» A review of twenty one metropolitan areas shows that most are seeing an increasing percentage of their population growth — or a decreasing percentage of their loss — in their core counties.
Last week, the U.S. Census Bureau released its annual population estimates for counties as of July 2011. These data provide significant insight into changing population trends in the United States, and the results offer considerable support for the argument that the country’s growth is moving back into its cities, at least to some degree.
National coverage of the data release focused on the fact that the data showed a significant drop in residents moving to exurban counties at the edge of metropolitan areas. The massive creation of housing at the far reaches of regions appears to have slowed to a trickle, and even the movement of the population from Northeastern and Midwest metropolitan areas to Southern and Western areas has decreased. The fastest-growing counties by numeric population change between April 2010 (when Census 2010 was completed) and July 2011 were counties that contain large central cities — Harris, Los Angeles, Maricopa, New York City (if the five boroughs are combined), and Miami-Dade.
Of 21 metropolitan areas reviewed (chosen based on their size and presence of a central city), just five saw decreases in the population of their core counties between 2010 and 2011 (Cleveland, Baltimore, Cincinnati, St. Louis, and Detroit), while two of those also saw declines for the metropolitan area as a whole (Cleveland and Detroit). Many cities that have historically had declining populations, including Philadelphia and Washington, grew quite strongly over the year-long period.
But most important was the change in growth dynamic within each of the metropolitan areas (MSAs). 14 of 21 central counties experienced an increasing percentage of their respective region’s growth compared with the period from 2000 to 2010. This means that new growth in most regions studied was more concentrated in the central county than it had been in the 2000s. For example, whereas just 3.8% of the Washington region’s population growth between 2000 and 2010 occurred in the District of Columbia itself, 13.4% of the same region’s growth between 2010 and 2011 occurred in the central city. Most extreme, perhaps, was the situation in Cook County (the central county for the Chicago region), which took in 51.3% of the region’s population growth between 2010 and 2011, while the county had declined significantly in population between 2000 and 2010.
Three of the central counties reviewed saw declines in population but as a percentage of the region’s growth, those decreases were lower than those seen in the 2000 to 2010 period, indicating improved conditions there (Baltimore, Cleveland, and Cincinnati).
The exceptions were St. Louis and Detroit, whose central counties continued to shrink faster than the surrounding metropolitan areas, and Los Angeles and Boston, whose central counties continue to grow but not as fast compared to their respective regions as they did in the 2000s.
|Changes in U.S. County Population, 2000-2010 and 2010-2011 (21 metropolitan areas)|
|Above in sixth column, blue numbers means central county(ies) with an increasing share of the total regional population; red numbers mean central county(ies) with a declining share of the total regional population. Source: U.S. Census Bureau.|
These data reinforce statistical and anecdotal evidence from cities around the country about growth patterns. The 2010 Census demonstrated that in cities across the country, downtowns experienced dramatic growth in the 2000s, even in cities that suffered from overall population losses. With county-level data now available for the period from 2010 to 2011, it appears that that growth has extended throughout many of those surrounding cities — or at least been strong enough to dilute the effects of losses elsewhere.
In addition, the apartment market, which is heavily concentrated in central cities, is experiencing record vacancies and rising prices, demonstrating increasing demand for that housing product. In the Puget Sound region, for example, three-quarters of apartments are being built in the City of Seattle and half are being constructed in or near downtown.
Finally, a study released last week by U.S. PIRG and the Frontier Group shows that all these people moving into American cities are doing so in part for transportation reasons. Per capita vehicle miles traveled peaked in the United States in 2004 (and in fact, total vehicle miles peaked in 2007), showing that people are traveling less by private automobile. The report documented that the decline between 2000 and 2010 was especially steep among young people, who are moving quite dramatically towards bike, transit, and walking modes. The share of 14- to 34-year-olds with no drivers licence increased from 21% in 2000 to 26% in 2010. These young people are moving away from cars for both economic and non-economic reasons — the high price of gas is of course an issue, but so is the instant communication made possible through advanced mobile phones, for instance. These changes in transportation preferences support the notion that central cities are coming back.
It is hard to know to what degree these trends are reflective of a still-recovering economy. High gas prices, mass foreclosures, and difficulty acquiring loans certainly are likely to encourage people to stay put, decreasing what is called domestic migration, or movement from one U.S. city to another. The nation’s largest cities, which generally have high rates of domestic out-migration, are mostly reliant on international immigration to grow. So if the economic situation improves, the trends affected U.S. demographics between 2010 and 2011 may no longer apply.
