Actually, Highway Builders, Roads Don’t Pay For Themselves

Cumulative Net Difference Between Spending on Highways and Highway “User Revenues”
Since 1947, American highways have run up a deficit bigger than $600 billion, in 2005 dollars. Source: ##http://www.uspirg.org/home/reports/report-archives/transportation/transportation2/do-roads-pay-for-themselves-setting-the-record-straight-on-transportation-funding##U.S. PIRG##

You’ve heard it a thousand times from the highway lobby: Roads pay for themselves through “user fees” — a.k.a. gas taxes and tolls — whereas transit is a drain on the taxpayer. They use this argument to push for new roads, instead of transit, as fiscally prudent investments.

The myth of the self-financed road meets its match today in the form of a new report from the U.S. Public Interest Research Group: “Do Roads Pay For Themselves?” The answer is a resounding “no.” All told, the authors calculate that road construction has sucked $600 billion out of America’s public purse since the dawn of the interstate system.

The Myth of the User Fee

First, let’s dispense with the idea that the gas tax – the primary source of financing for federal transportation projects – is a user fee.

“If you go to a state park and pay the fee to get in there, that’s a user fee,” report author Dan Smith, U.S. PIRG’s transportation associate, told Streetsblog. “If you’re driving down the road and you have to pay the toll for driving on that specific road, that’s a user fee.”

But people also pay gas taxes to fill up their lawnmowers. And those lawnmowers don’t usually end up on the highway. Just because you fill your tank doesn’t mean you ever drive on the roads funded by the gas tax you pay.

The Catch-22

Then there’s the huge contradiction underpinning the core arguments for highway expansion. Do new roads cut congestion, or do they “pay for themselves”? Highway lobbyists try to have it both ways, but the truth is that neither of these propositions hold water.

Highway expansions are often justified as projects that relieve traffic and, believe it or not, reduce pollution. So if a highway widening achieved its stated aims, it would cut congestion and fuel consumption, which would mean fewer gas tax dollars and roads that don’t pay for even a fraction of their construction costs. However, we know that new highway capacity doesn’t actually reduce driving – it induces more driving.

The additional traffic created by expanding highways does generate more gas tax revenue, but still not enough to come close to covering the costs of new roads.

U.S. PIRG cites the Pew Charitable Trusts’ SubsidyScope project, which found that “user fees paid for only 51 percent of highway costs, down 10 percent over the course of a single decade.”

Even if gas taxes were the direct user payment they’re made out to be, no one seems to have much appetite for making sure they actually pay for the infrastructure needs in this country. Gas taxes haven’t risen to accommodate more fuel efficient cars or even for plain old inflation. Nor have they compensated for the fact that driving is declining, meaning less gas consumption (but, puzzlingly, not less road-building).

The federal gas tax hasn’t gone up since 1993.

The Highway Funding System as a Subsidy for Driving

The argument that drivers pay for roads might be somewhat more credible if they weren’t taking money away from other public funding streams. Gasoline is exempt from sales taxes in 37 states and the District of Columbia. So rather than paying into the general revenues for the state, motorists are paying into an already narrowly prescribed pot of funding, which highway advocates want to see prescribed even more narrowly to exclude transit and bike/pedestrian projects.

In New Jersey, the savings on the sales tax exceeds the gas tax drivers have to pay. In that way, the government actually provides a financial incentive to purchase gas and drive. And since gas taxes are fixed and sales taxes are percentages of the purchase price, more and more states could end up with this perverse subsidy as gas prices rise.

Image: ##http://www.soundecoadventure.com/AnchWhit/TollBooths2.html##Sound Eco Adventures##
Photo: ##http://www.soundecoadventure.com/AnchWhit/TollBooths2.html##Sound Eco Adventures##

What About Tolls?

