Is Roanoke Really An Appalachian Comeback Story?
Yes! But the size of that comeback could be debated.
Today I’m going to look at a different Appalachian city than I did last time, when I looked at Huntington-Ashland. Now, I say Appalachian, but Roanoke, VA isn’t actually part of the Appalachian Regional Commission. In fact, ARC’s definition of Appalachia excludes virtually the entire Shenandoah Valley area, while weirdly reaching almost all the way to the Mississippi River.
This happened, historically, because Appalachia was not defined by any geographic or cultural commonality at all. It was defined by poverty. Think about that. The legal definition of Appalachia was set based on which counties were poor. Plus, many states lobbied to get more counties added in order to become eligibl for more funding. The inclusion of Mississippi is particularly egregious, as is much of the included area in Alabama, and some of the fringe-counties in Kentucky and Ohio. Meanwhile, decisively Appalachian regions in Virginia and North Carolina were left out of the definition.
But alas, that’s the definition people use. I, however, will insist on calling Roanoke an Appalachian city.
And Colin Woodard’s Politico Magazine article describing Roanoke’s recent history and revitalization is incredibly serendipitous, as he describes the process of an industrial mill-and-logistics town transforming itself into a knowledge hub. That is exactly what I argued Huntington-Ashland, and Eastern Kentucky on the whole, must do to survive.
I’m not going to take a long time summarizing Woodard’s excellent article. It is a tour de force in explaining how an industrial city can save itself. Read it. And then remind yourself that Pikeville, Kentucky has been expanding their new medical school recently, and that their population has been growing while surrounding areas shrink. Hmmmm….
Anyways, I’m not writing just for boosterism of Roanoke or to belabor my point that Eastern Kentucky needs knowledge hubs. Rather… I have to be a bit of a debby downer.
The opening to the article claims that the Census measured just 15 people in downtown Roanoke in 2000, but that over 2,000 live there now. This claim is wrong.
The city manager of Roanoke passed along me a New York Times Article that seems to be one of the original sources of this claim, although it says that there were fewer than ten people, and then (in 2012) over 1,200. The over 1,200 part is correct; the under 10 bit remains dubious. But beyond just being wrong, these figures create false hope.There are downtowns that have shrunk to finger-countable numbers. They’re not going to get thousands of new residents any time soon, no matter how well-managed an economic development program they get.
In fairness, this error might not be Colin’s fault. He said on Twitter:
And he bounced me over to the city manager, who seems to be a pretty on-the-ball guy, and who shared the article above with me, and passed my email on to the city planners. The planners then kindly (and promptly!) got back to me and (1) acknowledged that Census gives a very high population estimate and (2) suggested that locals viewed this figure with incredulity. They told me that “under 100” downtown residents was a pretty universal belief.
Update: Planning officials got back to me with more information, and noted Tract 11 includes the Roanoke jail, which Census counts. I looked at the Census data again, now knowing what to look for. In 2000, Census counted just 170 housing units in Tract 11 of which 134 were occupied, vs. 517 in 2010 of which 426 were occupied. Some of those 2010 units were in the added area from the Tract definition change, meaning that Core Tract 11 probably actually added something like 200 occupied units from 2000 to 2010. Estimates of the household population are 202 in 2000 and 620 in 2010; again, probably at least a few of those additions are in the new Tract 11 area, so growth is slightly overstated. The most recent ACS puts group quarters (mostly the jail) at 521 residents, which means the household population is probably about <1,000 people in Tract 11 as of 2014, but may be as high as 1,300–1,500 by 2016.
These revisions mean that all my Tract 11 data is skewed upwards, but they still imply that (1) population was never below 100, (2) population is not over 2,000, (3) growth rates have been impressive, (4) growth rates have not been as impressive as advertised. Many thanks to the diligent folks at Roanoke’s city government who helped me figure this out! I am not revising the rest of the article because I think it’s important to leave a record of times when we make mistakes.
