Quick Note on Los Angeles and Chicago Density and Modal Split

A long-running conundrum in American urbanism is that the urban area with the highest population density is Los Angeles, rather than New York. Los Angeles is extremely auto-oriented, with a commute modal split that’s only 5% public transit, same as the US average, and doesn’t feel dense the way New York or even Washington or Chicago or Boston is. In the last 15 years there have been some attempts to get around this, chiefly the notion of weighted or perceived density, which divides the region into small cells (such as census tracts) and averaged their density weighted by population and not area. However, even then, Los Angeles near-ties San Francisco for second densest in the US, New York being by far the densest; curiously, already in 2008, Chris Bradford pointed out that for American metro areas, the transit modal split was more strongly correlated with the ratio of weighted to standard density than with absolute weighted density.

DW Rowlands at Brookings steps into this debate by talking more explicitly about where the density is. She uses slightly different definitions of density, so that by the standard measure Los Angeles is second to New York, but this doesn’t change the independent variable enough to matter: Los Angeles’s non-car commute modal split still underperforms any measure of density. Instead of looking at population density, she looks at the question of activity centers. Those centers are a way to formalize what I tried to do informally by trying to define central business districts, or perhaps my attempts to draw 100 km^2 city centers and count the job share there (100 km^2 is because my French data is so coarse it’s the most convenient for comparisons to Paris and La Défense).

By Rowlands’ more formal definition, Los Angeles is notably weaker-centered than comparanda like Boston and Washington. Conversely, while I think of Los Angeles as not having any mass transit because I compare it with other large cities, even just large American cities, Brookings compares the region with all American metropolitan areas, and there, Los Angeles overperforms the median – the US-wide 5% modal split includes New York in the average so right off the bat the non-New York average is around 3%, and this falls further when one throws away secondary transit cities like Washington as well. So Los Angeles performs fairly close to what one would expect from activity center density.

But curiously, Chicago registers as weaker-centered than Los Angeles. I suspect this is an issue of different definitions of activity centers. Chicago’s urban layout is such that a majority of Loop-bound commutes are done by rail and a supermajority of all other commutes are done by car; the overall activity center density matters less than the raw share of jobs that are in a narrow city center. Normally, the two measures – activity center density and central business district share of jobs – correlate: Los Angeles has by all accounts a weak center – the central 100 km^2, which include decidedly residential Westside areas, have around 700,000 jobs, and this weakness exists at all levels. Chicago is different: its 100 km^2 blob is uninspiring, but at the scale of the Loop, the job density is very high – it’s just that outside the Loop, there’s very little centralization.


  1. Eric Doherty🌻 (@Eric_Doherty)

    Paul Mees claimed that while higher population densities make it easier to create high quality transit networks at a reasonable cost, the importance of this one factor is often greatly exaggerated. I think he was correct that correlation between density and travel behavior isn’t strong unless you cherry pick your examples. The transit, walking, cycling and road infrastructure is important. https://rabble.ca/books/sustainable-transport-key-saving-planet/

  2. Eric2

    I think there is a threshold of density above which transit usage shoots up. Streets have a fixed capacity (in vehicles per lane per hour), and it is difficult and rare to increase this capacity by adding lanes. So there is a hard limit to how many cars can enter an area at a time, and it’s not that high a limit. Below this limit, excess demand will result in additional car travel. Above this limit, all of the additional travel has to be non-car modes. So maybe you can estimate this capacity from street layout, and figure out that Chicago’s Loop has X jobs above the capacity limit, while LA’s centers such as they are have Y jobs above the capacity limit, and thereby estimate the modal split, and see if it matches the actual modal split.

    • Henry Miller

      I think you are right, but be careful not to confuse cause and effect. Transit can maintain higher density, but it may or may not cause it.

      If an area has poor transit and high density, then the people who own property there will tear down buildings for more parking, and will not fight road expansion to get cars around (so long as they are paid for the land of course). However if there is good transit they feel less need for parking and so will build denser when they build (to better use the space they have), and will fight more roads that they don’t see a need for.

  3. A commenter

    I wonder how much of LA’s underperformance is due to it’s historically high parking requirements. I think design matters a lot in how people get around, maybe more than overall density

    • Lee Ratner

      High parking requirements certainly doesn’t help. If it is easy for people to find a place to store their car at their destination, they are probably going to drive. However, not having great parking doesn’t make transit use more common. Oakland isn’t especially generous with parking spaces and many times you need to park on the street. Oakland residents aren’t getting on A/C buses when they want to go to Lake Merritt though.