Tolls are, indeed, an honest-to-goodness user fee, charging drivers directly for the road they’re driving on. But the overwhelming majority of roads are not funded by tolls. Local streets don’t have tolls. Rural highways don’t often have them. And tolls don’t come close to covering the costs of roads. According to U.S. PIRG, “In the 1950s, experts estimated that no more than 9,000 miles of highway (compared with the more than 3 million miles of highway in existence at that time) could support themselves with tolls.”

Founding Fathers

The report goes into ancient history (the Hoover administration), investigating the original intent of the gas tax at both the state and federal levels, and debunking the myth that they were always intended to pay only for highways. Indeed, federal gas taxes originated in the 1930s and were dedicated exclusively for highways only for a 17-year period, starting in 1956, covering the construction of the interstate highway system. Since 1973, the gas tax has been used for a variety of transportation programs and has even been used, on occasion, to pay down the deficit.

External Costs

And now the obvious: You can’t measure all the costs of driving with the price of asphalt. The U.S. PIRG report gives a laundry list of external costs associated with driving, including:

Photo:
Photo:##http://farm3.static.flickr.com/2581/3868301165_fe39dd4bf5.jpg##ret0dd/Flickr##
  • Changes in the risk of accidents, including injuries to non-drivers and damages to property.
  • Environmental and public health impacts, including smog, greenhouse gases, water pollution from highway runoff, and the impacts on wildlife and outdoor enthusiasts.
  • National security and economic implications of protecting access to foreign oil.
  • Increased pressure on those without cars.
  • Quality of life and the impact of roads on active transportation, such as walking and biking.
  • Car-centric development patterns, sprawl, and the resulting infrastructure costs for the expansion of water, sewer, and other services.

The report cites one study that finds that, just to pay for roads, user fees need to be 20 to 70 cents higher, and another study that finds that, to pay for external costs like these, we’d have to add another $2.10 a gallon.

The Cost of the Myths

“Road advocates use these myths about the gas tax being this user fee and that highways pay for themselves to get preferential treatment, and to get a larger chunk of the dedicated fund,” says Smith of U.S. PIRG. “Advocates of any type of policy would like a dedicated fund, because it is a stable source of funding.”

The myths associated with road financing put all other forms of transportation at a disadvantage, said Smith. “Conservatives say all other transit is social policy and should come from the general fund.”

With a Republican majority in the House, the myth that roads pay for themselves will again be enlisted to prioritize highways over transit, as the GOP begins shaping a transportation agenda around “getting back to basics” and cutting spending, especially for transit.

“We want to make sure that those falsehoods are not a part of this debate,” said Smith. “People will think twice before saying roads pay for themselves when the numbers say they don’t.”

  • Dj Z

    If the subsidy for cars is so small, why not just raise the gas tax to cover the gap?

  • garyg

    If the subsidy for cars is so small, why not just raise the gas tax to cover the gap?

    Since the subsidy for cars is so small, why are people making such a big deal of it, instead of focusing on the much larger subsidy for transit? Tell you what: I’ll agree to support the elimination of subsidies for cars if you’ll agree to support the elimination of subsidies for transit.

  • Tjon

    Another interesting take: while highways are comprised of only a right of way, transit generally includes a right of way _and_ vehicles (and their associated maintenance costs). Shouldn’t the highway subsidy be adjusted to reflect various automobile subsidies (tax benefits, cash-for-clunkers, and possibly auto company bailouts) as well? It seems to me that would be a fairer comparison.

  • garyg

    Shouldn’t the highway subsidy be adjusted to reflect various automobile subsidies (tax benefits, cash-for-clunkers, and possibly auto company bailouts) as well? It seems to me that would be a fairer comparison.

    Cash for Clunkers was $3 billion. That’s about six one-hundredths of one cent per passenger-mile, for one year. If the auto bailout was a subsidy to anyone, it was a subsidy to the shareholders and employees of the automakers, not to drivers. But it’s not really clear that the bailout was a subsidy of anyone, since they are supposedly loans that the automakers are repaying over time. And I have no idea what “tax benefits” you’re referring to.

    You’re really grasping at straws here. How much did mass transit get in TARP funds?

