America Becoming More Rural
A new Census definition has changed classifications slightly. It doesn't go nearly far enough.
Josh Zumbrun, the WSJ numbers columnist (“And Just Like That, America Becomes More Rural“):
With so much attention on the U.S.’s urban-rural divide, you might soon hear that the rural population in 2020 was much larger than in 2010.
That isn’t because people moved en masse to the country during the pandemic. It’s because the U.S. Census Bureau is updating its definition of an urban area, from one with 2,500 people to one with 5,000. That reclassified 4.2 million people, living in 1,140 areas of the U.S., from urban to rural.
This has real-world consequences: Access to many federal and state programs is based on whether an area is defined as rural or urban.
But it also presents an opportunity to take stock of how rural has been defined, and whether the definition really matches what people think.
“People’s perception of the norm for urban vs. rural can be very different than the government’s definition,” said Claire McKay Bowen, a principal research associate at the Urban Institute, a think tank.
When people think of urban areas, the images that spring to mind are probably of skyscrapers, buses, heavy traffic, bright lights, hospital systems and crowds of pedestrians, or of urban sprawl: the miles of subdivided suburbs and strip malls.
They don’t think of a town such as Salmon, Idaho, where Dr. Bowen grew up, with a population of just over 3,000. It is the county seat of Lemhi County, which is the size of Connecticut but with a population of under 10,000. School sports events were typically a 2½-hour drive—each way—for visiting teams. It is a five-hour drive to Boise.
Yet for over a century, the Census Bureau has tabulated numerous small towns like this as urban areas. (Rural areas are simply those that aren’t urban.)
It first calculated the urban population in 1874, in a “Statistical Atlas of the United States.” It was producing maps based off the 1870 Census and wanted to portray any cities of more than 8,000 as dots, then shade the remainder of counties to convey the density of the rural population.
As a supplement for the map, the bureau calculated the population of all the dots and labeled it the “urban population.” In early years, the population threshold for the dots moved around, but by 1906, the Census Bureau had settled upon 2,500 as the cutoff between urban and rural. In other words, the most widely used definition of urban didn’t result from any dedicated analysis but just someone attempting to create some cool graphics and tables.
As the country grew, the Census Bureau focused on measuring suburbs, small municipalities on the outskirts of larger cities that it includes in a city’s urban area.
Even as some urban areas became enormous—19 million people in New York and its environs; 12 million in and around Los Angeles—the cutoff for urban areas remained 2,500.
With the new definition, the rural population was 66.6 million across U.S. states and territories, or 20% of the total in 2020, compared with 18.7% under the old definition. If the old definition had remained in place, and the 4.2 million Americans reclassified as rural had instead remained urban, then the rural share of the population would have shrunk slightly, from 19.3% in 2010 to 18.7% in 2020.
Whether an area is classified as urban or rural can affect funding for healthcare, broadband development, transportation and so on. Some government agencies follow the Census Bureau’s classification, such as the Rural Health Clinic program. Others develop their own criteria. The Office of Management and Budget classifies entire counties as either metropolitan or nonmetropolitan based on proximity to a major city.
Any numerical cutoff is by its nature arbitrary. Whether the number is 2500, 25,000, or 25 million it’s not as though the character of the place would radically change were one person to move in or out.
Even beyond the number itself, what’s the unit of analysis? I live in Fairfax Station, Virginia, a DC suburb of 12,000-odd folks. Because of local zoning requirements that each lot be a minimum of five acres, it certainly feels rural. But I’m a ten-minute drive from multiple towns that feel more urban—and 22 miles from downtown DC, which feels more like what I think of as “urban.”
Zumbrun’s take:
Places with population just over 5,000 (or with over 2,000 housing units; another way to qualify) are still clearly small-town, not big-city, America. If they are culturally, economically and geographically more similar to sparser rural than denser urban areas, perhaps the threshold is still too low.
Again, I don’t think it’s that cut-and-dried. To me, it’s not just about the size of the town but its layout and proximity to major metropolitan areas.
My wife lived in Kennett Square, Pennsylvania when we first started dating. It’s about the same driving distance from Philadelphia as our current house is from DC. But it felt much more rural. Yes, it’s smaller—just over 5000 in population and just over 1 square mile in size vs 12,000 and 9 square miles—but it’s surrounded by other places that feel like small towns as well. While it’s the same distance to a major city (actually, two, if you count Wilmington, Delaware, which is even closer), you’re in what feels like a rural area in between. Also, it’s literally farming country—half of the mushrooms sold in the US are from the area.
Zumbrun, though, seems to want to classify more people as “rural”:
The Census Bureau originally proposed a threshold of 10,000 people (or 4,000 housing units), which would have reclassified an additional 939 areas with 5.7 million people as rural. It ended up using the more modest change to minimize disruption to the many programs that depend on Census definitions.
The World Bank has promoted an international standard that defines cities as having more than 50,000 people. Using this rule would put a total of 95 million Americans outside of cities—approaching 30% of the population.
