The problem with “public” transportation

“The els were quite profitable and transit companies were eager to build them, but the NIMBY interests didn’t like the noise they made ”

Noise and obstructed sightlines are real externalities. Maybe they should be acceptable externalities; I’d love to see you try to make the case. But just dismissing externalities that everybody knows exists with a pejorative doesn’t help much.

Thank you.

The els may have been profitable, but profit is not the sole criterion for whether something is good or not. It was necessary to put the trains underground, and it’s something that private enterprise was not able to do profitably. Doing something necessary that the private sector can’t or won’t do is the very definition of a public good.

Of course many of the private mass transit providers in the early 20th century and before subsidized their systems through real estate development. This is an area in which I think public transportation agencies should be moving towards, although it probably would not make enough money for the agencies to significantly subsidize operations.

I would love to assume that elevated transit construction is still a viable option today. Depending on who you listen to, the city’s drive to remove the Els in Manhattan was spurred by real estate interests who called the transit a blight. It’s an interesting piece, but sadly the realities of construction a transit system for public use today is infinitely more complicated than it was 100 years ago. ADA regulations, environmental impact statements, noise mitigation and building around dense environments makes the tactics for subway and elevated construction used then very much obsolete.

For sure, there is value in moving transit underground. But saying that subways “move” transit underground, I think, is misleading – it’s a choice between a few subways or a ton of els, and I’m not quite so sure that 10 subways lines are better than 30 elevated lines.

Private transit was nice while it lasted- but unfortunately it was driven out of business by government-funded highways.

Ooh, embarrassing! Thanks for the correction. Damnit, and that was a favorite factoid of mine, too. 🙁

Like I said further upthread, it is not (and never was) a simple choice between subways and els – if the government were burying every single elevated line, then that would be a different story. But in reality, only a small number of els (/potential els) became subways.

Also, Downtown didn’t get into the specifics of why subways were unprofitable, but one reason it cited was the city’s insistence on operators maintaining a uniform 5¢ fare (something which apparently the city did away with once it realized that was totally impractical) and all the other restrictions it levied on potential franchisees. Not sure if easing up on the demands would have made subways profitable, but it’s something to think about.

You can’t forget that the IND lines of the New York Subway, including the 6th Avenue, 8th Avenue, Queens Boulevard, Culver, Fulton St, and Crosstown Lines (generally Lines A-G), were built by the city. While it’s true that there was enormous private sector involvement in subway construction at the beginning of the century, by the 30s, it was all being built by the government.

I think it’s a useful distinction in that gov’t-run public transit is a huge money pit and that requires lots of (as of yet) politically unpalatable subsidies just to maintain (forget expansion), whereas private mass transit was profitable enough to fund its own expansions (when allowed to, that is – an increasingly rare occurrence as the century wore on).

Also, I’m not sure that the gov’t simply “authorizing” a transit line detracts significantly from its private-ness. If they are granting it an exclusive monopoly and shielding it from competition, then it’s relevant, but even the franchise monopolies of the late 19th/early 20th centuries weren’t that exclusive. Also, a lot of the elevated lines actually were built on private land – some of them ran in between alleyways of private lots. And finally, a lot of the streetcars nominally ran on public roads, but many of those roads and the neighborhoods that they ran through were actually constructed and maintained by the streetcar companies.

Loop real estate is interesting. In general, it benefits hugely from the L,
but property adjacent to the L tends to be have a slightly tougher time – at
least the ground and 2nd floor space. (This tends to be a bigger problem
further outside the loop where buildings are only a few stories high.)

The 2 obvious reasons are the permanent shadow, and the noise. The noise
could be mitigated with technology that already exists, or tubes like
Koolhaas’ at IIT. I think the noise mitigation would be a better investment
than new lines…

In feasibility studies of a circle line “outer loop”, they did consider
building the L above ground, because the costs of underground are so much
greater. (though digging in Chicago doesn’t require blasting like
Manhattan) I assume those plans are on hold, but haven’t kept up with
them…

Could it be that the causality between poor neighborhoods and els actually runs the other way? That poverty caused the neighborhoods to not get subways, not the other way around?

Does this mean that you think that public roads are a public good? Why do you think transit
is not a public good?

Second Avenue *did* have rapid transit a century ago!

But they tore down the el.

Promising to replace it with a subway.

No, private transit was driven out of business by the enormous superiority of private automobiles. This happened not just in the U.S. but throughout the developed world. Highways are funded mostly from user fees. Transit use today is subsidized at a vastly higher rate than automobile use.

Haha, you must be new here.

