The weird economic geography of transit use in Chicagoland

Stereotypically, American transit users are low-income. In some places, like Chicago, transit is good enough, at least in some places, for middle-class and even upper-middle-class people to use it. But even here, surely more money makes you more likely to drive?

But not quite. In the Chicago metropolitan area – meaning these numbers include the suburbs, not just the city – the median income for a person who drives to work alone is $39,957. The median income for someone who takes transit to work is $40,314.

It gets even weirder if you make a map:


In short, blue areas are places where the median transit-rider’s income is higher than the median car-driver’s income. As the colors get deeper red, transit-riders’ incomes are falling relative to drivers’.

The results are pretty much the opposite of what you would expect: the suburbs are almost entirely blue, meaning that an average transit-rider from, say, Naperville, is actually richer than an average driver. In the city, that’s reversed: drivers tend to be wealthier than transit-takers. (If you only look at the city, things look a little more explicable: transit-riders’ relative income is highest in gentrified North Side neighborhoods with good transit access, especially along the Blue, Brown, and Red Lines.)

What’s going on? I can’t prove it yet, but I strongly suspect that it has to do with where different kinds of jobs are.

Our normal paradigm of how someone chooses to use transit has to do with income: we assume that in any given situation, transit is probably some amount less convenient but also some amount less expensive than driving, and so people who are more price-conscious – that is, lower-income people – will be more likely to sacrifice convenience for the sake of saving money. Under that model, rich people should be more likely to drive pretty much everywhere, but as the convenience of transit improves – that is, in dense cities – the gap should narrow.

But what this suggests, I think, is that income is playing a smaller role than job location. If you live in the suburbs and work in the suburbs, transit is almost certainly a terrible option for you. Metra runs infrequently and there are few jobs to walk to from Metra stations; Pace also has long gaps between buses on many routes, and is too slow to make efficient trips across the vast distances beyond the Chicago city limits. That means that you’ll have a very strong incentive to drive, even if you’re low-income. (And recall that in places where transit is very, very bad, even most low-income people will find a way to get access to a car so they can participate in society.)

If you work in the city, though – in particular if you work downtown – transit might be a great option. Metra will drop you off within walking distance of your job, doesn’t require paying for downtown parking, and may very well be faster than taking highways into the Loop at rush hour.

That means that suburban residents who work downtown are much, much more likely to take transit than people who work in another suburb. Which wouldn’t mean anything by itself – except that the types of jobs that locate downtown don’t look like the types of jobs that locate in the suburbs. Although downtown Chicago has plenty of service sector jobs – the people who work in restaurants, supermarkets, and clothing stores – it also has a massive concentration of high-paying white collar positions. That means people who work downtown are disproportionately likely to have upper-middle-class or upper-class white-collar incomes. Which means, in turn, that transit riders in the suburbs are more likely to have those incomes as well.

Suburban jobs, in contrast, are a wider mix of service sector, blue collar, and higher-paying office jobs. That means that drivers – who will disproportionately be people who work in the suburbs – will, on average, have more average-looking incomes.

I would take two big things away from this. The first is that improving transit, which is frequently cast as a social justice issue (including by me), does not automatically benefit lower-income people more than higher-income people. (Which doesn’t mean that it isn’t worthwhile – there are other reasons to support transit.) In terms of who benefits, there’s a big difference between, say, adding a rush hour express train from Naperville to the Loop and upgrading suburb-to-suburb Pace buses – or, for that matter, buses in the city.

But the second is that you can create major transit benefits without actually doing anything at all to transit service. These maps show that wealthier people are taking better advantage of transit infrastructure in the suburbs – not because they have better access, but because of where their destinations are. That means that bringing a wider range of jobs to transit-accessible locations – in downtown Chicago, but also other Metra stops, or places where Pace buses converge, or other easily-accessible places in the city – may matter as much or more than building new lines or services. And recalling that most trips aren’t actually commutes, this applies to other kinds of destinations as well: grocery stores, other kinds of shopping, and even homes.

Which means that if you care about people being able to save money by taking transit – or the mobility of people who can’t drive, including the disabled, the young, and many of the elderly – a major part of our program needs to be about focusing new construction near transit. That doesn’t require any big multi-year studies or multi-billion-dollar federal grants: it just requires some zoning changes. Let’s get on that.

As an addendum: of course, in many places, even if transit-takers are wealthier, there are very few of them. If you remove all the places where fewer than 5% of residents take transit to work, you see that much of the Chicago suburbs is very car-dependent – but the overall pattern remains the same.



19 thoughts on “The weird economic geography of transit use in Chicagoland

  1. Thanks for this post. Let me just add that when I looked at various Westchester suburbs a few years ago, I saw the same pattern. In fact, this pattern even holds for people who work in Stamford and Greenwich: transit riders are richer than drivers, since transit riders are disproportionately likely to be reverse-commuters from New York who are attracted to the corporate HQs.