  • Re: Absolute versus relative.

    Disclaimer: This part of the post is not a post about the relative merits of the arguments; see my personal take at the bottom.

    garyg said, “The tiny government subsidies to car users have almost no effect on the market, because they are so small both in absolute terms and as a share of user costs”

    Delucchi said, “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.”

    Delucchi used “absolute” correctly, as it was stated that the metric is “absolute per passenger mile” versus “relative per passenger mile”. However, failing to say, “per passenger mile”, means that the “absolute cost” is in relation only to itself. If you want “absolute cost per passenger mile” you should say so.

    My personal take on this thread: This argument seems to be between garyg saying, “On a per-mile basis, subsidies of mass transit are greater than subsidies of automobiles.”, and other people saying, “But automobiles get massive subsidies in lots of places and lots of ways.”. Then, beyond those points, other people are effectively saying, “But the mass transit subsidies are worth it in a way that automobile subsidies aren’t.”

    Garyg appears to be frustrated that people don’t get that he’s making this solid direct-cost argument that appears to hold up regardless of how many reasonably-direct costs you include. However, the Iraq war (among other tangentially-related things) is not a direct cost, and thus not included in his argument.

    Others appear to be frustrated that garyg doesn’t seem to see the sheer devastation that cars are wreaking upon people and the environment, while requiring massive sums of money to do so. However, garyg appears to be concerned only with a direct economic argument, and thus death and destruction from vehicle usage and future flooding are irrelevant.

    So, can we now all agree with these things:
    1) Cars and mass transit are massively subsidized. (By “massive” I mean at least two billion dollars a year in the US.)
    2) The total cost of car subsidies are much greater than the total cost of mass transit subsidies.
    3) It appears that, on a per-mile, direct-subsidy basis, cars are currently cheaper than mass transit.
    4) Wanting to be able to get places without a car involved is a value that garyg, most Republicans, and many others don’t rate too highly. However, most everyone else in this thread, most Democrats, and many others do rate it highly.
    5) When garyg said, “TARP funds”, he meant, “stimulus funds”.

    Now, unless someone has actual numbers to refute point 3, could we agree to move on to just arguing over who’s right in point 4, since that one is mostly a personal value judgment, anyway?

  • Darren

    Thank you, Clay.

  • Tom

    If cars were the only available mode of transportation, cities as we know them would not be possible. For example, let’s consider NYC. Can you imagine how many highway lanes would be necessary to move the number of people that the subway can move in an hour? Likewise, can you imagine the sheer area that 8 million people would consume if every family needed to own at least one automobile? The economic efficiencies (ability to share goods, services and ideas easily) and energetic efficiencies (per person energy expenditure in a city is incredibly lower than in rural and suburban areas) made possible by cities are invaluable to our future. This is not simply economic analysis of consumers’ cravings. This about what will make prosperity possible when the end of cheap oil arrives, and the answer is not cars, roads, and the sprawl that they make possible.

  • Tom

    And wait just a second here. What about the Interstate Highway Act? How’s that for government interference? If it’s now clear that automobiles, in part, are degrading the quality of our cities, health, and environment, why must we put our foot down at a similar large-scale investment in transit?

  • hurotsin

    I think it’s pretty obvious that all modern transit requires subsidization at this point. If you took away the subsidies on both auto travel and mass transit, no one would be able to afford fuel to power their cars, and no one would be able to afford train or bus fares. That leaves us all riding donkeys around (assuming we can afford to feed them). And the fact is that if you took away either mode you would instantly eliminate the quality of life of a certain percentage of the population, which would have negative societal consequences which are not easily quantifiable, on top of the economic impact.

  • Al

    1. Good summary, Clay.

    2.
    “Even if ALL “free” parking were counted as a market-distorting subsidy, the value would be on the order of 5 cents per passenger-mile, which is trivial in comparison to the enormous subsidies provided to transit. But in reality the market price of most “free” parking is zero anyway, because the costs to businesses of pricing it separately would outweigh the benefits. That’s why they choose to bundle it instead, just as they bundle all sorts of other products and services they provide to their customers.”