Nevertheless, high gas prices appear to be a fact of the future, and the growth of downtowns, which occurred in the 2000s despite the mostly good economic conditions, both indicate that more popular cities may be a future fact of life. Even so, there is still work to be done: In only about half of the metropolitan areas did growth in the core counties as a percentage of regional growth match or rise above the existing population of those core counties as a percentage of the region’s population. In places like Dallas and Houston, for example, high growth rates in the core counties were not large enough compared to those counties’ share of existing population, meaning that effectively population continues to be distributed on the whole outside of the core. This was less true in cities like New York, Minneapolis, and Philadelphia, where the core county absorbed a higher percentage of regional growth than their existing population.
It should be noted that changes in populations of core counties do not always directly correspond with the performance of core cities — which is why I have included the rightmost column in the table above, showing the percentage of the core county’s population that lives in the central city (as of 2010). In some cases, county borders correspond directly with those of the central city (as in the case of New York City, New Orleans, and Baltimore, among others). In other cases, the central city is a relatively small percentage of the county population (as in the case of Miami, Atlanta, and Cleveland). Thus, while these data allow us to examine population dynamics of certain cities directly, it is inconclusive for others. Nevertheless, it is likely that the demographic situation experienced in core counties is similar to that seen in core cities. (And indeed, the trends recorded here appear to be just as valid for counties that are 100% central cities as those whose populations are primarily outside of the central city.)
The Census Bureau will release data for 2011 on central cities and neighborhood areas later in the year.
I should also say that the Census Bureau has previously released annual population estimates that diverge dramatically from the official decennial Census figures used for Congressional redistricting, which were last released for 2010 and which will next be counted for 2020. In 2009, for instance, the Bureau announced that Chicago’s population had declined by 44,000 since 2000; in 2010, it said the city had lost almost 200,000 in the previous ten years. Similarly, in 2009, the Bureau claimed that Atlanta had a population of 540,921; a year later, the city’s population had magically shrunk to 420,003. So we need to remember that these estimates are estimates.
* Both Cleveland and Detroit had both their central counties and MSAs lose population in both periods studied, and New Orleans had population loss for both central county and MSA in the 2000-2010 period. Thus, for example, Cleveland’s -86.0% from 2010 to 2011 was a relative improvement over -160.6% from 2000 to 2010, since this means that Cuyahoga County accounted for more than the region’s total loss of population in the earlier period but less than the total from 2010 to 2011 — which means population loss is spreading from the core county to the suburbs. St. Louis, Baltimore, and Cincinnati core counties all saw population loss in both periods while their respective regions grew, though each lost less as a percentage in the more recent period.
45 replies on “In New Census Data, An Improved Outlook For Core Counties”
This process will be slow, but it seems unstoppable now. Apart from Detroit, downtowns are just nicer to live in than rural areas these days, and the trends aren’t changing.
Los Angeles has crazy political geography and a bunch of the “suburbs” have far more coherent downtowns than Los Angeles itself does — it would be interesting to break down the numbers there along more useful lines.
St Louis City, which is technically a county of its own, contains just 15% of the metropolitan area’s population and consists mostly of downtown and slums. The slums are still emptying out, while downtown is experiencing an urbanist revival, still small in absolute numbers though. Thus the overall population is decreasing. To get a more accurate picture of the St Louis “core” you should look at St Louis City plus St Louis County – the latter contains the inner suburbs. I would be curious what the numbers for it are.
From what I hear, the same issue is present with Baltimore, though perhaps to a lesser extent. If so, then two of the three metros where the core shrank but the metro grew are due to artificial boundary choice rather than real demographic changes. Not sure what the story is with the third such metro – Cincinnati.
A more accurate description of St Louis City, like you correctly noted is also a county, is a rebounding core along with neighborhoods that exist next the metrolink, a stable south side and a north side that is becoming a ghost town. Need people to have Slums. and kinda take a offense to the term. St. Louis has some great neighborhoods with some great institutions and a lot of great events.
The biggest problem St. Louis is having at the moment is the declining number of families and/or smaller families. The biggest group to leave St. Louis city in the last census was childern if not mistaken. The public schools hit a low when I moved there in 2004, but have steadily improved and more and more charter/private school options exits in the city.
