Arnold Kling  

Mass Transit and Happiness

PRINT
European Productivity... Neuroeconomics...

Peter Gordon reports on declining use of mass transit.


As a group, the 20 largest U.S. metro areas declined in transit use (all trip purposes; thank you, Wendell Cox) in the 1990s. Not relative decline but absolute decline.

As a group the areas with new (post-WWII) metros (San Francisco, Washington, DC, Atlanta and Miami) lost even more transit users than the group of 20.


Yet, according to Robert Frank, we would all be happier if we rode mass transit.

For Discussion. What factors account for the decline in mass transit use? Is it due to flaws in public policy or to consumer preference?


Comments and Sharing





COMMENTS (11 to date)
Lawrance George Lux writes:

It is regretably Consumer preference. Time spent in their own private vehicles remains the only alone time most have anymore. The cellphone is trying to eliminate that time. People need space, which they can get nowhere, except that they drive away. lgl

Brad Hutchings writes:

Egads lgl!! You sure do give it the good fight here, but that explanation is laughable.

This is just a graph theory problem. For mass transit to work, it has to pretty much saturate the nodes of the graph (starting points and destinations) with paths that are fairly direct. This turns out to be a fairly easy task in densely populated areas and totally impossible in sprawl. Even regional mass transit, such as BART in the SF Bay Area, suffers from the problem of if your endpoint isn't near the BART station, it's probably more convenient to drive and negotiate/find partking.

You might imagine this problem as the Internet with a few modifications. Let's say that for you to browse an arbitrary web site, your request had to wait for a periodic critical mass of requests along one leg of the route of your destination (i.e. a bus that is filled to capacity with passengers). Except there are lots of legs on this route, so you wait and wait and wait. The Internet usually does not work that way, and except in very particular applications (e.g. low speed, high latency endpoints like dialup or satellite), we don't put up with such behavior.

You could all learn a lot from computer scientists. We have had this stuff figured out and modeled for years ;-). For moving people around, we are much better off providing capacity (traffic flow, parking, etc.) the way people choose to use it than shoehorning them into public spending boondoggles. Think about how the Internet works. That is how you make people happy.

Allan Schoenberg writes:

I think Brad hits it on the head from a scientific and network standpoint, but from a social viewpoint it comes down to control. I live in Chicago and take the CTA, but the lack of control and inconsistency drives me to, well, drive. I have no control when the trains are delayed or when I can't get a seat to read the paper. When I drive, even though it may not take much less time, I feel that I'm in control. The cost savings is an added factor, but if I have to be in the office for a morning meeting I'm not relying on the CTA.
Allan

I agree with Brad that mass transit is a network. Thus, its value increases rapidly as more endpoints are added.

Relative failure of mass transit may be a result of success of freeways and road network. Because, roads and freeways are well-connected, it is much easier to drive personal car than to take rail-based mass transit.

bob writes:

So, why do San Francisco, Atlanta and other cities continue to build mass transit boondoggles eventhough experience and science show it is a waste of resources? Los Angeles and San Jose are two recent projects that come to mind. San Francisco is extending BART to the airport.

Perhaps I'm hoping that everyone else will take mass transit and leave more bandwidth on the highway for me.

Are there situations where the science says mass transit is a good idea? How about, when there is a high volume destination and a high expense (long term parking fees) involved.

Brad Hutchings writes:

Bob,

If you can model the costs right, you can just use standard graph (network) algorithms to find things like shortest (least costly) path, etc. Assuming that people are rational actors and sufficient capacity exists, you can expect them (over time and en mass) to settle on the most efficient routes.

When I was growing up in the SF Bay Area (Danville) and my Dad was working in San Francisco, it was really interesting to watch his commute choices. He could have walked to the end of the street and taken a bus to the BART station in Walnut Creek then BART into the city, then walked two blocks to work. Bear in mind this was before cell phones and laptops, so he couldn't really do much work during his commute if he wasn't driving. When he had his small-ish motorcycle, he usually rode to the Oakland West station. When he had his Honda Gold Wing (which was stable on the Bay Bridge), he found a guy in the city who let him park in the corner of his garage. When he had a car, he usually drove to the Orinda BART station, bypassing all the other stations east of the tunnel. I guess you'd have to know the freeways and BART lines up there for it to make any sense, but when his long term incentive was to find the most efficient commute so he could spend more time at home, he was pretty good at doing that.