    The costs of providing parking is certainly not zero. The cost appears to be zero only because of the government-enforced oversupply (what would the market price of an apple be if merchants were required to provide one with every purchase?).

    I was visiting LA and reading the local paper, and I read a funny article. There was a local scandal. Businesses were required to provide sufficient parking for their (expected) customers. Many of them fulfilled this requirement by paying valet companies to guarantee the, say, 25 spaces that the restaurant was required to have. The scandal emerged that the valet company was guaranteeing more parking spaces than it had bought/rented land for, and so the local residents were demanding that the businesses shut down until they buy the required parking. This was front page news.

    But yeah, sure, the market rate for parking is zero.

  • Thanks so much for highlighting the PIRG report and this important issue. I’ve shared it with my No Car Go readers:

    http://nocargo.wordpress.com/2011/01/20/highways-arent-cheap-dates/

  • garyg

    Clay,

    Academic studies of the external costs of transportation include estimates of the costs of the Iraq War, “environmental devastation,” “death and destruction,” and other indirect costs of transportation. You seem to think these external costs apply only to cars, but that is incorrect. They apply to mass transit also.

    The combined total of all these external costs for cars is vastly smaller, per passenger mile of travel, than the direct public subsidies provided to mass transit. Adding up all external costs and direct subsidies for each mode, total subsidies to cars are maybe 10% of the total costs of cars. Total subsidies to mass transit are 70-80% of the total costs of mass transit. There is simply no way you can escape the conclusion that subsidies massively favor transit over cars.

  • garyg

    The costs of providing parking is certainly not zero.

    I didn’t say the costs of providing parking is zero. I said the market price of most “free” parking is zero because most parking providers bundle the costs of parking into the prices of other goods and services they provide. Thus, you pay a tiny bit more for the underpants you buy at Walmart than you would if Walmart did not provide parking.

  • Garyg: I’m glad the rest of my post was not controversial. Arguments of people talking past each other tend to annoy me.

    Now, just a couple of points on your reply to me:

    “You seem to think these external costs apply only to cars, but that is incorrect. They apply to mass transit also.”

    Sure. They just apply less. E.g., if everyone rode mass transit, fewer people would die than if everyone drove by themselves in a car.

    “There is simply no way you can escape the conclusion that subsidies massively favor transit over cars.”

    I could just look at the total cost of subsidies there are to cars, compare it to the total cost of subsidies there are to mass transit, and escape that conclusion.

    I could assume that, without the demand involved in gassing up millions of cars, there would have been no war in Iraq, and the military budget would have been massively smaller.

    Even without assuming those things, I could work off of a “per trip” metric, while also counting walking and bicycling trips of people who depend on mass transit.

    In other words, in order for me to accept your claim that there’s no other conclusion possible, we’d have to work off the assumption that nearly all the claims of the radical environmentalists are true, and then see what the subsidy is.

    So, please, unless you’re willing to argue assuming pretty much all leftist contentions (i.e., no “but this one shouldn’t apply” unless nearly all the liberals agree it shouldn’t apply.), please accept that I agree that the per-passenger-mile-subsidy metric appears to vastly favor cars. The numbers being what they appear to be, those in favor of mass transit would be crazy to argue for mass transit on that basis.

  • garyg

    I could just look at the total cost of subsidies there are to cars, compare it to the total cost of subsidies there are to mass transit, and escape that conclusion.

    No you couldn’t. Incentives depend on unit subsidies, not total subsidies. If the unsubsidized cost of apples and oranges is $1 each, and the government subsidizes 10 cents per apple and 80 cents per orange, the subsidies would massively favor oranges over apples, regardless of the total subsidy for each type of fruit, because the government is paying a much larger share of the costs of oranges.

    The numbers being what they appear to be, those in favor of mass transit would be crazy to argue for mass transit on that basis.

    But they do argue that. I’m glad you agree it’s crazy.