Great article….I like the breakdown with all the different percentage growth comparing growth in city versus the entire metro area. Unfortunately I feel New Orleans should statistically be an outlier because in the last decade of the extreme changes of the population due to the hurricane. However, living in New Orleans, I can attest to the fact that the downtown and immediate surrounding areas are experiencing a large revival that wasn’t there before the storm and population is moving towards a more central location, even if it is still less (total) than before Katrina. Examples include south market district and 1031 Canal street developments that are approved to be built right next to the new streetcar line they are building on Loyola avenue downtown.
Yes, I completely agree — New Orleans is truly an outlier here thanks to the effects of Katrina. It is hard to know how the city would have changed had the storm not occurred.
I’m having a hard time understanding the red/blue criteria in the table. If Houston has 50% of the regions growth in both column 5 and 6, shouldn’t its share of the regions population be growing (blue)?
In short, no. Harris County absorbed 63.3% of the region’s growth between 2010 and 2011, but it (the county) represented 68.8% of the region’s population in 2010, meaning that it absorbed less than its proportional share of regional growth. Houston (the city) represents 51.3% of the county’s population (and therefore 35.3% of the region’s population), but we do not yet have city-level data on growth from the Census Bureau.
So when do transportation models change? Lower VMT per capita is the new normal. Sadly, not too many regional transportation models yet reflect this reality. Of course, when you project more and more miles of driving, you plan more and more miles of highways.
It was always a Ponzi scheme with government absorbing the debt needed to “expand” the cheap-oil, consumer-driven economy. But now that the private side of that scheme doesn’t even attract the same number of buyers and consumers, it’s even more foolish to continue the public partner’s participation in the scheme.
Transportation models change when the old transportation modellers die or retire.
Kind of depressing.
It means it will take quite a while. Longer than it should.
I think Boston’s central county slower population increase is partly due to gentrification. for example – in the South End neighborhood – townhouses that had been split up into several rental units are reverting back to single-family residences, usually for wealthier older couples. So, a single building that used to have 10-15 individuals now has only 2.
Even the loss in cities like St. Louis, Cleveland and Cincy is due to gentrification. Smaller household sizes are how past Census challenges in these cities based on building permit data didn’t pan out in 2010.
Still, if you have improved per capita income, slightly less population ain’t such a bad thing for these smaller, older cities. That is to say, they may not have the bragging rights for “growth,” but their bottom lines (in terms of fiscal capacity) are still improving.
Yah, I’m wondering if absolute or percent population growth (or decrease) is the best measure of how cities are doing. While new housing starts were the highest in St. Louis in years for the 00-10 period, the actual population fell as families were replaced by singles and young couples. What is holding St. Louis (and I’m sure a lot of other central cities) back is the condition of the public schools. The lack of quality schools has led many young families to leave the city when their kids reach schools age.
In Boston, the city already grew faster than the suburbs. Going by the ACS, it posted double the population growth of the metro area between 2000 and 2009, and Cambridge posted almost as much population growth as Boston. Even going by the census, which lowballed all urban population growth numbers, Boston slightly outgrew the metro area.
The data, especially if you look at the Nytimes map, seems to show that pretty much all Midwestern cities continue to sprawl.
Milwaukee (my hometown) was interesting to look at because of how half of the city proper continues to grow in population and the other half (the North Side) continues to lose population (although the rate of decline seems to be decreasing). Interestingly also, wealthy postwar suburbs like Brookfield and the North Shore lost population while the exurbs grew by leaps.
Most other midwestern cities seem to continue to empty out pretty quickly..
The push towards sprawl in “Flatland” is partly due to the fact that, given the flat geography and political rules which allow suburban car-drivers to “free ride” on city services, it’s generally cheaper in the short term to sprawl. Changes in transportation costs change all that.
Regarding Milwaukee, it is a particularly weird case because the whole darn city is practically an exurb of Chicago at this point.
I would contest the idea that Milwaukee is an exurb of Chicago (at least yet, anyway). The two cities are growing together, but Milwaukee definitely has its own economy and growth patterns (which generally go towards Madison, actually) and only about 13,000 Milwaukeeans work in Chicago out of a metro of 1.7 mil.
I’m excited and worried about the coming spikes in the cost of driving. I think if it happens in the right way it could be a boon to places like Milwaukee and create sort of natural growth boundaries. But, if we keep delaying it and delaying it, I think it might be too late for a city like ours to adapt..
Great story. As futurists who predicted the halt of suburbs in 2012, here’s our next two predictions to watch:
1.Small town growth. The primary growth will be bigger urban centers, but smaller towns on train routes will also grow. Take a look at Galesburg, IL, for instance. On a great line into Chicago. It will double to triple in size.
