Density, Car Ownership, and What It Means for the Future of Los Angeles

The number of cars per person in Greater Los Angeles
The number of cars per person in Greater Los Angeles. For a full copy of the map, ##http://la.streetsblog.org/wp-content/uploads/sites/2/2010/12/LA-Vehicles-Per-Person.jpg##here.##

Density. Vehicle Ownership. The number of cars packed into a small area.

These are concepts that we discuss often on Streetsblog. Thanks to a UCLA research project undertaken by Professors Mike Manville and Donald Shoup, we can get an idea of some of the challenges Los Angeles’ planners face in trying to ween our city off the automobile.

This article will look at the population density, car ownership per person, and car ownership per mile maps and charts.  At the top of each article will be a “Streetsblog sized” map for Los Angeles with links to maps for San Franscisco and New York for comparison purposes.  All maps are based on information from the 2000 Census.

A huge hat tip to Katie Matchett, without whom this article wouldn’t have been possible.

Population Density:

Click for population density maps of Los Angeles, New York and San Francisco.

12 13 10 dense

Our first set of maps show the population density for the region. The greater Los Angeles region is often portrayed as a giant suburb.  When you look at the raw number for Los Angeles, as compared to New York and San Francisco, this characterization seems suspect.  After all, Los Angeles exceeds New York when it comes to person per square mile.

Population

Urbanized Area

(sq. mile)

Population Density

(person/sq. mile)

Los Angeles

11,874,000

2,980

3,990

New York

18,091,000

5,500

3,290

San Francisco

3,019,000

720

4,200

This table shows the total population, size of the urbanized area, and population density for each region. Note that while San Francisco has the highest population density, Los Angeles is more dense than New York.  However, when you look at the maps you’ll see that the population is spread relatively evenly throughout Los Angeles, while the other regions have dense centers with more sprawling suburbs.

In other words, when it comes to New York, our suburbs are denser than their suburbs; even though they have a much denser urban core.

It will be interesting to see in the 2010 census how much of a dense urban core has developed over the last ten years.  I suspect those dark green areas will continue to grow, but that nobody is going to start mistaking the Downtown with those of San Francisco or New York.

Vehicles Per Person

Click for the vehicles per person maps for Los Angeles, New York and San Francisco

The number of cars per person in Greater Los Angeles

It’s no surprise that New York had the lowest rate of car ownership per person, but I think that many people will be surprised to see that Los Angeles actually has a lower rate of car ownership than San Francisco. As with population density, vehicle density is more evenly distributed in LA than in the other two regions.

Total Vehicles

In Urbanized Area

Population

Urbanized Area

(sq. mile)

Vehicles per Person

Los Angeles

6,433,000

11,874,000

2,980

0.54

New York

7,771,000

18,091,000

5,500

0.43

San Francisco

1,769,000

3,019,000

720

0.59

Matchett provides a couple of other facts about Greater Los Angeles, which further illustrates what does and doesn’t determine car ownership

The three tracts with the lowest vehicles per square mile and per person are located in Long Beach, downtown Los Angeles, and San Fernando. A closer examination of the demographics of each of these tracts reveals that while each has a fairly large population (between 300 and 1,000 residents), the tracts have hardly any workers or housing units. This suggests that they may contain institutional uses such as jails, where vehicle ownership rates would be very low. The downtown tract might also include homeless residents without vehicles.

The three census tracts with the highest number of vehicles per person are located in the cities of Beverly Hills, Newport Beach, and Norwalk. The Norwalk tract also includes primarily commercial uses and has only nine residents, so the high vehicle ownership rate is likely due to the tract’s low population. The two other census tracts with a high number of vehicles per person are located in some of the wealthiest communities in the Los Angeles region. Given the strong correlation between income and vehicle ownership, this should come as no surprise.

In other words, Los Angeles’ surprising low rate of car ownership seems to have more to do with income than progressive transportation ideals.

Vehicles Per Square Mile:

Click for vehicle per square mile maps for Los Angeles, New York, and San Francisco

12 13 10 vehicle

From a planning standpoint, this set of maps and statistics, combined with the first set of maps and statistics tells the most important story.  What is the impact of population density on car ownership?  As we try to provide alternatives to the automobile, it stands to logic that dense development is part of the key.

Total Vehicles

In Urbanized Area

Urbanized Area

(sq. mile)

Vehicles per

Square Mile

Los Angeles

6,433,000

2,980

2,161

New York

7,771,000

5,500

1,413

San Francisco

1,769,000

720

2,460

Oh, well this is embarrassing.  Matchett explains.

The pattern of vehicles per square mile in each region basically mirrors the population density: tracts with high population density also have a lot of vehicles. This is true even in New York, with its lower vehicle ownership rates.

Honestly, I was surprised by this result.  I wasn’t exactly stunned that this was true for Los Angeles, because our transit system still needs some development, but that car ownership in New York and San Francisco was still high in their dense, transit rich areas?  I didn’t have a great explanation for this, but fortunately Matchett supplied an explanation from Professor Manville.