  • “No you couldn’t. Incentives depend on unit subsidies, not total subsidies.”

    Sure I could. Take it to the logical extreme — Say there’s one mass transit subsidy, period. Let’s say $200 per passenger mile for a monorail group-bicycle that travels half a mile.

    So, say this example costs a couple million. So, a couple million for monorail bicycle, and billions for car transit. Car transit is massively favored, because there’s just not that much investment in monorail bicycle. Sure, at that particular half-mile stretch the monorail bicycle is massively favored, but for the rest of the country there’s no comparison.

    “But they do argue that. I’m glad you agree it’s crazy.”

    Huh. I hope they use better numbers/more favorable studies or something.

  • Al

    Well, I don’t know about you guys, but if I were in charge of the transit system I would make a new policy: to send every train on a 30 minute detour along a little-used track. No stops, just barrel down the track at top speed, turn around, and come back. Waste everyone’s time, cost a bit more in energy and salaries and maintenance.

    But: by the per-passenger-mile metric, this is a great idea! For a minimal extra cost, you can double the trip length, and slash the per-passenger-mile subsidy required to almost half. Fantastic! Why hasn’t anyone thought of this before?

  • garyg

    Sure I could. Take it to the logical extreme — Say there’s one mass transit subsidy, period. Let’s say $200 per passenger mile for a monorail group-bicycle that travels half a mile. So, say this example costs a couple million. So, a couple million for monorail bicycle, and billions for car transit. Car transit is massively favored, because there’s just not that much investment in monorail bicycle. Sure, at that particular half-mile stretch the monorail bicycle is massively favored, but for the rest of the country there’s no comparison.

    Huh? Again, the total subsidy is irrelevant. What matters to incentives is the unit subsidy. If the government subsidizes each apple 20 cents and each orange 80 cents, the subsidies massively favor oranges over apples, because the government is paying a much larger share of the cost of oranges. If oranges are rare because there is little demand for them even at the massively-subsidized price, that just shows how much consumers prefer apples to oranges. Same thing with your (absurd) hypothetical.

  • garyg

    But: by the per-passenger-mile metric, this is a great idea! For a minimal extra cost, you can double the trip length, and slash the per-passenger-mile subsidy required to almost half. Fantastic! Why hasn’t anyone thought of this before?

    Er, because it’s obviously stupid? Why would transit agencies send trains on 30-minute detours that have no benefit but just increase costs and “waste everyone’s time?” People wouldn’t be willing to pay the higher costs of longer trips if longer trips didn’t provide a greater benefit than shorter trips. If shorter trips provided just as much benefit as longer ones, people would take the shorter trip instead.

  • Al

    Yes, exactly. It’s obviously stupid, but it also decreases the, as you say, “unit subsidy”, which you claim is the most important metric.

    (Regarding people taking the shorter trip instead: there is no shorter trip. All trains are rerouted.)

    If the per-passenger subsidy is $2, and the train runs 20 miles, then the per-passenger-mile subsidy is $0.10. If you send it on a 20 mile detour, so that it runs 40 miles, which requires an additional subsidy of $0.40 per passenger (the marginal cost of extra travel being small), then the resulting per-passenger-mile subsidy is $0.06. The plan is idiotic– a larger total subsidy for a worse result– but it improves the metric.

    Which is why the per-passenger-mile subsidy is a poor metric.

  • @garyg Let me try again by extending my hypothetical: monorail group-cycling would be massively subsidized in one place. Cars would be less subsidized per place, but overall would be much more subsidized. So, choose any individual place, and cars are much more subsidized. I.e., the average subsidy per possible places is higher.

    In less theoretical terms, Acela in the eastern states might be more subsidized per passenger mile, but where I am, the subsidy is all for cars.

    And remember that I’m not trying to convince you that the argument is right (I don’t care that much, and there’s no way I’d convince you, even if I did care.), I’m merely trying to show that there are ways to “escape the conclusion that subsidies massively favor transit over cars”.

