What Suburban Poverty?
Myth: American cities have undergone inversion, in which poorer people are more suburban than richer people.
Reality: at least on the level of people commuting to city center, wages generally rise with commute distance. In particular, the phenomenon of supercommuters – people traveling very long distances to work – is a middle- and high-income experience more than a low-income one. This is true even in Los Angeles, a Sunbelt city with more of a drive-until-you-qualify history than the Northeastern cities. The only exception among the largest US cities is San Francisco, and there too, the poorest distance is 5-10 km out of the Financial District.
All data in this post comes from OnTheMap and is as of 2017, the latest year for which there is data. The methodology is to define a central business district, generally a looser one than in past post but still much smaller than the entirety of the city, and look at people who work in it and live within annuli of increasing radius from a specific central point within the CBD. OnTheMap puts jobs into three income buckets, the boundary points being $1,250 and $3,333 per month; we look at the proportion of jobs in the highest category.
I report the annuli in kilometers, but technically they’re in multiples of 3.11 miles, which is very close to 5 km.
| City | New York | Los Angeles | Chicago | Washington | San Francisco | Boston |
| CBD | 3rd, 60th, 9th, 30th | I-10, I-110, river | Congress, I-90, Grand | 6th, R, river, E | Broadway, Van Ness, 101, 16th | I-90, water, Arlington |
| Point | Grand Central | 7th/Metro Center | State/Madison | Farragut | Market/2nd | Downtown Crossing |
| Jobs | 1,017,175 | 310,111 | 558,379 | 249,707 | 441,104 | 241,775 |
| 40k+ % | 68.7% | 68.8% | 70.4% | 69.8% | 73.2% | 71.7% |
| 0-5 km | 211,910 | 22,557 | 67,348 | 56,578 | 86,845 | 41,912 |
| 40k+ % | 79.8% | 44.6% | 84% | 75.1% | 76.6% | 70.5% |
| 5-10 km | 205,215 | 38,986 | 91,332 | 56,154 | 67,063 | 52,499 |
| 40k+ % | 63.6% | 53.2% | 70.1% | 61.5% | 68.3% | 64.3% |
| 10-15 km | 172,117 | 42,391 | 88,604 | 38,233 | 49,111 | 30,619 |
| 40k+ % | 51.9% | 65.1% | 58.8% | 66.2% | 73.6% | 73.3% |
| 15-20 km | 101,543 | 41,229 | 67,620 | 23,589 | 35,692 | 20,444 |
| 40k+ % | 62% | 71.3% | 67.3% | 69.8% | 74.5% | 76.8% |
| 20-30 km | 92,871 | 53,809 | 68,571 | 27,921 | 40,170 | 29,271 |
| 40k+ % | 74.4% | 75.6% | 73.5% | 75.8% | 76.9% | 79% |
| 30-40 km | 61,236 | 33,051 | 49,374 | 15,568 | 33,395 | 17,511 |
| 40k+ % | 81.1% | 77.6% | 76.1% | 78% | 80.1% | 77.8% |
| 40-50 km | 37,931 | 17,561 | 41,745 | 8,403 | 20,509 | 12,738 |
| 40k+ % | 82.1% | 81% | 78.7% | 82% | 82.4% | 78.7% |
| 50-60 km | 26,746 | 13,853 | 25,872 | 3,346 | 15,981 | 9,321 |
| 40k+ % | 81.3% | 82.3% | 74.9% | 76.7% | 78.7% | 76.9% |
| 60-70 km | 21,860 | 8,561 | 14,940 | 2,596 | 14,682 | 6,101 |
| 40k+ % | 80.3% | 83.6% | 74.5% | 76.9% | 73.3% | 71.5% |
| 70-80 km | 14,007 | 7,720 | 5,471 | 1,444 | 9,151 | 4,757 |
| 40k+ % | 77.8% | 79.7% | 72.4% | 79.3% | 69.8% | 74.2% |
In all six metro areas above except Los Angeles, the income in the innermost 5-km circle is higher than in the 5-10 km annulus. In Chicago that inner radius is in fact the wealthiest, but in Boston it’s below average, and in New York, Washington, and San Francisco it is poorer than wide swaths of suburbia. There is always a large region of poverty in an urban radius, which is roughly the inner 15 km in Los Angeles, the 5-20 km annulus in New York, the 10-15 km radius in Chicago, and so on.