I’ll try to make this brief: Highways are now subsidized (mostly!) out of user fees, but this was not always the case. Furthermore, eminent domain was used very liberally to create the highways. And finally, local roads – the ones that actually have the ability to take down private mass transit – have never been funded out of user fees, and to this day are still not.

A quibble? Why are subsidies for local roads a quibble, but subsidies for mass transit are relevant to the conversation? Do you have some data on local road spending that you’re not sharing with the class?

“Why are subsidies for local roads a quibble”

Because they are tiny in comparison to transit subsidies. If you think you have data showing otherwise, please present it. In fact, how much are you claiming is provided in subsidies for local roads per year, and where are you getting your numbers from?

The histories I’ve read of the subway have it different. The city’s patrician elite planned for rapid transit from the 1850s. It permitted the els as a makeshift solution, but preferred subways because they’d be much faster. In addition, at the time, in the 1860s, there was no electric train traction yet, so els were run on dirty steam power, whereas subways could be pneumatically powered. By the time electric traction was feasible, in the 1890s, the limited speed of the els was already a serious problem. The intention behind rapid transit was to suburbanize immigrants and the working class, sending them to single-family detached houses with separation of uses in order to turn them into proper middle-class Americans; speed was crucial, to maintain adequate range (that’s why it took until the 1920s to build crosstown subways in Manhattan).

After 1900, municipal control was more supported in New York. The politics of subway construction revolved around the city’s immense corruption; initially the Rapid Transit Commission and Parsons got the support of the reformers, since they viewed them as independent of the political machine, but afterward the reformers grew weary of the IRT and considered it equally corrupt. But to combat that corruption, the city introduced competition rather than municipalization, letting the BRT run some transit lines in Manhattan, and threatening the IRT that if it objected, it would give the BRT even IRT extensions like the Lexington Line.

Congrats, Stephen, you’ve made it big – you’ve got Mixner Gordy garyg DillonS Mixner commenting on your blog.

Jim,
I’d be glad to discuss this topic further, as one of the main objectives of this site is to dispel the myths that are commonly held about the current government transportation system. Perhaps a better forum is in the comment section of the article which I think handily destroys the pervasive myths about “roads paying for themselves.” etc. http://www.marketurbanism.com/2008/07/30/urbanism-legend-gas-taxes-covers-all-costs-of-road-use/
Perhaps you could read the article, poke some holes in it, provide some analysis that counters what I have to say, and we can have a healthy debate.
Adam

The reality is that transit users are subsidized at a vastly higher rate than road users. If transit were priced at anything like the market rate, transit users would have to pay enormously higher fares than they do now. That would dramatically reduce transit demand and make dense, urban, transit-oriented lifestyles even less attractive than they are now. The transportation system in the United States has become overwhelmingly dominated by cars DESPITE government subsidies, not because of them. The same is true in virtually every other industrialized democracy.

If you seriously think subsidies favor using automobiles over using transit, I’d like to see your evidence and analysis. The difference isn’t even close. Direct subsidies alone pay for about 70 cents of every dollar in transit spending. As Mark Delucchi put it: “the total subsidy to transit greatly exceeds the total subsidy to auto use, per passenger mile, in both absolute terms and relative to the prices users currently pay.”

I have no idea how you think cars could have become dominant because of subsidies rather than in spite of them given that subsidies so overwhelmingly favor using transit over using autos. Autos are the dominant mode of transportation even in Europe, where gas taxes are enormously higher than they are in the U.S.

300 posts, but no evidence or analysis to support the proposition in question. Indeed, there is very little about transit subsidies at all, let alone an attempt to demonstrate that auto subsidies are greater.

300 posts, but no evidence or analysis to support the proposition in question.

You don’t get to just change the proposition and pretend we’re dodging it. Nowhere in this blog does Stephen or I propose or claim know exactly to what extent greater subsidies to one mode causes one mode to be more prevelent.

The world isn’t a zero-sum game. Just because kumkwats are subsidized 90% and apples are subsidized 20%, doesn’t mean the subsidies to apples are ok.

My point isn’t that we should ignore subsidies to transit – I said it’s bad. Randal and others already beat that dead horse. My point is, and always been that transportation, in general has been heavily subsidized since the progressive era. That includes cars. You can pretend to deny it, but I think we’ve covered that topic better than any site on the web.

I don’t know to what extent each cause of the dominance of the automobile is. But I know one thing, roads in America have been grossly subsidized for the past century, and economics tells us that if you subsidize something, you get more of it. So, you have more auto travel than you would otherwise, and more transit travel. The end result is seen in the land use pattern, and cant be definitely quantified in the dominance of uses.

If you have countr-evidence that roads aren’t subsidized, present it. Otherwise, there is no burden of proof on me to prove your unprovable opinions and speculations.