  4. Martin

    I think this generally illustrates that indeed it is impossible to have a strongly centered city center with mostly car access, and a low share of public transit. It is simply a geometric impossibility to have so many cars driving to a very small area that concentrates all destinations.

    The reverse is however not as true. You can maintain a relatively high level of density, if also the access points are more distributed across the place. Cars still face some geometric problems, but not nearly as severe as in the strongly centered case. You find this in third-world cities, with weak transit, whereas a result by necessity the urban structure must be quite decentralized as well.

    The problem is that the second case supports transit much worse than the first case, and if you want to transform the urban structure of such cities you indeed face a very big problem. The second case also of course does not really support either walking or biking as reasonable transportation options.

    To some extent, I think cities are a bit beyond hope for radical reform (except perhaps to develop some urban islands), in particular as it is likely that in the future self-driving semi-public transit (with say 5-10 passengers per car) likely will be able to meet the transportation demands of such places quite well. In other words, it will continue to be a viable functioning urban form, but will (because of that) continue to be as unpleasant to live in.

    • Eric2

      “self-driving semi-public transit (with say 5-10 passengers per car)”, being inherently slower than driving, won’t attract car drivers unless you give it separate lanes to speed it up. If car drivers won’t give up a lane for 100-passenger buses, I don’t think they’ll give up for 5-10-passenger vans. The lanes also have much more utility for a bus following a fixed route.

      Also they have a 5-10 passenger vans in my location (with computerize ordering and route scheduling, but not self driving of course) and in my experience they are worse than buses, not to mention rail. Their advantage over buses is in personalizing the trip to bring it closer to each passenger’s origin/destination. But this equally means that the route is full of diversions and turns and is much more nauseating than the bus while being only marginally faster. I much prefer to walk a few minutes to a bus stop at either end of a long straight route. The problem seems inherent – in any multipassenger vehicle, a more convenient trip for me means a more slow and nauseating trip in order to be convenient for the other passengers. Better to have a frequent bus network with a bit of walking from the stop.

      • Martin

        So I also think everyone who can afford it will own/order a personal self-driving car instead. I do however think that for everyone else something along the line of a 5-10 person vehicle will easily outcompete a traditional bus network (on price if nothing else). Particularly given that now in a city like LA where the bus share of transit trips is already very low, the service is very bad, and many buses will average less than 10 riders for a large part of the day. This will largely be an upgrade for transportation options in poor transit cities in the US. If the city-wide usage of a 5-10 person vehicle goes up, it will also be relatively easy to coordinate roughly similar starting points/end points, in particular during rush hour. This is in clear contrast to something like a shared van from an airport which serves very marginal demand.

        • Henry Miller

          Who is this everyone else that cannot afford a personal self-driving car? Not enough people to support a network of 5-10 person vehicles.

          If you don’t buy the latest new car every 3 years, but instead buy a 3 year old used car and driving it for 9 years or so, then the personal car is probably cheaper. Shared cars they have to fix the worn out seats and other cosmetic issues, while most people with a use car just live with it.

          Those shared 5-10 person cars also use a lot of fuel – do they even have enough ridership to be less fuel than private cars? Remember they sometimes have to drop the last person off and then are running empty to pick the next person up, while the private car just sits in a parking lot.

          A good fixed route mass transit system is the only way to usefully bring people to transit. The economics of small vehicles just do not work out to be enough better than a private car if you can afford one. (it might be slightly cheaper, but not by much and your personal car is always in your driveway when you want to go) Though I’m talking about the full costs, with subsidies it might be cheaper for you because someone else is paying part of the price.

    • Lee Ratner

      Tokyo is relatively multi centric, not only in the 23 wards, but in the Greater Tokyo Area but also has probably the best or at least one of the best transit systems in the world. I think transit with multiple centers is possible as long as you have ways to get from one center to another without having to cross into the center every time.

      • Mark

        The difference between Tokyo and LA being, of course, that each of Tokyo’s multiple centres is the size of a large North American downtown. If you rammed 23 Chicago Loops into the LA Basin, it would probably have pretty good transit ridership too!

      • Martin

        I think it is more accurate to see Tokyo as pretty monocentric, with a gigantic center in the middle that has a somewhat uneven density (the central 5 wards, and some adjacent areas). That is clear when you look at the commuter-train network. It would be pretty unusual to take a long within-city train trip that would not pass through this area.