This of course does not take directionality into account. In Chicago, it is especially important – to the north, there is wealth at all radii, and to the south, there is mostly poverty. In contrast, in New York directionality is less important, and it is in a way the purest example of the poverty donut model, in which the center is rich, the suburbs are rich, and the in-between neighborhoods are poor, without wedges that form favored quarters or wedges that form ill-favored quarters.
The importance of this is that because the inner and outer limits of the poverty donut are slowly moving outward, there is talk of suburbanization of poverty – or, rather, there was in the decade leading up to corona, but I suspect it will return once mass vaccination happens. However, even now, American cities are not Paris or Stockholm, where wealth mostly decreases as distance from the center increases, even though both cities have intense directionality (rich northeast, poor south and west in Stockholm, and the exact opposite in Paris). The poorest place remains the inner city, just beyond the near-downtown zone at what I would call biking range from city center jobs if any American city had even semi-decent biking infrastructure.
This contrasts with various schemes to subsidize suburbs that assume poverty has already suburbanized. Massachusetts, where even in the inner 5 km radius the $40,000+ share is below average, has a concept called Gateway Cities, defined to mean roughly “low- and lower-middle-income cities that aren’t Boston.” Of those, about one, Chelsea, is inner-urban, while the others include Springfield and various ex-industrial cities that are generally no poorer than Boston and lie amidst suburban wealth, like Lowell and Haverhill. Based on the idea that Massachusetts poverty is in the Gateway Cities and not in Boston itself, it justifies vast place-based subsidies that mostly go to people who are decently well-off while Dorchester has to beg for slightly better public transportation to Downtown Boston.
In New York, one likewise hears more about the poverty of Far Rockaway than about that of Harlem. There’s this widespread belief that Harlem is no longer poor, that it’s fully gentrified because there’s one bagel shop on 116th Street that caters to a mostly white middle-class clientele. This is related to the stereotype of the Real New Yorker, weaponized so that the cop or the construction worker who is a third-generation New Yorker and lives at the outermost edge of the city is an inherently more moral person than the Manhattanite or the immigrant and is the very definition of the working class while earning $90,000 a year. This goes double if this Real New Yorker lives on Long Island, usually with some catechism about how the city is too expensive even though the suburbs are about equally costly. The one place-based policy that would benefit the city, having the state integrate its schools with those of the generally better-resourced suburbs, is unthinkable.
It’s notable that this discourse that overrates how poor American suburbia is comes exclusively from people who tend to sympathize with the poor. People with Thatcherist attitudes toward the poor abound in the United States, and tend to correctly believe that the inner city is poorer than the suburbs, and if anything to overrate the extent of urban poverty. In either case, the conclusion groups of Americans reach is that the government must subsidize the suburbs further; all else is just motivated reasoning.
In reality, if one has the Thatcherist or Old Tory moralistic attitude that poverty is a personal failure then, with reservations, one should continue believing the large American city is inherently immoral. But if one has the attitude that poverty is a social failure that is solvable with social programs, then one must realize that there is more of this in central cities than in their suburbs, even faraway suburbs that are called drive-until-you-qualify because they are slightly poorer than some other suburbs, and therefore if anti-poverty programs must be place-based then they should be urban.
Pedestrian Observations 
The claim isn’t about absolute numbers but rather that poverty is rising in suburbs faster than in cities.
The distance-to-CBD definition of a suburb is not good. Many of the annuli here contain a mix of SFH suburbs and high density urban areas. The 15km ring in Chicago Loop contains Edgewater ($50k median income, 13,000 p/km2), Oak Park ($100k median income, 4300 p/km^2) and Riverdale ($14k median income, 800 p/km^2). You can’t conclude much by putting all those into the same bucket.
+1
The Edgewater annulus definitely does not include Riverdale, which generously seems 18+ km from the Loop. Edgewater is at most something like 12 to 13 km from the Loop.