From a policy standpoint, this suggests that simply increasing density is likely to exacerbate rather than mitigate congestion–something we see borne out by most congestion data. (Increasing density and congestion can sometimes allow people to make more trips while avoiding congestion, but the congestion itself is still bad). So the trick for transportation and land use policy is to find ways to pull apart density and vehicle use. That’s what pricing does, and that’s what minimum parking requirements do the opposite of. Parking requirements make it very easy for increases in density to move in lockstep with increases in vehicles, because new dwelling units automatically include housing for cars.

So there’s the challenge for our local planners and transportation engineers.  As Los Angeles grows and becomes more transit diverse in the coming years, the city, county and Metro needs to get rid of parking minimums in dense, transit-rich areas and find other ways to encourage people to not feel the need to own cars.  It sounds as though Metro ought to be urging cities to relax their parking requirements in the areas around their new rail lines or, at a minimum, get them relaxed for the developments on land that the agency owns.

  • This is an awesome analysis! Go Katie, Mike and Shoup-dogg.

  • Love the wonky entry. The first step in democratizing planning is democratizing the understanding of complex issues. Well done, DN, KCM, MM, and Shoup-Dog.

  • S,

    I was inspired by your post on The Source on Friday.

  • Hahahahhaaaaa! Awesome :-)

  • anty

    So many questions! Are these numbers for the total population or just those of driving age? Is there a way to get at vehicles per household? Do all three metro areas have similar proportions of typically undercounted populations (or are these those adjusted numbers conservatives like to pitch a fit about)?

  • Great post, but links to Vehicles per Person maps aren’t working!

  • Katie M.

    I’m always happy to do my part promote transportation geekery.

    Feel free to post your questions about the data, I’ll do my best to answer them (or get the ever-brilliant Mike Manville to chime in).

    Katie

    http://www.wherethesidewalkstarts.blogspot.com

  • Derek

    “After all, Los Angeles quadruples New York when it comes to person per square mile.”

    But 3,990 is not quadruple 3,290.

  • The maps are a little misleading.

    For example, the New York City map seems to cover a ridiculously wide area, from central New Jersey all the way out to the tip of Long Island. This includes extremely dense Manhattan, but the edges of the map are quite rural in nature and are naturally not dense at all.

    The Los Angeles map, on the other hand, is the map that we are all familiar with. The edges are still rural, especially in the mountainous areas which are too rugged to be developed, but everything else from Redondo Beach to Pomona would appear to be as developed and spread out as the cliches have previously rumored.

    There’s no scale on the New York maps. If these maps are the same size and if they represent the same amount of area, then clearly Los Angeles deserves its crown as the king of suburban sprawl.

  • Kenny

    The vehicles-per-person maps are there, but linked to improperly. You have to delete the “2010/12/13/” from the URL, and then it works fine.

  • Thanks for the comments on the maps guys, it should be corrected now.

  • Carlton Glüb

    I think Shoup at one point did a piece about the perils of Average Density.

    Especially when you’re talking about transit – where the most important factor is density in the immediate vicinity of the transit stop – average density needs to be taken with mounds of salt.

  • Sarah

    This is awesome! I especially love the maps (and can’t wait to see how they change with the new Census data).

  • I would find all of these statistics much more interesting if they focused on central cities rather than urbanized areas, for the simple reason that in EVERY region, urbanized areas outside central cities are typically car-dependent suburbs.

  • Katie M.

    @anty: 1. The numbers are vehicles per total population, not just those of driving age. 2. It is possible to calculate vehicles per household, and I’m assuming Shoup and Manville will speak to that when they publish their full analysis. In the meantime, in LA it’s 1.67 vehicles per household. 3. Not sure about that one, perhaps someone else knows the answer? If I had to guess, I would think LA had the highest number of typically-undercounted folks.

    @James: The maps show metropolitan areas. According to the Census Bureau, “The general concept of a metropolitan area is that of a large population nucleus, together with adjacent communities having a high degree of social and economic integration with that core.” So you can see that the concept is not so much geographic as economic, which is why the scale of the maps differ so much.

  • Carlton Glüb

    @Michael,

    Yours is a really key point. San Fran always looks like it’s miles ahead of LA on bike and ped mode share, but the geography demographics of that city are totally different.

    In particular, San Francisco is physically 1/10th the size of LA.

    If you cut out the 1/10th of LA that’s just, say, Downtown and Koreatown, LA’s level of biking and transit would look much better by comparison.

    Likewise, but in reverse, if the City of San Francisco included the entire peninsula – Daly City, Atherton, M, and all the sprawling rest – it would probably look a lot more like LA.

  • The US Census bureau arbitrarily sets the minimum density for urbanized areas at 500 people per square mile. At 2.5 persons per dwelling, that translates to 3.2 acres per dwelling gross. In other words, the minimum is exurban sprawl. The main lesson from these maps is that New York has a lot more exurban sprawl than Los Angeles.

    On the other hand (and echoing Michael Lewyn’s comment), New York has a larger percentage of its population living in neighborhoods where walking, biking, and transit are truly viable and convenient as travel modes for everyday needs.