  • @Al Your argument is theoretically sound, but in practice it seems unlikely that people would want to game the metric.

    Certainly, I think the per-trip metric is better, but you could say, “And people could take hundreds of trips to a neighborhood store.”

    Is there some metric that doesn’t run into your theoretical problem, where the numbers can be skewed if people intentionally do something stupid?

  • garyg

    Al,

    You’re not listening. We’re measuring the costs and benefits of real trips, not hypothetical trips that people would not make, such as your absurd hypothetical. For about the fourth time, we know that longer trips provide more benefit because people would not be willing to pay the higher costs of longer trips if they weren’t getting more benefit in return. They’d take a shorter, cheaper trip instead. People would not be willing to commute 30 miles if they could get the same benefit from a 1-mile commute. Typical benefits of longer commutes are cheaper housing and better schools. People would not be willing to travel 5 miles to a supermarket if they could get the same benefit from a half-mile trip to a neighborhood convenience store. Typical benefits of longer shopping trips are lower prices and a larger selection of products. Your claim that the value of trips has nothing to do with their length is contradicted by fundamental economic principles of the relationship between costs, benefits and demand. Not to mention being contradicted by simple common sense.

  • Al

    1. My point is that the per-passenger-mile subsidy is a poor measure if it results in ridiculous trips being judged better than non-ridiculous trips, purely because they’re longer.

    As for what metric is better, I would say that it would have to take into account the following: per-passenger subsidy, per-capita subsidy, pre-capita cost to the traveler (if you can trade a high direct cost for a low subsidy, then that might be a good idea), externalities like pollution and health impacts, the time-cost to the traveler (and also to others, like drivers on roads parallel to the rail line), the number of accessible destinations, resistance to oil shocks, and, yes, per-passenger-mile subsidy.

    Finally, transit promoters like to play down the value of transit to poor people, because “let’s make it difficult for poor people to come here” has killed more than one good transit project. But the value exists, and it’s a good one, because it removes a barrier to entry to society.

    2. People might also not be willing to commute 30 miles or travel 5 miles to a supermarket if the roads hadn’t been rebuilt for high speed through-traffic at the expense of local traffic; if there were no free parking requirements at the endpoints; if the price of gas reflected the environmental & social cost; and if there were any alternatives to driving, not just transit, but simple things like sidewalks and safe and convenient street crossings; and zoning that allows small local businesses to operate without paying for a large parking lot.

    I’m not saying that a longer trip can’t have more value than a shorter trip. I’m saying that public policy has been single-mindedly focused on subsidizing long trips, far past the point of getting a good return.

  • garyg

    1. You are confused. We’re trying to measure transportation costs and benefits. The unit of measure itself doesn’t “take into account” the individual components of those costs and benefits (subsidies, user costs, external costs, etc). It is used to calculate them. Any conceivable unit could result in “ridiculous trips being judged better than non-ridiculous trips.” If you make ridiculous assumptions, you’ll get ridiculous results. If you think there is a better unit for transportation cost-benefit analysis than passenger-miles, what is it, and why do you think it’s better?

    2. You’re still missing the point about the relationship between costs, benefits and trip length. Whatever effect the other variables you list have on costs and benefits, costs increase with trip length. A 30-mile bus trip costs more than a 1-mile bus trip. A 30-mile car trip costs more than 1-mile car trip. The longer trip consumes more fuel, more time, more wear-and-tear on the vehicle, and in the case of the bus trip, more in labor costs for drivers, mechanics, etc.

  • Al

    Hmm. So you mean that in addition to the variation in cost per passenger mile you also look at the variation in benefit per passenger mile, so that for instance three miles of travel on a city street may have a higher cost than on a rural highway, but also has a higher benefit (many more destinations)? That is something I can get behind. My main beef with the measure was that it implicitly compares a bus moving a person three miles in a city with a car moving on a highway, finds that the car is cheaper, and that, presumably, buses should be replaced with cars. But if you also add in the benefit of traveling three miles in each situation, then you can start making a meaningful comparison (and that does neatly account for ridiculous detours, because though the cost per mile might go down, so would the benefit per mile).