There’s no question most Americans live in and around the major cities; over half live in the 72 largest. But the official rural population count has long been held down by an arbitrary definition. The country is more rural and small-town than we think.
There’s no obvious way to code “feels like” into Census designations but 50,000 seems like a more reasonable threshold than 5000. It’s not much of a city if it has fewer people than that. And I’m naturally predisposed to using international definitions to make comparisons more uniform.
The real problem is that “urban” and “rural” rather clearly aren’t the only categories. Places with more farmland than people are rural. Places with skyscrapers and concrete are urban. Most of the land area is something in between those.
A recent discussion here made realize just how widely we in the commentariat differed in our opinions of rural vs. urban. For me the changeover is based on expectations of government. Do you get your water from a public main and send your wastewater down a sewer? That’s urban to me. Do you expect publicly funded streetlights? That all roads will eventually be plowed by the government when it snows? That you can call the police and expect them to do something if your neighbor is having a loud, wild party at 2am on a Tuesday? That if that same neighbor erects a foul smelling pig sty in their backyard that you can rely on zoning to get it shut down?
I have the biggest disconnect with people who don’t feel that suburbs are urban. In some ways I think of suburbs as being more urban than the cities they surround. I guess because I can’t think of anything less rural than a Karen, and they are predominantly creatures of the suburbs.
@MarkedMan: I think of “rural” as synonymous with “farming” or “ranching” but that’s obviously not the Census viewpoint. We’re on a private road funded by the HOA and each household is on its own well and septic system here. My folks were on septic but had county water. Most everywhere else I’ve lived we had city/county water and sewer.
On some level, this seems like it should be based on density rather than raw numbers.
For example, the township I live in has roughly 20,000 people in it, spread over 22 square miles for a density of around 850 people / sq. mi
Meanwhile, the neighboring town has around 4,000 people in less than half a square mile for a density of around 6,200 people / sq. mi
It seems weird to suggest our township is urban but the town is rural
It’s also kind of fuzzy what is meant by “area”? Are the township and town two separate areas? Are we part of their area despite being a separate municipality? Are we all actually part of Philadelphia?
@James Joyner: You sound like you are in a rural/urban transition area. Septic and well sounds rural, HOA definitely sounds urban. FWIW, I’ve owned two houses that were septic and well, and one that was septic but water from a main. Other than a brief stint in a very, very rural village in the Peace Corps, the rest of my time has been definitely urban.
I guess propane vs. municipal gas is another dividing line.
@MarkedMan: I gather, but have not seriously researched, that this area was zoned at 5-plus acre lots precisely because that was the threshold at which the county was not required to provide those services. We have our own propane tank as well. Then again, I had one of those (albeit much smaller, since I only needed to power a gas fireplace) at my house in Troy; the neighborhood was simply all-electric.
@Stormy Dragon: I agree that density can be problematic, but maybe for a different reason. A lot of outlying suburbs have learned the lessons of the originals and now zone for exclusively single family homes on multi acre lots. Yes, they have wells, septic and propane, but they also expect a lot of services and are very heavily zoned. Don’t even think about running a commercial business unless you put it in one of the carefully chosen corridors. And no heavy manufacturing, ever. There is a sizable local police considering the number of residents and they spend a great deal of time looking for people “who don’t belong there”. I’ve owned homes in places that were close to that level, and believe me, if you erected so much as a tool shed without getting a permit some busy body would rat you out. That, to me, is the antithesis of rural.
@Stormy Dragon:
The census bureau also uses a method of measuring population density to determine rural or urban. Basically it looks at how many people live a certain distance from a center point.
@James Joyner: Fairfax County is so weird sometimes. I am probably less than five miles from you as the crow flies but we live in a townhouse development with probably 20 families on one cul-de-sac and needless to say we have county water and sewer. Even the nice new McMansions up the road are basically on top of each other. But 10 minutes drive and it’s rolling hills and horse country.
I like it, actually, but it’s so different from the Detroit area I grew up in.
The change meant rural population increased from 19.3% in 2010 to 20.0 in 2020. So I’m still delighted the GOPs have decided to be a rural party.
Karl Popper observed that generally definitions are constructed left to right. What are the characteristics of what we call “puppies”, the essence of puppyhood? Science tends to define terms right to left. Juvenile canines form a class useful for discussion, let’s call them “puppies”. It’s a question of what the Census Bureau is trying to capture by calling some areas rural and others urban? The article makes it sound like they’re not capturing anything well defined, it just grew.
I wrote my first paragraph as an example of the bad reasoning that flows from loose definitions. Being Republican, or MAGA Republican, may correlate with urban/rural, but doesn’t define it. In politics, rural is a poor stand-in for lower education and economic status. Economically it would make more sense to talk about agrarian/resource extraction/tourism v manufacturing/commerce/technology. Culturally, rural is a rough proxy for low education, low income, scarce cultural resources. For many purposes inner city ~= rural.