        However, that is not really my point. I think rail transit indeed can uphold nearly any amount of density. It was more that if you have a more car-centric transportation usage, and medium-high density, that is easier to achieve with a less mono-centric layout.

    • Henry Miller

      People will use transit over driving if transit is convenient. That means not too slow, and very predictable. It doesn’t have to be faster than a car, it just needs to get you there in a reasonable amount of time – under half an hour.

      Strong centers make that easy as since everyone is going to the center you can provide service only that way, and if it is still too slow you add express service (skip stop…) to speed it up.

      You can provide good transit without a strong center, but it is much harder. Now since people are going every which way you need service in all directions from everywhere. My best suggestion is your provide fast service to local centers (never more than 10 minutes ride), and then express service between all the local centers. Make transfer easy – many people will have to make a 3 transfer trip to get where they want to, which is already a big no-no in transit so you need to do everything you can to make it easy.

      • Lee Ratner

        I think that price also comes into play. You need to make driving expensive enough not to be worth it for many people.

        • Henry miller

          Stop, that line of thinking just gets you voted out of office and all your reforms repealed.

          Even though there is little a politician can do about gas prices, running on an ‘I will reduce the price of gas’ is a popular platform. If you can do something about the high cost of driving you look like you filled your promise.

          Even in the time when you (or your side) is in power, you will spend all your political capital making driving expensive, and so be unable to make transit better. Since most people don’t have a realistic option to not drive, your higher driving costs will make them mad.

          Focus on making transit better. Give people a realistic option to not drive. That means serve more than people who work 9-5 in the center of the city (note singular center). People who work non standard shifts. People who are going to the doctor, church, and other non work trips. People who work in the suburbs.

          Sure you can’t serve all of that easily, and some might be impossible. But there is a lot of easy stuff we are not doing and could. Learn to build and run cheap transit so you can build and run more for the dollar. Spend effort figuring out where people really go and redesign your system for them.

          • Lee Ratner

            I know that this is politically impossible in the United States. I am still doubtful that a pure carrot approach can get more people taking transit though.

          • Henry Miller

            Perhaps, but if you suddenly gave me a reasonable budget (with no political opposition) it would be several decades before we would have the type of transit system in place that we can start talking about the other incentives.

            Today we need to focus on making transit useful at all, not how to force people to ride it.

  5. Phake Nick

    This blog remind me of some statistical metrics used in Japan when measuring activeness of a region.
    Night time population … The ordinary population data, counted by home of people
    Day time population … Count where people locate in the daytime, aka if people work at or study at certain area, then they would count as a member of population of the place they work or study in, but if they either work or study in the same place they live in or if they do not work/study then they would remain as a member of the place.
    Inflow/Outflow population … Indicate the movement between different towns/districts/etc in the daytime due to people getting jobs or schools.
    Daytime-nighttime population ratio … Find out the ratio of daytime to night time poulation. If a place have more population at night compared to day time then it mean the place is a bedtown where people town to go to other place for work. On the contrary, if a place have more people in daytime than at night then that mean more people come to the place and the higher the ratio it mean the more it can attract people into the area.

  6. Lee Ratner

    I’m wondering if Australian cities would have been a good model for American or Canadian transit during the 20th century. They are basically large sprawling connections of suburbs, Melbourne and Sydney are even less dense than the Dallas-Fortworth Metroplex if wikipedia statistics are correct, but have high frequency commuter rail networks. Melbourne also has the world’s biggest tram system and Sydney is building light rails and true subways. Maybe that would have worked better for the United States Sun Belt cities than a transit system designed for a denser city.

    • Alon Levy

      Yeah, I should say, at the 100 km^2 scale, Sydney and Melbourne have low job centralization by New York or Paris or Vancouver standards (but high by Los Angeles ones). I don’t know what their structures of job density are, though – I just saw this on a list of census areas and their job densities, I don’t know if they’re centered on train stations the way much of Stockholm is.

      • bensh3

        Sydney for sure qualifies, with centers like Chatswood and Parramatta. It’s fortunate that Sydney’s motorways run parallel to rail lines, whereas in Melbourne there are even fewer useful highways to begin with.

      • Lee Ratner

        Australia for some reason decided not to rip up and even expand on their rail networks even though it was a very suburban place where nearly everybody could afford a car like the United States and Canada. I can see Los Angeles County having something like an Australian high frequency commuter rail network. BART in the San Francisco area is something like an American version of this.