You are absolutely right, I messed up with Google Maps. Edgewater would be more on a line with Englewood ($19,854, 3,265 p/km^2, 12km to State/Madison), and Riverdale would be more like Evanston (median income $103k). I think my point stands that these rings contain many communities of differing suburban-ness and poverty.
Yeah Alon did mention that Chicago definitely is not isotropic in how the poverty is distributed.
So is the analysis ignoring people who don’t commute to CBD?
I noticed this too, and was a bit disappointed by the analysis. “Supercommuters are a middle and upper-middle class phenomenon” ties in with middle and upper-middle class jobs being more urbanly concentrated. The working-class and working poor in the suburbs and exurbs are rarely commuting to the CBD! ‘Rarely’ of course permits a lot of exceptions, but plenty of commerce happens outside the CBD proper, and that commerce employs plenty of people outside the UMC bubble.
“The one place-based policy that would benefit the city, having the state integrate its schools with those of the generally better-resourced suburbs, is unthinkable.”
Yes making students commute two plus hours a day by bus is practical and a good idea. Man and people say I don’t live in the real world.
At my public high school plenty of students commuted 2 hours a day (1 hour each way), and there was no shortage of students and parents willing to accept that commute in exchange for going to the school. Admission was highly sought after and quite competitive. Heck, people pretended to live with their aunts in the city rather than their parents in the suburbs just to get in, and commuted from said suburbs every day.
Well that sounds like just as F– a situation.
I chose my high school in Germany in part because it was in a more interesting place than the alternative and the bus ticket would be valid all day every day
That’s 1 hour each way rather than 2 hours – very different. And I agree there is no shortage of bright teenagers willing to work themselves half dead for a shot at an Ivy League university – but this isn’t and shouldn’t be the norm for everyone.
They already have long commutes on school buses to greenfield school campuses, they’re just segregated.
And then they would have a long ride from there to wherever you imagine sending them.
“they’re just segregated.”
Well to some people its a religion instead of reality.
Well, if you shove them on public transit that may actually be the same amount of time or faster than having to wait for a school bus.
When I was applying for schools in New York, I got into a school that was a 1h30 transit ride from my house that started at 8, and a closer “neighborhood” school that was too close to qualify for the transit pass, but was quite a walk away, and the school bus that I would’ve been put on would’ve been a 1hr ride for a 7AM start. So going to the farther school actually meant I could wake up later.
Other than a few Northeastern cities and Chicago there is no public transport. If they are so far out in suburban splendor they need a bus to get to school they would need a bus to get to the public transport. If there was any.
No they didn’t. Mostly they had parents drive them 10 minutes to school, until they were old enough to drive themselves in 10 minutes.
Citation needed
My experience in a suburban US high school
Your experience does not necessarily represent a norm.
Here’s parents in Charleston, SC complaining about a 5:50am pickup time for elementary school kids: https://www.wistv.com/story/32815540/parents-kids-to-be-picked-up-by-school-buses-before-6-am/
Fairfax County VA had a documented case of a school bus route doing pickups at 5:45am: https://www.edweek.org/leadership/opinion-in-the-dark-about-early-morning-school-buses/2013/01
Raleigh, NC with 5:25am pickups: https://abc11.com/archive/9282462/
Orlando with 5am pickups: https://www.orlandosentinel.com/opinion/os-school-start-times-scott-maxwell-20150418-column.html
The issue is that over the years school districts across the nation do not want to pay so much money for school buses, so less buses do the same job and take more time.
at least on the level of people commuting to city center, wages generally rise with commute distance.
Because there are plenty of low paying equally shitty jobs closer to home.
“The one place-based policy that would benefit the city, having the state integrate its schools with those of the generally better-resourced suburbs, is unthinkable.”
Except the New York and Los Angeles school districts provide a test case for this. Both cover the entirety of large cities (and in the case of LA some adjacent communities. They fully integrate core cities and suburbs; both districts have more students than the populations of some states. The wealth from the Upper East Side funds schools in the Bronx, students in Howard Beach have the same curriculum and policies as schools in Bed-Stuy. Same for Marina Del Rey and West Compton, the part of LA with the 90210 area code and East LA.