So we can track down this 10/15/under 100 residents claim: it is based on non-quantifiable local knowledge. This is valuable knowledge! However, I will say that, no offense to the very helpful folks working for Roanoke, I trust the Census way more. Census makes an effort to include many people who may be invisible to local governments: illegal residents, people residing in non-residential spaces, people who regularly sleep in a place but technically maintain legal residence elsewhere, people making a very long-term hotel stay, etc. Many of these “residents” can be invisible or undercounted when approached from an administrative or real estate-development perspective, because they are unlikely to buy, often lack property rights, or have incentives to limit their visibility.
Update: The above paragraph looks way smarter now than when I first wrote it, given that fully 1/2 to 1/3 of the tract’s population was exactly one such oft-undercounted population. I was pretty much just guessing at the time.
In other words: it is entirely possible that city leaders are honestly and fairly presenting the information that passes in front of their eyes when they say the downtown population was under 100 people, and yet they may still be off by orders of magnitude. As a rule, I take Census data, and I am extremely skeptical of the idea that Census overcounted 10–100 people as 874 people…but it is worth noting that Census data may not represent the population relevant for all city services or real estate development endeavors.
The Actual Census Record for Downtown Roanoke
Step back a few months and you’ll remember a similar post to this one. I was covering the comeback of Des Moines, Based on a Politico Magazine article by Colin Woodard, in which he stated that the population of downtown Des Moines rose from 1,000 people to over 10,000 people. His source in that case was city officials and real estate developers. Same source in Roanoke. These estimates in Des Moines were, as I showed at the time, incredibly off-base. They are wrong for Roanoke too.
Roanoke’s downtown population has shown a moderate, but very encouraging, amount of population growth.
Now, to be clear, Roanoke is a booming town, and it is experiencing real recovery, and we should hold it up as an example of “What Works”! The main thrust of Woodard’s article is completely correct! But the vision being offered here of a downtown population boom in small cities… sorry, density-boosters. It’s not true.
We can show this easily. Here’s a map of Census tracts in 2000 and 2010:
And here’s an aerial photo of that same area:
Notice the higher density out towards the right. That’s Census Tract 11. Let’s zoom in on it.
Yep, looks pretty much like a downtown area. Denser, taller buildings, more business, apartments, etc. Great. We have found downtown. It is Census Tract 11, maybe Tracts 10–12.
Now we can look up Census-measured population for Tracts 10, 11, and 12:
In 2000, Census Tract 11, which is synonymous with downtown, had 874 residents. By 2010, that rose to 1,204 residents. By the 2010–2014 ACS, it was somewhere between 1,159 and 1,529, midpoint at 1,344. So that seems like steady-ish growth from 2000 to present.
But 874 is much bigger than 15, and 1,529, the maximum population, is much less than 2000.
Over 15 years, the average growth rate from 15 to 2,000 people would be 39% per annum. The actual growth rate was probably less than 3.8% that much. In other words, real estate developers overstated the growth rate of the downtown population by a factor of 10. That’s about what happened in Des Moines too. I was unable to find good real estate data to give me a sense of how home prices have changed along that tract, sadly. However, Roanoke’s city planners tell me there has been further development since 2014, and they expect the population is 1,550–1,700 in the downtown area, which seems totally plausible to me. If it’s 1700 in 2016, then that’s a 4.2% annual growth rate since 2000, a really good growth rate for a downtown! To be clear, even these much more sedate estimates I’m suggesting still show that Roanoke is experiencing top-shelf growth rates.
Update: Adjusting for the group-quarters population, the annualized growth rate, assuming the 1,700 total population for 2016, would be about 11.8%, so much higher than the numbers I listed, but, again, still vastly lower than the figures given in the original article.
But, to the point, correcting these well-intentioned but nonetheless erroneous factual claims is the reason why I run this blog: because the numbers people believe about local-area population are deeply colored by how much skin they have in the game. You think Presidents fudge the data? Wait until you see real estate developers.
