSAs like Roanoke.

    Still, even accounting for some margin of error at the individual MSA level, the overall numbers are scary.

    1. larryg Avatar

      looks like they do phone calls and internet and not sure how they determine who left.

      do they call the same households form prior survey’s? If someone is gone – do you assume they have moved completely out of the MSA?

      How do you know if they just didn’t move to a new place in the MSA?

      the data just looks funky…. I spent a little time on the American Survey website…


      I think Mr. Florida might be well advised to not take the data at face value and do a little due diligence…

      it says things like:

      ” Multiyear estimates cannot be used to say what is going on in any particular year in the period, only what the average value is over the full period.”

      that’s sufficient to give me some pause.

  3. DJRippert Avatar

    The Dems had “Cash for Clunkers”

    Maybe the Republicans should consider “Makers for Takers”

    Ha ha!

    I think you have to consider retirements in the data. An MD who retires in Fairfax and moves to Florida doesn’t add to the employed – educated pool in Florida.

  4. Ghost of Ted Dalton Avatar
    Ghost of Ted Dalton

    Good post. Although the more you look at the national picture, I’m not sure that Virginia is any different.

    If we want to “tell the hard truths”, then it’s pretty clear a lot of conservative ideas have to vanish. Not all, but a lot.

    For instance, if the future really is a megacluster of talent in dense areas, then it’s simply time to quit subsidizing rural areas in anything but agriculture. It’s also time to really double down on education. We can argue about school choice, etc., but if this is the future, then American society has got to do its very best to educate children. And yes, whether it’s school choice or traditional public schools or charters, that means tax increases and putting more money into developing intellectual capital at all levels and ages of society. Abolishing the bachelor’s degree is probably a good idea. BUT….with that comes the idea that everyone is going to constantly take courses at universities for their entire adult life. Which is a good thing (building new skills on a consistent basis), but again, we need to pay for it.

    I agree with your smart growth posts. There are a lot of good ideas out there on both the left and right for dense, urban living spaces. I love your recent posts on public spaces as well as new ways of viewing transportation. If this is the future, then we’ve got to reorient policy towards that model that either incents this model or discourages the rural/exurban model as it currently exists.

    But I will guarantee you this. The right will fight like savage dogs if you try any of this. Look across the country. The only way Republicans win in almost any state is through taking 70+% of the rural vote. GOPers know that the model that you and larryg and I and a lot of other posters think is the appropriate 21st century model will lead to the extinction of the Republican Party as we know it. Now, I think the GOP could reinvent itself. But they don’t want to. They understand that this model probably demolishes so much of what the current party stands for…..anti-gun control, anti-tax, anti-public education, anti-gay marriage, anti-abortion…..these are all products of rural, exurban thought for the most part (except the abortion issue). Whether you’re Jim Bacon (who favors more private sector solutions, but is willing to have some state programs) or larryg (a little to the left of Bacon), if you believe that the 21st century model is the megacluster, then you ultimately believe in some degree of communalism and community. The idea that public spaces should work to connect people to start conversations. The idea that dense development can lead to greater cooperation. Etc., etc.

    Look at the kookiness over the UDAs a few years ago (Agenda 21, etc.). If you really believe in the model you advocate, then you need to vote a straight D ballot the rest of your life. I’ve tried for years to talk to Republicans about a lot of the things you advocate. It terrifies about 90% of them. Unfortunately, the party has been taken over by a group of pseudo-libertarians who look askance at anything that even hints of communalism or cooperation.

    There was a time that I voted about 90% GOP. But as the years have passed, the GOP has become the most intellectually hollow institution in our society. There are 3 planks: No abortion, no gun control, no new taxes. That’s it. There’s absolutely zero vision for 21st century America.

    1. larryg Avatar

      re: ” There are 3 planks: No abortion, no gun control, no new taxes. That’s it. There’s absolutely zero vision for 21st century America.”

      anti-immigrant, anti-Hispanic, anti-gay-lesbian, anti-black, anti-muslim, anti-public-school, anti-science, pro-creationism, pro-corporate money in elections,

      Howard Baker died this week – the ghost of the GOP.

      what the GOP wants today is a Tea Party version of Ronald Reagan!

  5. The Lynchburg numbers aren’t so alarming, as it has a relatively high college population- around 16,000 in four year schools. The net out-migration figures seem to make sense… and the dropouts can continue to head off to greener pastures like Richmond, Roanoke, and Danville!

Leave a Reply