The result of pooling all of the wealth instead of letting Staten Island go their own way? Two school districts with terrible performing schools and in many cases worse outcomes for students than in places with lower school funding.
It’s hard to see how, in general, funding schools in low-income areas the same as schools in high-income areas, will not help things.
For most places, schools in low-income areas get less funding than schools in high-income areas. That certainly isn’t good for students.
Most school funds are not used for things that benefit education. A nice football stadium doesn’t benefit education. Nor does a the latest electronic whiteboard (instead of a chalk board) make much a difference. Money to schools mostly goes to pay bureaucratic administrators, buy the latest teaching fad textbooks (ie line the pockets of politically powerful publishers), gold plate the buildings, and make the sports stadium better.
I’m not against more more money where it helps. However there is a lot of ways to waste money in education and politicians are less willing to look that way.
Wait, what? New York does not have school integration. That the Upper East Side exists doesn’t mean it does; as a city, it’s 33% non-Hispanic white, but the public schools are collectively only about 15% white, because white parents flee to the suburbs. You do not need to have a clean cleave as between Detroit and its suburbs to have segregation. (And no, the performance at NYCDOE isn’t terrible; it’s one of the better inner-city districts in the US. Very different from Los Angeles with its Californian levels of school funding.)
There is little interest in integration from wealthy districts because it would cost them money. The term “good schools” is often just shorthand for “well funded schools serving wealthy, predominately white students”. The first step is to get rid of local funding. Schools should be funded based on need (which means poorer districts would have *more* money, not less). But of course, while America talks a lot about education, the truth is, we don’t give a shit about children (case #342, the pandemic, and how creating a safe environment for kids was lower priority than opening the bars).
The really sad part is that an integrated school system benefits all students. The assumption is that only poor Black or Latino kids benefit, but wealthy White kids benefit just as much. As someone who was voluntarily bused in the 60s and 70s, it is pretty clear why. Suddenly you are forced out of your little socio-economic bubble. It is bizarre that some parents still encourage their kids to take a “gap year” and travel around Europe, to see other cultures, but they ignore the value of cultural mixing that can occur in their own city. I got to see the real world — or at least the real America — very early on. and was a lot smarter for it.
Education in the United States is very decentralized, especially in funding. At most the states just set broad standards and do standardized testing. Even in states require that funding be rather equal across the districts like California, you have wealthier districts asking for a certain donation for each kid in school, can be from $500 to $2000, and doing big fancy fund raising galas. For public schools. I’m sure that rates equalization in other states will lead to the same issue.
Berlin suburbs across the state line (e.g. Schönefeld or Potsdam) all behave a lot less like east German towns of that size and a lot like wealthy communes unsure where to dump their excess money. Similar things apply to Landkreis München, but at least Munich has a wealthy state government to shower them with wealth
Figures for Canadian metropolitan areas looking at all households (not just commuters) show a similar trend: https://uwaterloo.ca/atlas-of-suburbanisms/sites/ca.atlas-of-suburbanisms/files/uploads/files/hh_income-final.pdf
(NB: Ignore Toronto… its CMA is so different from the Greater Toronto Area that is useless for this sort of analysis).
Ooh, interesting how Vancouver is the exact inverse of an American city – the 5-10 km annulus is the wealthiest, including most of the Westside (but not Shaughnessy, which is in the 5 km circle!) and North and West Van.
All of metro Vancouver could move to Brooklyn. The rest of British Columbia could move to Queens. Metro areas in the U.S. that would be comparable to Vancouver B.C. would be Portland OR or San Antonio TX.
https://en.wikipedia.org/wiki/List_of_metropolitan_statistical_areas
Google Maps says it’s three and half-ish miles from Grand Central Terminal to City Hall. Things will look different and work different depending on the size of the metro area.
“Myth: American cities have undergone inversion, in which poorer people are more suburban than richer people.” In 43 years in the United States, I have literally never heard anyone say this. Discussing gentrification and “hip” young people moving to denser cities, it’s stated that the *trend* may be towards richer people moving to urban areas, but I had heard/read no one claim that the overall demographics are even close to the point at which the city centers are wealthy.