Two final notes: First, the definition of Tract 11 changed from 2000 to 2010. It grew! If you look at the maps, you’ll see that part of Tract 10 was moved into Tract 11. Most of this is non-residential use or parking. But there is one new apartment building with new residents. In other words, some of the observed increase in Tract 11 is an illusion due to changing tract definitions! And indeed, when we add Tracts 10 and 11 together, we find that population fell from 2000 to 2010 from 3,659 to 3,396 people. For the 2010–2014 ACS, it is somewhere between 3,110 and 4,080, with a midpoint showing some growth to 3,595. Second, I also looked at 1990: Tracts 10 and 11 had the same definitions in 1990 as in 2000. Tract 11 had 1,024 people in 1990, and the 2 tracts together had 3,829 people. Ergo, this “only 15 residents” can’t apply to 1990 either. Here’s a population graph for Tract 11, Roanoke’s Downtown:
Roanoke Really Is Growing
We Should Celebrate the Real, Incremental Victories
Virginia’s “independent city” system, whereby cities of a certain size become county-level entities, is quirky, so makes the data a bit different to present than usual. Below are my city and county population estimates for Roanoke City & County:
The big shift in the 1970s is due to a city land annexation, as described in Woodard’s article.
Note that, before that annexation, population was declining in the city. And after it, the decline continued until the early 2000s. Then, growth returned!
This kind of comeback is not common. Now, in fairness, continued growth for the surrounding county suggests that the wider area was not experiencing serious demographic weakness. For example, here’s metro-area population:
I’ve included a comparison to Huntington-Ashland because I really do think the Tristate Area has a lot to learn from Roanoke’s success. Despite a sluggish 1980s and 1990s due to the loss of key employers like Dominion and Norfolk-Western, the Roanoke metro area managed to experience some continued growth, and today is approaching Huntington’s population.
This is nothing to sneeze at! The principal city of Roanoke has added over 6,000 residents since a low in 2004. The surrounding county has added another 7,500, and the rest of the metro area has added another 7,000. Now, granted, annual growth rates have still generally been below the state average. But from 1980–2000, Roanoke metro area’s growth rates were 1/3 the state average. Today, they are approaching 2/3, and in some years have beaten state growth rates. Many post-industrial Appalachian towns would kill for that kind of relative strength.
Conclusion
I don’t want to downplay what Roanoke has accomplished, rather I just hope to encourage some realism in what other cities can expect. Roanoke is an honest-to-goodness comeback story, showing how a focus on local asset-building and knowledge-generation can fuel regional economic recovery. After losing over 12,000 residents from its peak, Roanoke has now recovered more than half the losses, and revitalized its local economy as well. Whatever Roanoke is doing, it is working. Other cities can and should seek to emulate successful components.
However, Roanoke’s downtown population almost certainly did not rise from 15 to 2,000 people. More plausibly, it has risen from about 800–900 to about 1300–1700. That is a remarkable increase that would make many cities giddy, especially when paired with growth in the whole city, surrounding county, and metro area generally. There is no need to use dubious data to exaggerate Roanoke’s population recovery to make clear the key point that Roanoke has turned itself around.
Check out my Podcast about the history of American migration.
If you like this post and want to see more research like it, I’d love for you to share it on Twitter or Facebook. Or, just as valuable for me, you can click the recommend button at the bottom of the page. Thanks!
Follow me on Twitter to keep up with what I’m writing and reading. Follow my Medium Collection at In a State of Migration if you want updates when I write new posts. And if you’re writing about migration too, feel free to submit a post to the collection!
I’m a native of Wilmore, Kentucky, a graduate Transylvania University, and also the George Washington University’s Elliott School. My real job is as an economist at USDA’s Foreign Agricultural Service, where I analyze and forecast cotton market conditions. I’m married to a kickass Kentucky woman named Ruth.
My posts are not endorsed by and do not in any way represent the opinions of the United States government or any branch, department, agency, or division of it. My writing represents exclusively my own opinions. I did not receive any financial support or remuneration from any party for this research. More’s the pity.