    What if we changed the minimum density of urbanized areas to a level that most would agree is actually urban — say, 4,800 people per square mile (3 dwellings per acre gross)? All of the statistics would change quite a lot.

  • Love this conversation – how can we get our transportation agencies to engage in this type of data – esp at the Regional level. It’d be great to see SCAG do some analysis like this with their transportation modeling and forecasts and pricing strategies – seems like would inform SB 375 strategies excellently – love how Mike M spells out next steps clearly – a great read.

    and lol – my burbs are denser than yours – nice.

    Shout out to SBLA – for facilitating some great conversations these days

    Last request – can we get a citywide car sharing program back then? I heard zipcar couldn’t compete cause of high car ownership – this article and the fact it’s still working in SF seems to kinda debunk that

  • Fascinating maps – thanks for sharing.

  • I agree. Fascinating maps. Thanks for sharing.

  • Drew Reed

    Awesome data, thanks for the post. Despite all the analysis suggesting that this might actually be a bad thing, I couldn’t help but be a bit proud to see LB with one of the lowest rates of cars per person. Also, I wonder what a bikes per person map would look like?

  • Stephen Hemenway

    The borders of the San Francisco area do seem a little arbitrary.

    @Carlton Glüb

    “Likewise, but in reverse, if the City of San Francisco included the entire peninsula – Daly City, Atherton, M, and all the sprawling rest – it would probably look a lot more like LA.”

    The San Francisco map, and therefore the study I believe, does include Daly City and Atherton and even a little further south down to the borders of San Jose. I suppose San Jose could also be included in the map. It be interesting to see how this affects everything.

    @ Katie M

    The map of the Bay Area does seem somewhat arbitrary in that it doesn’t include Marin, which has a HUGE degree of social and economic integration with San Francisco. Much more percentagewise than the southern end of the peninsula which borders directly on San Jose.

  • Stephen Hemenway

    Great post. This is really fascinating in a lot ways. Another map that would be very helpful to have is how many miles are driven, or how many hours on the road per car. I own a car and live in Oakland, but after 8 1/2 years my car has 30,000 miles on it, since I don’t need to drive very far for anything and I commute on Bart into SF. It’s the driving I believe that adds the most to the pollution. Suburb x could have 30 cars that drive 3 miles a day. Suburb y could have a 5 cars that drive 30 a day. Which is worse environmentally?

  • Alex

    don’t get too excited- a huge portion of Long Beach is covered by POLB, which doesn’t have many cars!

  • Matt

    Is there any way to access this data? I’ve been looking for vehicle ownership data in LA for a while, and all I’ve found from the census is the data that contains “0 cars per household” “one car per household” etc. I much prefer the “vehicles per person” and “vehicles per square mile” way of displaying the data.

  • ohnonononono

    I actually think vehicles per person is a weird metric if it includes children, which I assume it does. A single mom with 2 kids now has 0.33 cars per person in her household? That’s a bit misleading. The Census Bureau usually refers to workers over 16 in its commuting data, so I think that’d be the better denominator.

  • ohnonononono

    When the Census says “urban” they mean “not the rural hinterlands.” Urban, suburban, exurban; it’s all under the broad heading of urban in contrast to the rural, non-metropolitan, agrarian parts of the country.

ALSO ON STREETSBLOG

L.A. vs. S.F.: How Does Transportation Really Compare?

|
Last week, the Los Angeles Times published an article titled, “San Francisco residents relying less on private automobiles.” It is summarized at today’s Metro transportation headlines. The Times highlighted recent good news, reported in early February at Streetsblog SF, that 52 percent of San Francisco trips are taken by means other than a private car: […]

Some Highlights From Yesterday’s Live Ride Share Conference

|
Yesterday Transit Center, the Natural Resources Defense Council (NRDC), the Southern California Association of Governments (SCAG), Move L.A., and the Shared Use Mobility Center, joined forces with two dozen other organizations and businesses to host Live Ride Share. The conference was billed as the “first to focus on shared mobility in Southern California [and] highlighted […]

Integrating Land Use and Transportation II: The Hills

|
Yesterday, Streetsblog introduced a six-part series by Mark Vallianatos looking at how city leadership can start truly integrating land use and transportation in the six geographic zones he outlined: parks, hills, homes, boulevards, center and industry. Yesterday he outlined the series and wrote about parks. Each section includes a “preffered mobility” that the land use […]

Integrating Land Use and Transportation III: Homes Zone

|
Tuesday, Streetsblog introduced a six-part series by Mark Vallianatos looking at how city leadership can start truly integrating land use and transportation in the six geographic zones he outlined: parks, hills, homes, boulevards, center and industry. First, he outlined the series and wrote about parks. Yesterday, “The Hills” got it’s turn. Each section includes a “preffered […]

City Makes Space for Zipcars on Public Streets

|
Zipcar parked at Gramercy Park in NYC in 2006. Photo: NannyinNY/Flickr For car-sharing to every truly work, the cars used in the program need to be easily accessible to the members who have access to rent them.  To that end, the City Council has cleared out 10 parking spaces near USC and UCLA to pilot […]