Really? I hear it all the time. I read it all the time. It has been become a major discussion point in transit circles, let alone outside it. For example: https://seattletransitblog.com/2020/11/12/kc-council-quarrels-over-metro-service-guidelines/.
It is worth noting that the plan is to simply move the money to the southern suburbs of Seattle, based on the assumption that it needs the most in terms of extra service. Partly this is due to the relatively high ridership during the pandemic, partly it is because people just assume that poverty has moved to the suburbs. The logic for the former is full of holes. The evidence for the latter is lacking.
Unfortunately, this is a very common assumption, and it has real, negative impacts. If your goal with a transit plan is to serve as many low income, historically disadvantaged people as you can with the available money, they you should still base your decision on data. Unfortunately, efforts to do that fall to the wayside, and stereotypes govern the decisions: https://seattletransitblog.com/2020/11/30/choice-riders-and-not/.
I heard an analyst who I generally respect (nobody you’re likely to have heard of) tell me, “Harlem is the new Upper West Side,” and that was in 2007. And in Massachusetts, the concept of Gateway Cities assumes poverty means Worcester and Springfield and Lowell and New Bedford and not Dorchester and East Boston and Roxbury.
I’d be interested to see the analysis sliced the other way – what is commute distance as a function of income? My expectation would be the shortest commutes would be at the lowest incomes (because low-paid service sector jobs are everywhere) and the highest incomes (because they can afford to live close to their sparsely-distributed jobs).
Even if the low income jobs are everywhere, they aren’t neccessarily distributed in exactly the same way as the poor people. You might have suburban clusters of warehousing and big box stores quite removed from the unfavoured quarter where the poor live.
In New York the distribution of low-income jobs is for the most part the same as that of high-income jobs except less concentrated. So the job centers are the same (except for the airports, which are big job sites only for the poor), but the overall concentration of jobs in those center is lower.
But it doesn’t usually matter that much for transportation, because those off-CBD low-income jobs are generally sites of consumption for the middle class.
Alon, you of all people know how weakly centered American cities are, and then proceed to completely discount the experience of low and middle income non-CBD commutes, which are the majority of commutes in ALL American metros. The poor people in the burbs aren’t commuting to the CBD and have a REAL problem of high transportation costs, high housing costs AND dispersed social services that your analysis completely ignores.
I redid the analysis without the CBD commutes. I’ll blog it in a few days, but LA still has city center poverty and suburban wealth, and all the other regions except SF have prominent poverty donuts. This matters, because LA is unusually weak-centered, and its inner 10 km radius is way poorer than the outer radii of Boston and such that get vast subsidies for commuting to Downtown Boston (SCR, again).
The table makes a good case that having transit fares increase with distance for trips over 10km is broadly reasonable. It doesn’t seem to imply much about non-transport-related anti-poverty schemes since the poorest residents of the outer suburbs don’t commute to the CBD (or anywhere at all, in many cases) and are therefore not counted here. And even the transit fare case suffers from the potential confounding factor where if long-distance commutes are sufficiently expensive only the rich will have them.
I can’t find the link (no pun intended), but Seattle Transit Blog published a very good article on the paradox of the unfair fare. Each kind of fare regime produces trade-offs that create an unfair class of riders, and it’s unavoidable. The article outlines the benefits and drawbacks of flat and incremental fares by time, distance or geographic zone.
I think there are a couple reasons for this. Partly it is due to people spotting trends, and then assuming they have played out. If you asked me the percentage of people who have land lines, I would say somewhere around 20%, knowing that trends take a while to play out (it took a long time for people to replace their VCRs). The reality (as of early this year), is 40%. Yes, it is less than half, but it is still a huge number, way more than I thought, and probably way more than the average person assumes. This really doesn’t matter much with phones, because the trend is clear. But it matters a lot with housing, because things can easily reverse themselves. (Cities were dying, then they were gentrifying, now they are dying again (depending on who you ask)).