    This has to do with agglomeration economics. The CBD has the most expensive land. This means that the uses that will locate there a) are not land-intensive, b) benefit from agglomeration more, and c) can afford to pay high land rents. All of this points toward office towers, populated by professional services firms (hence, they’d like to locate near other major firms) that pay high wages. Even industries like tech don’t have to be in the CBD, and that’s why we see Google operate out of the Meatpacking District and not Midtown. This agglomeration effect is fractal in the sense of appearing to a smaller extent in secondary subcenters, so that when transit is somewhat useful for reverse-commuters, we see the same pattern. (I suspect that if you count tech shuttles as transit, transit users’ incomes well exceed those of drivers for people working in Mountain View.)

    1. Thanks! Yeah, obviously we can’t fit all employment in downtown, for both space reasons and the economic ones you mentioned. I think this is a case for a) a strong grid network of bus lines where possible, and b) planning for lots of other relatively transit-accessible secondary nodes.

  2. Reading through I’m glad we came to the same conclusion. It’s corroborated by my own experience growing up in Highland Park. Dad worked in the loop at a high-paying job at First Chicago/NBD/Bank One/Chase and took Metra every day, only driving maybe once every 3-6 months. A nice benefit for me and my brother as we became old enough to drive, since we could just use Dad’s car. Mom worked as a teacher in Northbrook, not making nearly as much as Dad, and she had to drive. So that fits the pattern.

    Overall, I think it also shows how important it is to consider service availability when measuring mode share. So for example, is it really fair to measure transit use throughout the whole metro area when in the suburbs generally the only viable option is suburb-to-downtown transit trips. Thus if the mode share for that specific kind of trip could be measured, I think the data would be a lot more useful.

  3. Very interesting, helps explain some oddities in how transit riders are viewed. This is certainly about job location, but it’s also about commuter rail as a mode from the suburbs to Downtown. Commuter rail is pleasant compared to other transit modes, often economically segregated (doesn’t stop in low income neighborhoods), very heavily subsidized and generally considered high status.

    1. I’m not sure that’s right: I’m working on a post for City Observatory on this theme, and you see the same pattern in the Twin Cities and Seattle, which don’t have extensive commuter rail but do have extensive suburb-to-city express buses. What matters is the service, which can be high-quality buses or commuter rail.

  4. I suspect the pattern is the same in Toronto too. Commute mode share to downtown jobs is about 3/4 non-auto (mostly transit) for both workers living in Toronto proper and those living in the suburbs. For city residents they’re mostly taking the subway, often using feeder buses or if they live close to downtown streetcar/cycling/walking. For suburban residents it’s mostly commuter rail and commuter bus. The commuter transit is a premium service, fast, comfortable, and expensive, and the users appear to be overwhelming upper middle/upper class white collar downtown workers, especially during peak periods.

    However it is not very heavily subsidized, in fact GO transit has the highest farebox recovery in North America. I wouldn’t say that it skips low income neighbourhoods either. There is a relative lack of stations in the urban core but that’s true regardless of whether the neighbourhood is poor or wealthy.

    Job location definitely is a big factor, as I said auto mode share is similar for downtown workers living in Toronto proper vs the suburbs, despite the suburbs being less dense, wealthier and more auto-oriented in built form.

    On the other hand, auto mode share for downtown residents reverse commuting is almost double that of suburban residents commuting to downtown, so that despite downtown’s smaller residential population, the number of downtown residents commuting by car is similar to the number of suburban residents commuting to downtown by car.

    Office parks have abundant free parking, and are much more likely than residences to be located near highways (at least in Toronto) making them very accessible to cars. On the other hand, the massive parking lots often separate them from transit, and spread out built form and location on curving circulator roads and highway frontage roads (rather than arterials) makes them difficult to serve by transit.

    I suspect that most cities will have much higher transit mode shares for downtown jobs than even jobs located along rapid transit outside downtown. Most North American cities have radial rapid transit systems, so while Elizabeth and Brooklyn residents can get to Manhattan easily by subway/commuter rail, getting to Elizabeth from Brooklyn would be less competitive relative to driving. And New York still favours transit much more than most in this regard since the geography forces drivers to take a similar path as transit riders when going from one part of the suburbs to another. Looking at google maps, it suggests going through Downtown Chicago to take transit from Elgin to Naperville or Rockville, MD to Tysons Corners, while drivers can just take a much shorter route cutting across the suburbs.

  5. While I suspect locating non-downtown job centres near transit won’t yield the very high transit shares of downtown, it’s still a good first step to locate them along a rapid transit line or at the very least so that they are easily accessible to the arterial bus grid. Once you do that, perhaps eventually you can build up the ridership to serve these job centres with additional transit lines, including possibly multiple rapid transit lines.