        • adirondacker12800

          BART sucks. It’s all local all the time. In nice round numbers it has the same amount of route as Washington D.C’s Metro and Washington has twice the ridership. Or Atlanta has half the amount of route and half the ridership. And you don’t want to compare it to things in the Northeast or Chicago.

          • jonahbliss

            I’d say it’s fairer to categorize it as “all express, all the time.”

          • adirondacker12800

            You can categorize it as super duper extra special fantastic. It would still make all the stops all the time.

          • Wanderer

            Stop spacing on BART is very wide, once you get out of San Francisco and central Oakland. They’re typically 2-3 miles apart, even more in some places. That’s not metro local stop spacing. Chris Spieler thinks that BART should add more stations in low income East Oakland, there are only two on a roughly 10 mile segment between Lake Merritt and San Leandro. BART’s original planners said BART is the express service, local service needs to come behind it.

          • Mark

            I reckon that’s a pretty fundamental challenge with building a metro system that runs 75 km from the city centre. Infrastructure is expensive, so 4-tracking it doesn’t really make sense for the vast majority of its (low-demand) length, but the status quo is too widely-spaced to be terribly useful. Seems to me the solution is just to not build metro lines 75 km from the city centre.

          • Lee Ratner

            BART was always intended to be a high frequency commuter rail service and there needed to be a way to get people from the East Bay suburbs into San Francisco and Oakland without them driving. MUNI light rail is the real metro service from San Francisco.

          • jonahbliss

            The RER makes “all the stops all the time” but we still understand it to be an express complement to the Metro. Here’s an example that makes it clearer, if you’re going from SF Civic Center to Balboa Park it’s 4 stops on BART, but at least 16 or more on Muni (depending on the line)

          • Eric2

            It’s 9 stops from Dublin/Pleasanton (the far end of the east bay) to Embarcadero. You could easily fit another 5 or even 9 stops in there and it would still be an acceptable journey time, even from the very last stop on the line.

          • adirondacker12800

            I’ll bite. If I’m at Civic Center and want to go someplace other than select places on Market Street or Mission Street I’ll use MUNI. Or someday in the future I can be at Beale Street just south of Market. I can walk about a block to BART and go allllllllllllllll the way to San Jose’s Didiron Intergalactic Multimodal. Or I can a walk a bit farther to Transbay terminal and take Caltrain. Or if I’m feeling flush, spring for the fare on high speed rail.

          • Sassy

            If you’d add local service by quad tracking the existing corridor, then you can add local service just by adding more stops and having timed overtakes. Outside the central section, there’s room for those services. The problem is that you’d either have to build more capacity into and through SF, or turn trains around in Oakland, but that has nothing to do with the demand in the deep suburbs not justifying quad tracking.

            The other problem is that success would involve tons of transit oriented development at those new local stops, which is unrealistic consider the lack of transit oriented development at the existing stops.

          • jonahbliss

            I’m probably opening a can of worms here, but… if you compare WMATA and BART, they get almost the exact same number of riders per *station.* Add more stations, and you’d get more riders – as the current system is so spread out that a new station’s walkshed wouldn’t compete with an existing station;s.

          • Lee Ratner

            @Sassy, they are doing much more transit oriented development around BART stations these days. The entire Walnut Creek Parking Lot is getting turned into apartment buildings. I think similar developments are happening at other stations but I’m not going to travel to another station and investigate. Defeating NIMBYs is tough just about anywhere but California is a special case.

            The Bay Tube is running at peak capacity or at least it was before COVID. If Californians were willing to splurge, a second Bay Tube running from downtown Oakland into Alameda and then into SF on the east side before connecting to the market street tunnel would be a good idea. Like many good ideas, it won’t happen.

          • adirondacker12800

            If the tunnel is at capacity under the Bay it’s still going to be at capacity west of the shoreline.

        • Tom M

          Perth almost did. Under the conservative state government in the late 1970’s, the passenger service on the line from Fremantle to Perth was suspended,with the line to be turned into a busway. However, with the change in government in the early 1980’s, the line was returned to passenger service and electrified. Electrification sparked an change in fortune for the rail system, with many new lines since constructed to keep pace with the city’s growth.

          • Tom M

            And the Sydney Tram Network, that was larger than Melbourne’s, was ripped up in 1961.

          • Lee Ratner

            Perth is one city and the least major of the big metro areas in Australia.