We are an 80/20 country (CONUS) with 80% of the people living east of a line from Winnipeg to San Antonio. But it is even more so, if you lop off California and continue that line through OR and WA, then you have just 30 million between the two lines. Just a bit more than the NYC metro area. Yet there are significant urban areas out there. It also really shows why the Electoral College protects the interests of this “minority” from the whims of those who can’t see past the Hudson or the Shenandoah Valley.
@JKB: The Electoral College wasn’t intended for that and doesn’t really even do that so stop trying to justify it that way.
It seems to me that the division between urban and rural should be based on how far one has to travel to get to a location where all services are available. If you live in a location where the emergency room, for example, is less than an hour away, you live in an urban area.
@JKB:
You cannot just chop off the most populous states to make an argument. It’s dishonest and dumb.
If you want to point out that vast, low-resource areas are thinly populated, say that. If you want a government the prefers vast, low-resource, sparsely populated areas, say that.
I prefer the argument that Americans go feral if they aren’t by a large body of water (ocean or great lake), or if they live in a former slave state.
(I would consider the slave state part to be dishonest if you couldn’t see the confederacy in modern maps of things like teen pregnancy per capita, syphilis per capita, child poverty, books read, etc. It’s clearly a major influence continuing 150 years after slavery, and seems to be more influential even than large bodies of water, so two driving forces in America)
I will care about the definition when it enables a significant policy shift for rebuilding the small cities that are the economic hubs of rural America.
If kids in Kansas want a future they need to get out of Kansas and move to one of the great metropolitan areas in our country, and that just hollows out the state. And we don’t want all those Kansas kids messing up our nice metro areas with their weird cow smells.
And even within a state, Washington has a lot of smaller cities hurting in rural areas that are basically dead zones.
This might be the first of 23 steps that lead to the revitalization of Centralia, WA, but it’s far more likely to be a random bit of movement around the edges of policies that amount to nothing.
@Just nutha ignint cracker:
The definition 30 years ago was more than 25 miles from any city of at least 25,000 people which is more in line with the way you’re thinking.
As someone who has spent the majority of his adult life working in big cities, NYC, LA, Atlanta, London, Paris, to me it’s simple to divide rural and urban. But it’s probably not the working definition. When I lived and worked in Miami, the locals would get pissed when I’d call Miami “a quaint fishing village”, because that is what it seemed to me, compared to LA and NYC and London. You forget how big places like LA and NYC are until you go to a place like Austin, or Memphis, even Atlanta, or where, 30 mins away from the city center, you see properties with 5-10 acres for less than 300K,
For example. Atlanta is an urban area. Winston, GA, which is only 30 mins (28 miles) away, is definitely rural. Marjorie Taylor Greene’s very rural district starts 20 mins from downtown Atlanta. Seems to me we should be looking at density more than just people.
@Gustopher:
Sedgewick County (Wichita) is over 500,000 now and growing steadily. Johnson County (KC suburbs) is over 600,000 and growing somewhat faster. Tulsa County, not far across the border in Oklahoma is 700,000 and growing steadily. A bit farther, Douglas/Sarpy Counties (Omaha) in Nebraska are approaching 800,000 and growing briskly.
Brain drain from the rural parts of those states into those successful metro areas has been going on for decades. The kids don’t seem to have much problem adapting.
@Gustopher:
Indeed. And even some on the fringes of major metros. For example, my little town, Kelso, WA (population ~12k [off stage–“SAAAAAA-lute!!”]) is, depending on which measures you bias towards, either the second or fourth poorest “city” in the state despite being across the river from and sharing a border with Longview, WA, and being less than 50 miles from Portland, OR. What really surprised me, though, was that Pullman, WA is the consensus poorest city despite being the home of WSU. I guess a university only creates so much pull as a driver of the economy.
@Michael Cain: True, but over the past 30 years, 25 miles has become a much shorter distance. When I first moved to SW Washington back in 1994, I had kids in my school that had never been to Portland. These days, their teachers live in Portland. Two or three years ago, when I-5 was closed with a landslide between Portland and Longview for 2 or 3 days, schools closed because too many teachers couldn’t make it to school until the road was opened for single lane traffic. Then the district moved to 2-hour late start for an additional day or two.
@Just nutha ignint cracker: My how the time flies, Covid happened 2 or 3 years ago. The road closure was two or 3 years before that.
@Just nutha ignint cracker:
Which is why the CB futzes with their definition pretty much every census, trying to keep it meaningful.
Fundamentally the problem is trying to divide a very diverse country into only two categories. Ultimately it obscures more than it reveals. The Census ought to get rid of that and add more categories.
@Gustopher: There’s also the push that BigAg has had on hollowing out the rural population. Farms get bigger, but the number of people required to carry out the work drops. How many “family farms” are left?
Rural towns have been evaporating because the farms don’t have the population density anymore.
One of the reasons I consider states like Kansas “green deserts”. They’ve got the plants, but the population density is getting closer to what you would see in deserts.