2.30% decline in existing home values by 2020. Watch out. Existing home values won’t come back. New housing will grow tremendously in urban centers and towns with light rail and train stations nearby.
Clarification: Existing SUBURBAN home values will decline.
The growth in urban populations is definitely a result of rising fuel prices and high unemployment, but that doesn’t mean the trend towards public transit and bikes will reverse if/when fuel prices go down or the employment picture improves. Numerous studies have shown that when people begin riding bikes due to rising fuel prices they continue to ride even when gas prices go down. Urban living, public transit use and biking are addictive and a habit that is easy to keep once you start.
or a result of a drop in crime.
I would also predict that suburbs in the US will become more and more urban. Silicon Valley and Orange County are car-oriented sprawl, but they are job rich and increasingly dense places.
Suburbs that are far from jobs will go into permanent decline. Riverside/San Bernardino lacks good job opportunities and is separated from the rest of the Los Angeles by a handful of congested highways, without jobs it will not survive. Galesburg IL will never grow very much. It is 3 hours away from downtown Chicago, no one will ever commute this far.
Galesburg has the advantage that it’s an actual town with actual shops and whatnot, predating the rise of the modern auto-oriented suburb — and it is on a train line. It may dwindle but it won’t vanish. Riverside and San Bernadino proper have such characteristics too, though less so.
There are some endless-car-sprawl locations which really are doomed; the outskirts of Phoenix come to mind as the most extreme example I can think of. The entire area’s going to shrink (city in a desert, global warming) and it’s extraordinarily car-dependent.
If the cities boom rents go up encouraging businesses to put the back office in places like Riverside or Galesburg. Makes the commute 15 minutes.
The growth of urban areas and the decline of far suburbs seems promising. This should quiet those sprawl advocates who trumpet that people want to live in widely separated homes and drive cars everywhere (not ride transit).
Something that, again, bugs me, is how many commentators here are not only interested in providing transportation alternatives, but – this time – angrily waiting for “the doom of suburbia” as if the sizable portion of the population who lives there not because they are forced, but because they like the lifestyle, was irrelevant in this debate.
I’d hazard a guess that only about 25-30% of suburbanites live there because they “like the lifestyle”. It’s really irrelevant what people like anyway, if suburbia becomes economically unfeasible, people will leave it. Some will return to the old cities, some to small towns, others, the countryside. The only “force” that will be used to effect this will be market forces.
“Some will return to the old cities, some to small towns, others, the countryside.”
Small towns and the countryside are now known as “exurbs” – not what you want. (The difference between now and 100 years ago being that [due to farm automation] most people in semi-rural areas now have to commute long distances to large office centers.)
Some will return to the old cities, but even they have become less dense, and density cannot be constructed overnight.
The bottom line is that many people will suffer if the change does not come slowly.
“The bottom line is that many people will suffer if the change does not come slowly.”
One way to segue slowly out of the suburban/car/highway model we have into something else (i.e. something more urban in form) would be to have government programs encouraging densification, the moving away from a car-centric model to one centered more on public transit/walking/biking, prioritizing infill and brownfield development over greenfield development, higher gas taxes w/ more of the proceeds going toward public transit and “complete streets” programs, and alternative fuel research and promotion. I’m guessing though, this would be decried as “Communism”. I suspect we’ll continue to promote the current model until it crashes suddenly; there won’t be any “soft landing”, at least not in the US. I supect even after this crash we won’t learn our lesson…there will be a clamoring for leaders who can restore $2.00/gallon gas so we can all continue happy motoring and living in 8k square foot houses in the middle of nowhere.
“Small towns and the countryside are now known as “exurbs” – not what you want. (The difference between now and 100 years ago being that [due to farm automation] most people in semi-rural areas now have to commute long distances to large office centers.)”
True. Truly rural living should be something for farmers, foresters, and other people who actually *work* on the land. “Gentlemen farmers” should have to pay their own way and not live off government subsidies to endless rural highway and roads so that they can play on their pretend farms. Suburbia and exurbia have kind of ruined the city/town/countryside distinctions.
I don’t think government needs to start cherry-picking winners and loser on real estate markets or creating expensive and usually inefficient programs to favor “densification” (sic).
A good approach would be for the gov’t to create some infrastructure for urban/regional transit, define TOD areas along the axes and let the market sort out what will be build.
Assuming the exogenous trends will force some higher density transit-accessible development, providing the infrastructure and the proper zoning provisions would be enough.