          • Tom M

            Adelaide would be the “least major” of the big metro areas (>1 million)

  7. jb

    The charts reminded me that Philly underperforms LA in both population density metrics, and is only marginally higher in activity center job density, but has stronger numbers for non-drive alone commute (albeit weak for the Northeast.) If you were to drop SEPTA’s regional rail onto LA, how much would that impact SoCal’s commutes? (Or for that matter, if you were to shrink LA’s roads to Center City’s size / cut down parking to match Philly’s unparked rowhouses…)

    The other variables that I think are unexplored in the piece are geographic barriers / toll roads (and I guess the self reinforcing local customs / vibes) – every US transit city basically has some hard edges and road constraints that push people onto transit; how would LA look different if the Sepulveda Pass behaved more like the Bay Bridge? (On the flip side, this is why Miami’s terrible transit use is even more disappointing…)

    • Alon Levy

      As of the 2019 ACS, the Philadelphia CSA’s commute modal split is 8.2%, of which commuter rail only contributes 1.8%; the Los Angeles CSA’s is 3.8%. (I haven’t seen an MSA-level breakdown of transit shares by mode of transit – I’m basing this on B08006/sex of workers by means of transportation to work, not the finer-by-geography/coarser-by-mode means of transportation to work by selected characteristics.)

      • adirondacker12800

        That isn’t particularly clear. I can make assumptions, it’s not 3.8 percent of workers in L.A. use rail. Not cargo cult Metrolink, anyway.
        There is “drive alone to work”. “use public transit”, whatever that is, and “something else”. carpool, walk, bike, canoe across the Charles from Cambridge to Boston. And work-from-home. Philadelphia likely has a higher share of not-drive-alone and not-public-transit because it’s more likely to be able to walk or bicycle to work. I suspect work-from-home might be higher in Los Angeles because it’s easier to do if you have a three bedroom with a two car garage.

        • Alon Levy

          (Yeah, the LA CSA is 3.8% all transit. Nowhere in the US is 3.8% commuter rail; New York is the highest at 3.7%, followed by Chicago at 3.1%.)

          Yeah, I’m specifically looking at public transit whereas Rowlands is looking at everything omitting driving alone. US-wide, the most popular mode that is neither driving alone nor public transit is carpooling, which in most of the country (though not in LA) is lower-income than both – it’s for people who live and work in the same geographies as people who drive alone but don’t have money for universal car ownership or occasionally do but choose to carpool for green or communal reasons.

          The New York CSA is also 5.5% walk, yeah (LA: 2.2%) and 5% work from home pre-corona (LA: 6.2%). Poking around, the pre-corona wfh rates don’t look like they correlate with much – Dallas is 6.5%, Houston 5.1%, Miami 6.4%, Atlanta 8.4%, Boston 5.6%, Detroit 4.6%, Philly 5.6%, the Bay Area 6.2%, DC 6.1%, Chicago 5.7%.

  8. Robert Campbell

    One of the difficulties with density calculations is the presence (or lack) of water used to calculate density. New York City and to a lesser degree, Boston, have many shore areas which give a larger “land mass” which are not built upon (like Jamaica Bay). Some census tracts are either park land or bodies of water. It’s a mess to get populated census tracts data. Los Angeles has a very smooth shore line and not too much of it. Use of urbanized area statistics can bet a better answer for urban density. At this point there is no good census data for the 2020 census.

    • jonahbliss

      This is a pretty slippery slope argument – but I’m pretty sure most water adjacent metros could make the same excuse. LA / OC has the BAllona Wetlands, Dockweiler, the mouth of the San Gabriel River, Seal Beach National Wildlife Refuge, Bolsa Chica Ecological Reserve, the mouth of the Santa Ana River, Upper Newport Bay, and depending on how you want to count this – the country’s busiest port areas. Then there’s of course the region’s notorious mountains, which are generally parkland…

    • adirondacker12800

      The watery census tracts or the park tracts or the tracts for reservoirs or … wouldn’t have many people in them or any people in them and wouldn’t get included in the “urbanized” tracts. The suburban tracts that are all-mall wouldn’t either.

  9. Ryland L

    Its sad. LA has so much potential. If anything the trend in recent years has been further commercial decentralization: Downtown’s rebirth has sputtered, while hybrid work has made “edge cities” like Century City less relevant.

  10. Pingback: News Roundup – Seattle Transit Blog

Leave a Reply

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

WordPress.com Logo

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

Facebook photo

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

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.