There is another issue, that is based on the idea that everyone has the same wants and desires. If you want a nice three bedroom house with a nice yard and room to park a couple cars, your best value is the suburbs. If you assume that this is a universal need, and you assume that everyone would rather live in the city, then it stands to reason that poverty has spread to the suburbs. But this just isn’t the case. Plenty of low income people would rather live in an apartment in the city, and sacrifice the amenities some consider essential.
Also consider the lens the problem is being seen through. An academic lens looks far at broader categories, at longer timelines and in aggreated statistics. This post is written in an academic style, with a thesis, data and arguments.
A media lens looks at trends, snapshots in time and gathers more granular data, anecdote and narrative.
These aren’t opposites but contrasts. Each has its advantages, but media and academic texts each have their own narrative holes and never get to “the truth” by themselves. The reality is that poverty continues to concentrate in inner cities, but it doesn’t mean that poverty in the suburbs is non-existent or is not growing. Media reports can see what academic analyses miss, and vice versa.
My general sense is it’s far wiser to conduct transfers individually and not based on place, or group, or service type, or whatever else. Sacrifice some perfection for simplicity. EG I think our transit would be far better (and equitable, in a case of literary irony) if we had kept equity concerns out of it.
In policy thought, I sometimes see a counterproductive dismissal of other people’s justifiable interests. I’m strongly in favor of upzoning, eg, but consider someone who *recently* bought into a very expensive metro. Our system for schooling definitely has antisocial elements, another eg, but consider that parents’ sacrifices for their kids’ schools are often among the biggest sacrifices they make. Those individual choices aren’t antisocial; harm to those individuals’ interests is a cost that might need to be borne for the greater interest, but recognize it’s a cost, mitigate best you can, and be respectful to the people you expect to bear it.
Intended as general food for thought, not as a targeted criticism, so forgive my vagueness.
Well, others have already pointed out that non-CBD commutes shouldn’t be ignored. I’m going to point out two more things.
1. It is outright ridiculous to treat San Jose as a suburb of San Francisco, and a little bit ridiculous to treat Baltimore as a suburb of Washington DC, although it is not ridiculous to treat Newark as a suburb of New York City.
2. For other cities you have circled an area of around 5 to 6 km^2 (for NYC, 3 km^2) for your model CBD. On the other hand, for Los Angeles you have circled more than 10 km^2, but only 310000 jobs are inside the “CBD”, of a city with 10 million population. As you have noted, your model does not work for Los Angeles. Your solution to enrich the urban core also won’t work for Los Angeles, because in fact Los Angeles does not really have a CBD at all.
Frederick, the Los Angeles city population is 4 million. The Los Angeles County population is 10 million. And the Los Angeles Central Business District has the largest single concentration of jobs in Southern California.
California is peculiar in that cities have to develop as cities, and balance land uses among residential, commercial and industrial. Proposition 13 forces the fiscalization of land use. Unless an area is stupendously wealthy and can live off of property taxes alone, most suburbs have to take on commercial activities like a shopping center, auto dealership row or industrial park.
Is Oakland a suburb of San Francisco even though it’s smaller? Did San Francisco become a suburb of San Jose because Santa Clara County had room to grow? They are all principal cities. Stockton is a principal city yet also serves as an exurb to the Bay Area and Sacramento.
Baltimore isn’t a suburb of Washington D.C. They are in their own metropolitan statistical areas. So are San Jose and San Francisco.
Frederick is probably using consolidated statistical areas, which do couple Baltimore-DC and San Jose-San Francisco.
The whole point of CSAs is that they have MSAs in them. If there was a lot more commuting between them they would get merged into one MSA.
It’s seldom that a CSA devolves into a coherent MSA, largely because the definitions are political rather than academic. If an area loses MSA status as a distinct category, it’s a psychological blow if a principal city gets demoted. It’s also a fight politicians might want to avoid, since the fights would be brutal and the stakes so small. (I had argued in the Governance and Liberal Cities thread that MSAs and CSAs would be a better subsidiary unit of government than states as they currently exist since they are more coherent economic units and would better balance out some of the wide political and economic disparities created by the current regime.)