As for the
it is just an unwarranted attack on small town living.
First, only a tiny fraction of total workforce and relatives (children, very elderly) are involved in actual rural work. Actually, you’d find that most people doing dirt work on farms (that are heavily automatized anyway) live in bigger cities and commute to their jobs like people commute for a – say – remote hydro plant or oil drill site without actually living close to it. People working a farm don’t necessarily are tied to the land and have interest on homeschooling children, or even putting the whole family to work in the same trade!
I agree that there more equilibrium is needed, but bear in mind more and more jobs are now non-5-to-9-in-an-office anymore. Telecommuting and many other trends are increasingly allowing people to dissociate ever more place(s) of work from residence by reducing the commute needs. Especially peak weekday commute.
Rural living should indeed be for people doing rural work.
Small town living is not the same as rural living. Look at Ithaca, NY. Ithaca == medium-small town. Trumansburg == small town. Village of Dryden = small town. Enfield == rural.
It’s not a good idea to fill up the rural areas with tract houses; we need those areas for farming. It’s been happening anyway, sadly. “Agricultural preservation” laws are trying to maintain the distinction.
There are many ways to get around these laws, especially if one is rich enough to buy the land and just not farm it altogether.
Moreover, a many cities are not surrounded by arable land, particularly in the Southwest and West Coast, so the point is moot.
The idea that people should have a “lifestyle” matching their work is kinda repugnant to be honest. We got past that state already.
I’d hazard a guess that only about 25-30% of suburbanites live there because they “like the lifestyle”.
The other 70% are there because the choice was living in suburbia or living under a bridge. It’s very difficult to choose to live in a dense walk-able neighborhood when all that is allowed to be built is single family houses on quarter acre ( or bigger ) lots without any sidewalks.
Most recent subdivisions (post early 1990s) do have sidewalks as a mandate. In many cases, sidewalks within the subdivision are indeed good quality, wheelchair accessible and safe (with a buffer to the road lanes themselves). More problematic are issues of proper pedestrian crossing on arterials bordering a subdivision, for instance. Crossings do not transmit a safety feeling when seldom used, so that small enhancements like special lighting, protuberant (in relate to the lanes) crossings or traffic lights that can be set on by pressing buttons are good measures.
As for the preferences on house tipology, I think that is very hard to measure on a strict defined % base, because the ability to chose your “dream house tipology” is vastly constrained by your ability to pay for it, which is heavily dependent on zoning, real estate market conditions etc.
What I do think is that less restrictive zoning could help. However, it is very hard to conceive how HOA could work well if they were responsible for subdivisions containing both houses of different type (rowhouses, single-detached and apartments in low-rises) and retail activity. The immediate interests of people in each condition will often clash and that would make HOAs an enormous mess. For instance, those living right close to a retail lot might not want any night time activity there, whereas those living 3 blocks away might want to get it.
So unless there you were suggesting to get past the concept of HOAs, I see some challenges there. But solutions are surely possible, like covenants that provide, ex-ante, extensive lists of what is or is not allowed to happen in each lot.
It’s very common for the subdivision to have sidewalks and for the arterials to NOT have sidewalks.
which is of course useless.
The state DOTs need to start putting sidewalks on their state highways by default.
Most arterials are roads kept by the counties, not by the state DOTs.
Which gets right back to the zoning easements that you rejected below on the ground that people bought houses in outer sprawl suburbia and not houses in suburban village clusters that under were not permitted to be built under the original restrictions on development in that area.
How you manage a multi-use walkable development easement in a specified vicinity around a stop on a specified quality of transport corridor is by fighting over the location of that stop. Once that fight is settled, the location of the prospective suburban village core is settled. Those which have been well sited will have suburban village cores developed in pursuit of the opportunities available within the easement zones.
The locations will commonly be either on an artery or main through access road, for road oriented common carrier corridors such as priority Express Bus, BRT, trolleybus or tramways, or at a common back where the stop is a station on a rail alignment presently dividing existing suburbs on both sides.
If what people want is a small town lifestyle, then its a lot easier to create a small, multi-use “suburban village” in a strategic location in a suburb than for everyone living in the suburb to try to find a small town to move to.
That is, the difference between a common suburban and a small town is the lack of any “downtown” in the suburb. So, for example, a multi-use district a quarter mile radius around a strategic common carrier transport point, with a denser residential easement a half mile radius around that point, would occupy 0.2 square miles and 0.8 square miles directly modified, but certainly a three mile radius around it would feel like a suburban fringe of that suburban village, so the suburban village would “redesign” at least 28 square miles of suburb.