  • Jason

    garyg

    Regardless of subsidies and costs per mile, cars are the leading killer of people in the US ages 5 – 24. We’re averaging 33,800 fatalities a year. That’s the equivalent of twelve 9/11’s a year. That’s incredulous for any first-world nation to accept. Everyone on this forum has had friends and family members killed in a car…there is no price you can place on these lives. You’re chance of death or maming on public transit is exponentially less than when sitting behind the wheel. Our PTA held a bike-to-school event in November and counted 42 parents texting while driving in the school zone. People can yell for enforcement all they want, but you simply can’t police that many people. Sadly, you’ll simply sigh it off as a “cost of doing business”. Since when did your convenience trump human life? By this afternoon alone, 30 people in this country will be dead from cars. Hell, by the end of the week we’ll be at 500. That’s four fully filled 737’s falling from the sky…and that’s just a week! How does that register with your economics?

    What we’ve seen in the US in the past decade alone is a massive investment in alternative modes of transit. Combined with that we’re seeing an inverse relationship between higher transit ridership transit/bike-use and lowering of auto fatalities:

    http://en.wikipedia.org/wiki/List_of_motor_vehicle_deaths_in_U.S._by_year

    With the increase in transit use, we’re also seeing a correlation to personal savings rates:

    http://www.creditwritedowns.com/2010/02/chart-of-the-day-u-s-savings-rate-over-last-60-years.html

    Annecdotally speaking, my wife and I gave up a car in 2009. We used the savings from taking public transit (approx. $25K over two years) to start a small business in our community. We now employ 12 people and will grow to include an additional 10 by the end of April. From the success we’ve had, we’ve convinced other friends to do the same.

    And no matter whose politics you side with, we’re funding both sides of this war by our dependency on oil. Not a smart move for a nation who prides itself on innovation and independence.

  • garyg

    So you mean that in addition to the variation in cost per passenger mile you also look at the variation in benefit per passenger mile, so that for instance three miles of travel on a city street may have a higher cost than on a rural highway, but also has a higher benefit (many more destinations)?

    No, I mean exactly what I have said several times now. Longer trips cost more than shorter trips. People wouldn’t be willing to pay the higher costs of longer trips if they didn’t get a benefit in return. That’s how we know longer trips provide more benefit than shorter trips. I’m not sure why you’re having so much trouble understanding the basic economic relationship between costs, benefits and demand.

  • Al

    Well, sure, that’s true for a given location. But it’s not true across locations: a trip in New York City probably costs more– and people are willing to pay more– than a similar or even longer trip in Montana.

    What I object to is the idea that you can find the average per-passenger-mile cost of traveling by car over the whole nation, as if you can average out the cost of a commuter’s time in traffic in LA with the cost of rust problems in Maine and the gas on a coast-to-coast trip, then average the per-passenger-mile cost of the NY subway with buses in suburban Seattle, compare the two figures, and come up with a meaningful conclusion.

  • Gonewest

    Ok, I admit I stopped reading the back-and-forth after a while, so apologies if this was answered somewhere above.

    But…

    What does Delucchi say accounts for the massive subsidies for public transportation? I don’t see any backup for that part of the analysis. Are buses and light rail merely victims of economy of scale (meaning, they never achieve a volume of ridership that allows them to reach their true potential efficiency?) I’m just asking.

    And secondly, in Delucchi’s article he rolls in “total operating plus capital costs” for public transportation. I assume that includes the cost of the bus or light rail itself, the maintenance garages, the bus stations, and so forth. But I don’t see where he estimates the corresponding cost of car *ownership* per vehicle mile — auto lease payments or auto loans, the value of the physical structure where I park my car at home, my maintenance fees, insurance, registration (ok, small), etc.

    Am I reading his paper right? Is it correct to capture the capital costs and operating costs of the public transit systems without accounting for corresponding costs in the case of a car?