CSAs do serve a useful purpose, even though as you correctly point out it can overamplify commuting patterns when they really might be occurring on the edges in small numbers, though probably sufficient enough to support some form of peak-hour commuter bus or rail service or lifeline local buses to the service’s edge. In Northern California, you see this in the Solano County area, where the parts closer to Vallejo tend to commute to the Bay Area and Vacaville to points east tend to commute to Sacramento. San Joaquin County (Stockton) has this phenomenon, too, though Stockton is a substantially large city in the shadows of the Bay Area and Sacramento. So you have some communities that have commutes to somewhere in the Bay Area to the west and southwest and Sacramento to the north. (Note: I am referring to MSAs here. The Sacramento example doesn’t necessarily imply commuters are traveling within walking distance of the Capitol. It can be an inter-MSA commute from, say Lodi to an industrial park in Elk Grove, Vacaville to Davis or Marysville to the SMF airport.
Definitions come from the American Community Survey and other sources. It’s where people commute. It evaluates things on the county or county equivalent level. Have to cut it off somewhere.
So, previewing the post that I will upload tomorrow: the Baltimore and San Jose issues definitely matter, in opposite directions. If you do the same analysis with the same metro center points but without the CBD restriction, then you see a dip in incomes for people within 50-70 km of Farragut, because that’s where Baltimore is, and Baltimore is big enough to be a large share of the population within that annulus. In San Francisco, which is the closest thing to a drive-until-you-qualify region, it’s the other way around, i.e. you see a bump in incomes in the 40-60 km zone because that’s where the southern Peninsula towns are, like Palo Alto. However, it’s completely legit to look at the behavior of CBD-bound commuters as a subclass, esp. when you have Massachusetts saying that better commuter rail to New Bedford will improve social equity.
I don’t count the Los Angeles city population because the boundary of Los Angeles city looks like a product of gerrymandering. I know it isn’t, but the looks are similar. And is throwing out Santa Monica or Beverly Hills or Huntington Park or East Los Angeles actually reasonable at all?
Los Angeles “CBD” having the largest concentration of jobs in SoCal is also a moot point. I’m going to quote an article from City Journal here.
‘This downtown push has done little to resuscitate the city’s economy, however, probably because it misses the entire point of Los Angeles, says Ali Modarres, chairman of the geography department at Cal State Los Angeles. Remember that L.A.’s economy is built on multiple cores, which operate largely independently of downtown. Downtown Los Angeles employs a mere 2.5 percent of the region’s workforce; New York’s central business districts, by contrast, employ roughly 20 percent. “To put the entire focus of development on downtown L.A.,” Modarres says, “is to ignore the historical, cultural, economic, social forces that have shaped the larger geography of this metropolitan area.” ‘
Frederick, you specified Los Angeles as a city of 10 million. Los Angeles has defined borders as a city, as does the county, which does have 10 million residents and also has odd shaping because of mountains and population clusters. This does create odd commuting patterns. Southern California has plenty of intercounty commuting patterns as well, primarily from the Inland Empire.
Governmental units and commuting patterns don’t neatly line up, and commuting patterns tend to focus on highway movement while looking less at surface street traffic, where most vehicle distance travel occurs.
The City Journal quote has a bitter motte-and-bailey flavor to it. Ali Modarres’ quote is factual on its own, but in service of a hard-to-defend argument. City Journal conflates governmental units and regional patterns. Also, it is premised on counterposition where one might not exist. It treats economy as a zero-sum competition pitting the central business district against everyone else, and the downtown push is an economic pull against other neighborhoods, cities and counties. And if multiple cores can operate independently of downtown, that would imply each core has agency. By Modarres’ and City Journal’s own logic, New York’s CBD(s) (it has multiple districts too!) are failures because they failed to capture the 80% of jobs outside of them.
The “On the Map” website is great, but limited. It looks at employment, but not household income. Does anyone know of a resource for the latter? Ideally it would present the data as this does — focused on concentration. For example, the number of people per square meter living below the poverty level (or twice the poverty level, etc.). This is a hot topic in King County (Seattle) where they are considering changing priorities for the bus system (https://seattletransitblog.com/2020/12/11/updating-metros-service-guidelines/). There is very little information as to how they have determined “equity” measures, and I’m curious as to whether it comes close to lining up with concentration of poverty.