    The advantage of downtowns is that they are not only served by transit, or even rapid transit, but multiple transit lines with direct service to much of the metropolitan area. If you look at some cities overseas like Seoul or Tokyo their built form and transportation systems have matured to the point where they have several rapid transit lines criss-crossing each other all over the place, so that you have several nodes that are truly downtown-like. I’m not sure if any North American cities have reached the point where secondary job centres truly need more than one rapid transit line, but focusing job growth around transit definitely seems very important if that’s the future you want to head towards.

    1. It would be interesting to do the charts for a city like Houston, which is multi-polar and has express buses converging on two of the subcenters (Uptown and Texas Medical Center) in addition to Downtown. In many of the areas around Beltway 8 and beyond, commuter transit to park and rides is the only transit in town, with the possible exception of a bus running once an hour or similar.

    2. Right, yeah. It’ll always be less convenient to get to non-downtown areas, even if they’re along transit lines–especially in a highly radial network like Chicago’s. But there are still a lot of improvements in accessibility to be made. I think the best “multi-line” secondary nodes in Chicago are places like Jefferson Park, downtown Evanston, and Oak Park, where there are L stations, bus hubs, and commuter rail stops. Unfortunately, I think Evanston is the only one doing any serious TOD around that hub.

      1. Oak Park is also doing some significant development around the Harlem/Lake Green Line Stop. New high rise and low rise residential buildings are being constructed with attached retail / commercial space. These are replacing one parking garage at lake and forest and 2 large surface parking lots at Harlem to both to the north and south of the el / metra tracks. It’s a good start

  6. (this is a non-technical comment from a car and transit user in Rome). In my opiniion, nothing “weird” about it – economics are the main tool to use in planning, but you have to be totally systemic in putting the data together. My job requires going to clients’ in all areas of a supremely traffic-challenged city: My choice of car or transit is based exclusively on what’s available at the destination location, and the choice can sometimes be painfully expensive (=parking fees).

  7. As Nicolas said, I think this might be a common pattern for any city that has express (as in relatively high speed) transit to downtown and the downtown has a high portion of well paying white-collar jobs. I’d expect the same for much of Long Island.

    I know some of the highest transit ridership spots of Long Island are affluent communities, though not all the high transit use neighborhoods are affluent. Scarsdale in Westchester has a median household income of $233k / year and a public transit commute share of 40%, a more middle income community number has somewhat lower transit use.

  8. A few questions:
    1. Chicago is already doing what you’re asking, and it’s not helping black people. When the Cook County Land Bank reinvested money won over banking crimes, the money went intentionally to upper class areas, so black neighborhoods were dropped. Another accidental slight against our black citizens, on transit specifically, the Lincoln bus and several others in black neighborhoods were slashed. The Blue Line and the Red line somewhat were renovated. It would be a great thing to build up the city for the people, but the Cook County Land Bank and the Mayor should not be calling the shots!
    2. Why do Chicagoans accept this? Most of people have a racist feeling that blacks are criminals and don’t belong in their neighborhood. Besides not being racist, another way to counter this is if Chicago were able to hold credit independent of the equity tied up in their houses. Property value engenders a hysterical fear that any sort of bad reputation in their neighborhood will cost you your retirement. This in turn supports the police, tough on crime policies, a new cycle of underdevelopment etc.

  9. Two comments:
    1. Work commuting is not about convenience. it is all about time and predictability. Commuters choose a route that is the fastest way to get from home to work day after day after day AND if the chosen route runs off schedule how often is it off schedule and by how much time is it off schedule. For instance, at most non-center city El stations in Chicago there are daytime parking issues due to car-to-El commuters. The car-to-El commuters likely have a bus route as an option in lieu of the driving portion of the commute but they drive, I think (and I used to be one), because the bus system is much less predictable than the rail system. (I’d be curious to know what percentage of ridership is this type and wondered why the CTA doesn’t operate parking garages at some mid-city el stations adjacent to highways like Jefferson Park or Irving Park on the Blue line).
    2. Chicago’s Commuter rail and el systems are such extremely radial systems that the only real use of the system – especially at the periphery – is for work commuting. Are there any systems in the US that are more of a gridded pattern than radial to further test this hypothesis? For instance, the inner portions of the DC Metro system is relatively gridded and less radial (though once your get 5 or so miles form city center it becomes radial.

    1. 1. “Time” and “predictability” are both, for me, part of “convenience.”

      2. Yeah, exactly. And are there? Not really; there are systems where the “center” is larger than others. In New York, for example, most of Manhattan south of Central Park might count as the “center,” but it’s still usually a pain to get from, say, outer Brooklyn to outer Queens.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s