I suspect that all of us who have to go from point A to point B a lot (commute, favorite place, etc.) are a lot better at maximizing efficiency than central planners devising routes and schedules. The reason mass transit can work effectively in NYC, for example, is that it saturates, so there is little incentive to search out a better route for marginal gains.

-Brad

Tom writes:

What factors account for the decline in mass transit use? Is it due to flaws in public policy or to consumer preference? The obvious answer is "both". The flaw in public policy is that planners and engineers think they can outguess the market. They can't, of course, because consumer preference will, in the end, dictate how people commute. Consumer preference changes with time, for reasons that planners and engineers are unlikely to predict accurately; e.g., changes in the number and geographic distribution of commuters caused by population growth and changes in relative real estate prices and municipal tax rates; changes in the geographic distribution and mix of jobs caused by population shifts and the ever-evolving mix of businesses (owing to national and global as well as local influences); generally rising affluence, which militates toward driving (control and prestige) rather than taking mass transit (being one of the sweaty herd); and, most importantly, efficiency (from the consumer's perspective). As Brad Hutchings says: "For mass transit to work, it has to pretty much saturate the nodes of the graph (starting points and destinations) with paths that are fairly direct. This turns out to be a fairly easy task in densely populated areas and totally impossible in sprawl. Even regional mass transit, such as BART in the SF Bay Area, suffers from the problem of if your endpoint isn't near the BART station, it's probably more convenient to drive and negotiate/find partking." Some proportion of consumers will find mass transit efficient in densely populated areas. An increasingly smaller proportion of consumers will find mass transit efficient in outlying areas, that is, in urban sprawl. And sprawl is what we have.

Lawrance George Lux writes:

You all insist the problem is finding the quickest commute, but I have lived in the less densely populated States most of my life. People build in the Country, though it quadruples their commute time. No one carpools! One-third drive their kids to school, ignoring the school bus system. My current location leaves the major shopping areas either 30 or 70 miles away. Most of my neighbors go shopping three or four times a week. No one carpools. Your statistical graphing means nothing. The American is addicted to getting out on the open road in his own personal vehicle. lgl

Mike writes:

It's disappointing to see so many comments, and yet none have pointed out that almost none of the total cost of an individual's choice to drive on a given trip is paid for at that time, directly, by that individual.

We get very upset over low farebox ratios (20% or so) for mass transit - but when compared to the fact that most spending, even that directly on maintenance and construction, has no link at all to the decision to drive that day... it's fairly obvious why driving increases. It's a rational response to try to get the most out of the dollars you have to spend anyways (the days I take the bus or ride my bike, I don't get a rebate on the property or sales taxes spent on roads!)

Kari writes:
What factors account for the decline in mass transit use? Is it due to flaws in public policy or to consumer preference? The obvious answer is "both".

Agreed. It's not just about networks and nodes, or even time. In New York and other cities with successful mass transportation, the cost of driving yourself around (time spent in traffic jams, paying for parking, owning a car in the first place, etc.) exceeds the cost of riding the subway or bus (limits on destinations, loss of privacy and personal space, personal safety, etc.). In most cities, that is not the case.

I lived in and around DC and its successful Metro for four years, two of those years within 20 yards of an express bus stop that went directly to a Metro station, and two years within two miles of a Park-and-Ride Metro station. While the Metro was clean, safe, and fairly time-efficient, I took it to work only in bad weather, or when some mass protest or another was reported to be snarling traffic. I drove because (a) I liked the time alone to think and relax before a stressful work day, (b) I had more control over my movement, and (c) I had a parking space waiting for me. However, I often took the Metro to get to entertainment districts, mostly because (c) did not apply.

...but when compared to the fact that most spending, even that directly on maintenance and construction, has no link at all to the decision to drive that day... it's fairly obvious why driving increases.

Hmmm. At least where I'm from, about half of the road budget comes from state gasoline taxes, and another half from the federal goverment's gas-tax-funded highway trust fund, so there is a pretty direct link. Now, whether drivers think of that every day they choose to drive is another matter, but the fact is that consumers do pay to drive on public roads.

Kari writes:

Err, that's "federal government." In the future, I'll stick to typing it as "gubmit."

Comments for this entry have been closed
Return to top