Of course, many newer suburbs are not laid out in a way that make it possible to locate this suburban village just anywhere … but the location problem for a suburban common carrier transport stop is pretty much the same location problem, so the fact that the stop is targeted for a suburban village core wouldn’t seem to change the location problem in terms of locating it where it is as readily accessible as possible to its core hinterland.
If people want a small town lifestyle — there already ARE a lot of small towns. The problem is usually lack of jobs in those towns!
There was a period when we knew how to deal with this problem. Each town had this thing called a “railway station”, and the people living in the small town rode the “train” to work and back, and…. well you get the idea.
Very few people actually want to live in car-dominated car-centric 1950s Levittown suburbia. Streetcar suburbs are *way* more popular.
Do you have any base to make such an overreaching claim? Because if very few people indeed wanted the modern suburb, no amount of real estate lobbying would be able to push people to buy such houses.
That many people don’t actually enjoy driving a lot doesn’t mean that people would enjoy riding a train, or a bus, or walking. Many people dislike certain aspects of life, and for the majority of them using a form of transportation to/from work, anyone, is just a chore, not an experience the persons awakes longing for.
This is one of the major errors I find in so-called “transit activism”: it assumes that because people don’t like to drive, they would like to do anything else. They don’t: people just would like effortless transportation with comfort, safety, privacy and, of course, zero cost, while allowing them to also live in a nice residence with amenities and comfort and of course zero cost.
As zero cost housing and transportation don’t exit, people put up with commuting as a chore.
To make an easy comparison: that many people don’t like vacuum cleaning their house doesn’t mean they would love old-fashioned sweepers and wet cloth as means to remove dust from their houses.
Another comparison: that many people don’t like flying in an economy-class between America and Europe doesn’t mean that they’d either be willing to jump into ocean liners taking 5 days to complete an otherwise 9-hour flight ordeal.
If we go by “what people want”, of course we know that people do not want sprawl outer suburban developments to the extent that we have built them, because the market for sprawl outer suburban developments is worse than the market for residential property as a whole.
People tend to buy houses that are on the market in the region they are looking for a house in, and the property development that we have been subsidizing, and in part requiring with zoning regulations, makes it far easier to build modern outer suburbs than any other type of residential development.
If there are houses in old streetcar suburbs or old small towns that have had modern suburbs grow up around them, those houses often sell at a location premium, and of course houses and townhouses in old walkable inner suburbs often sell at a quite substantial location premium.
Indeed, its hard to see how your critique actually applies to either to a policy of removing sprawl subsidies, or to a policy of providing an automatic set of zoning easements that permit the establishment of multi-user, more densely settled suburban villages within existing suburbs. If, in fact, people do indeed prefer sprawl outer suburban developments, they will continue to buy sprawl outer suburban properties anyway.
My specific concern is not to overturn the HOA system, which works well in most cases as a more accountable body than most city governance models.
Once again, the same misconception: people are, or shouldn’t be bound to work or seek for work only in the jurisdiction they live in. This is a medieval concept attached to the serfdom principles of the time.
The major function of transportation is to increase access for people, so that they can reach ever-increasing (and never decreasing) areas to seek for work, daily (non-holiday) leisure, social activities etc.
Traveling unnecessarily long distances for daily commutes wastes resources, causes pollution, and creates traffic issues. Yes, people should live relatively close to where they work.
Lawrence, this is a whole new discussion, an interesting one though: in our modern World, which implies households where all adults work or are seeing work, and where changing jobs is more frequent, making what is probably the biggest financial decision of a lifetime (buying a residence), which is subject to high transaction costs (from taxes/fees to the whole expense of moving itself) based on some job placement that is unlikely to last 10 years, if so, is risky.
The days of lifetime employment in a single facility are over.
Of course, single-person households could have an easier time dealing with that, but even so the costs of moving, if you are anywhere above the “bare bone student room” level is an expensive proposition, especially if you own or have a mortgage…
In quite a lot of areas, they are a relatively small share of the total residential housing stock.
In what were once the rural areas adjoining a growing urban area, the countryside between the small towns will have been washed over by outer suburban development. And in those cases, houses in those small towns and the new developments in the immediate vicinity often enjoy a premium … but the incentive of the provisions for income tax capital gains from property development is to develop a pure greenfield site, so even outer suburban developments built to gain a location advantage from a nearby existing town tend to be located where their access to that town is car dependent.