OK.  Yes, the huge subsidies to transit causes sprawl too.  

Some day, I’ll address that in a fundamentals of urban economics post.  In the meantime, you can read plenty about transit subsidies at Randal’s blog.

http://marketurbanism.com

I didn’t say transit subsidies promote sprawl. I said they promote transit-oriented patterns of travel and urban development. I’m not sure why you think that means sprawl. Transit-oriented urban development generally means denser, more compact urban forms, where trip distances are shorter and car travel is more difficult and expensive. The examples most often cited by proponents of TOD are developments near rail transit in places like Arlington, Virginia and Portland, Oregon. Without transit’s massive subsidies, such places would be even less common than they are now.

Technological triumphalism? Also, can you imagine the nice 3-track stacks and flyovers that would be necessary wherever lines crossed? Imagine the interchange of Times Square above ground. It exists in East Brooklyn, and it’s not a pretty site.

Obviously, the original construction of highways and Interstates, indeed the original roads, were constructed at great cost. Such great cost that we could probably not reconstruct them today with user fees. The original roads were constructed before there were users, and the gas tax pays for some maintenance at the very most. You can’t argue with the fact that there is a huge road network and a really really crappy piddly transit system in this country. Obviously, obviously the greater incentive is towards driving. We don’t even have legitimate sidewalks in many communities, but there is always pavement for vehicles.

What do you think would happen to the market for autos, and auto-oriented development, if private businesses were no longer required to subsidize parking, and if cities raised street parking to market rates?

The premise of the first part of your question is false. Private businesses are not required to subsidize parking. They are free to charge their customers for it.

As for street parking, the market rate for parking in most places at most times is zero, or close to zero. Unpriced street parking can only have a significant effect on demand in dense urban areas where the market rate for parking is significantly higher than zero. Street parking in such areas is already often priced via parking meters. The effect of raising the price of street parking in these areas to market rates would depend on the difference, if any, between current rates and market rates. In areas where the difference is large, it might significantly reduce car use in that area. In areas where the difference is small, it would probably have little effect. With respect to car use in total, the effect would probably be negligible. Indeed, raising the price of street parking in dense urban areas might actually increase the amount of auto-oriented development by shifting demand from those areas to the suburbs, where land is cheaper and parking is already mostly free. People who are willing to pay, say, $5 an hour to park downtown when they go shopping may not be willing to pay $10. Instead of driving downtown to shop, they’ll drive to the suburban mall instead.

As I have already pointed out umpteen times…

Yes, you definitely have pointed it out many times. Have you ever considered the possibility that people are reading your statement and simply disagreeing that it’s the most relevant fact?

By the way, here’s an example of a non-subsidy form of government intervention that also massively favors transit: immigration policy. For a number of reasons — income, culture, the location of immigrant communities — immigrants use transit at a much higher rate than the native-born population. In California, immigrants comprise about a quarter of the population, but more than half of all transit commuters. In recent decades, the foreign-born share of the population has increased significantly as a result of the liberalization of immigration law. This has helped to sustain transit demand in the face of the continuing shift to automobiles among the native-born population. I’m not saying that sustaining transit was a goal of changes to immigration policy, but it has been one of the effects.

Most evidence suggests that the immigration rate is likely to slow in the future. And existing immigrants become less likely to use transit over time, as they assimilate and increasingly adopt the lifestyles of the native-born population. This does not bode well for transit, or transit-oriented urbanism.

Your argument that most transit expansion was built during a time when the subways were owned by private companies – so private companies are the best constructors of transit is a non sequitor. In many other places of the world (and in the US) this phase of rapid expansion happened while the system was owned by the public – citing one example where rapid expansion happened during a time when it was owned privately does not make a very strong argument. The pro/con el debate is largely unrelated to the issue of who owns the system (although one could make the argument that the public tended towards the more longterm solutions – in this case the subways).

The corollary of your thesis is that privatization of the system will suddenly make it profitable is advantageous at best. Privatization is not a magic solution that will somehow bring you from 40% to 100%+ fare recovery ratio. If it is able to do it, than mostly by shutting the whole system down except the few profitable lines — which means many people will be cut off from service (and network effects from shutdowns will make the remaining lines unprofitable as well). And that is where transit is a public service — you want to provide the service to everybody as best as possible, instead of cutting off whole neighborhoods because of a lack of profitability. A public system should accept _some_ unprofitable service in order to reach more people.

A possibly better solution, rather than flat-out privatization, is to privatize operation of the service. One could requests bids to operate parts of the system; and whoever can offer the requested service (schedule+capacity) at the best value, will be able to get the contract. This will still involve subsidies from the one ordering the service (the municipality), but will mean that the infrastructure ownership stays in the hands of the public – to assure the continued existence of the service for one; and keep the control of the system and its expansion with the public. The trick is to leverage the market incentives without giving away control of essential service to the moody market gods.

Sorry, but your links do not address the issue. All you’ve provided is a list of posts, some of which argue that some kinds of government intervention favor autos rather than transit (other posts in your list simply don’t have anything to do with that question at all).

Again I ask: where is your evidence and analysis suggesting that other forms of government intervention favor autos to such a degree as to more than offset transit’s huge subsidy advantage? Where is your evidence and analysis suggesting that other forms of government intervention favor autos at all? And by “other forms of government intervention”, I mean other forms of government intervention in total, not just isolated examples.

All subsidies for highways are by definition subsidies for driving, there’s nothing to confuse there. You are right that the cost to operate and maintain a vehicle is a major expense for the user, something on the order of $5,000 per year on average I believe, or roughly $0.50 per mile.

Still, even if all user fees were used only for highways (which I did not ignore by the way, I clearly spelled it out), that only covers 67% of the total expenditures on highways. The government is spending $182 billion on highways and only taking in $122 billion in user fees per year. The remaining $60 billion is a subsidy, which with 254.4 million passenger vehicles in the US works out to about $235 per vehicle per year in subsidies which doesn’t sound like a whole lot compared to $5,000.

That $235 per vehicle only applies to highways, as I mentioned before. The local road network isn’t a part of that at all, and that makes up the vast majority of the pavement in the country. Even including that number (which is a pretty elusive one unfortunately) won’t bring it up to the $5,000 level of course, but it’s not insignificant in the equation. Also, that $60 billion in subsidies to highways is less than ALL spending on transit.

I’ll say again, there’s nothing to confuse. Regardless of what the rate or actual amount of the subsidies may be, they’re still subsidies. A highway subsidy is a de facto driving subsidy because highways are almost exclusively for cars. If you want to poo poo that for being a small part of the cost of driving, that’s fine, but it’s still a subsidy and it still encourages driving. A big reason for that is because those costs are hidden. Drivers don’t have a little counter on their dashboard that shows how much their trip is actually costing them, so the connection isn’t made in their brains. It’s all perceived as free, so it’s overused, but it’s costing us real money. $60 billion is a lot of money, and breaking it out into a per-mile cost is very disingenuous when comparing driving to transit because transit trips are almost always shorter. It should really be looked at on a per-trip basis, or at least indexed to account for both the per-mile value AND the number of miles traveled.

Your comment about “alleged local road subsidies” is very disturbing. There’s nothing alleged about it. We have just a hair over 4 million miles of roads in this country, and about 8.5 million lane miles (hence most roads are just two lanes). Only about 23% of those are State and Federal routes that get money from those sources (even interstate highways are owned by their respective states, so very little is actually owned by the Federal Government). County roads make up 44% and municipal roads (township, village, city, etc.) make up 32% with a little left over for weird things like national park roads and such. That 75% of the road network gets no gas tax revenue and must be made up for by the general population.

Local roads are funded almost entirely through city/township/county property taxes, sales taxes, and local income taxes. Those numbers are huge, something like half of all road expenditures, but they rarely make it into highway statistics because, well, they’re not highways! Yes those roads are needed whether cars use them or not, but that pavement and those storm sewers aren’t free. Roads were widened to accommodate more cars, they were repaved with asphalt instead of brick or dirt, curbs and gutters were added in many places to handle the extra runoff etc. Even if people pay a lot of their own money on their own cars to use these roads, the roads are still incredibly expensive.

That’s the real issue here. Even if the cost of driving is mostly born by the driver through the operation of their car, we as a society are still spending enormous amounts of money supporting the road network, and we can’t afford it anymore. The owner of that huge McMansion out in the suburbs may be bearing the majority of the cost of owning that house because he’s the one paying the mortgage, but we’re all paying for the extension of the water and sewer lines and the roads to serve it. So many of the indirect costs, the unfunded liabilities, the opportunity cost of using the money or the land for roads instead of other things, the externalities of pollution and sprawl, the war in Iraq, military escorts for tanker ships, these are all biting us in the ass now. We’ve put together a system that requires people to spend a huge amount of their earnings on transportation exactly because they must provide the “rolling stock” themselves. There’s a huge opportunity cost in that as well. That money could be going to a bigger house, or the kid’s college fund, or any number of other goods and services, ones that actually appreciate in value rather than depreciate.

Conflate state-owned highways with roads, anyone?

I shall now attempt to demonstrate that auto subsidies are significantly greater than generally realized. Ready? Here it comes: “Donald Shoup.” Bam.