  • Gonewest

    Thought a little more about this during my commute home. Let’s say I lease a car at $350 per month, insurance at $100 per month, maintenance at $25 per month, and I drive 15,000 miles per year. Let’s also say registration fees are negligible, and I’ll ignore the cost of ownership of my garage (because I don’t want to bother right now). And I’m ignoring fuel costs because I think Delucchi includes that in his “oil use, water pollution” row.

    So total ownership cost for a car is $5,700 per year, or $0.38 per vehicle mile. (Yes, I know leasing is a relatively expensive way to use a car but I didn’t want to get into arguments about the average maintenance cost of used cars of varying age and quality, etc).

    Which makes the per-vehicle-mile cost of a car similar to that of riding a bus — inclusive of subsidies and externalities — and assuming it’s correct to count costs of owning a car as a “subsidy” in the car column.

    Someone is going to challenge this… please help me understand why this isn’t valid.

  • Al

    I think the people who argue against mass transit aren’t particularly concerned with total cost, but with subsidies only. Doesn’t matter how expensive it is, if the user pays his own way, and no one gets a free ride. Also, centralized planning is inherently bad and inefficient.

    They might also argue that if the costs are similar, then the car is clearly superior, because it is faster (most of the time) and can take you to a greater variety of places.

    I think this is bunk, and my counterargument is approximately:

    Cars are only as cheap per mile as they are if you drive a LOT. If you live in the city and your day-to-day life is contained within a dozen miles, you’ll get nowhere near that, since you still have the big fixed costs. Transit, on the other hand, is at its best in dense areas and is great for the person who only needs a handful of miles (even if they were more expensive ones). This is especially true for low-income folk. Also, the costs of the car, particularly the ones that are usually ignored by promoters, are at their highest in cities (parking, both at home and away, traffic delays on limited roads, for both you and other drivers, pollution-health costs). The typical response at this point is to say that this only proves that cities are obsolete relics, and the conversation degenerates.

  • Gonewest

    My gut says transit works better in dense areas with high utilization. My ad hoc thesis: Dense areas lend themselves to better coverage via transit routes. Better route coverage encourages more riders. More riders brings down the amortized cost per trip. The more I ride transit the less I need a car. The fewer cars in a city the less space allocated to parking. Less parking, more density. And that closes the cycle.

    Additionally, if I live and work in a dense area then my typical commute is shorter. I spend less time in a car. Commute time is an opportunity cost that I can otherwise spend with family, at work, arts, volunteering, church, or writing comments on Streetsblog… To my way of thinking I would (and in fact I do) trade the longer commute to a suburb for a shorter commute without a car and the freedom to spend my time and money elsewhere.

    And when I am in a properly designed city I find I spend my extra time and money in public spaces, interacting with other people, and it’s just a generally more interesting and engaging life.

  • Phil from NZ

    There are distinguished economics texts (eg by Colin Clark) that analysed “externals” of roading and concluded that external BENEFITS outweighed external costs to such an extent that public subsidy of roading was entirely justified.

    I have never seen a study that scientifically contradicted this; all of them proceed from a position of total ignorance of external BENEFITS.

    This is far worse in the case of city roads than it is for highways.

    (By the way, the taxation revenue from new car sales, accessories, repairs and servicing, etc etc; are also substantial and should be taken into account. The government would not have this revenue WITHOUT roads and autos.)

    We simply would not have a modern economy as we know it without “auto-mobility”. Even the term “external benefit” is really far too weak. Auto-mobility has simply been THE most major enabler of wealth creation for decades. Richard Florida is right; most of the wealth creation in the world in the last 50 years has occurred in cities. “Density” per se OBVIOUSLY does not correlate to wealth creation, otherwise Calcutta would be more prosperous than Houston. The explanatory factor here is that Houston has much higher “enabling” of economic activity between market participants; culture is one feature, but “auto-mobility” is an important one.

    Basic geographic and spatial economics teaches the connections between infrastructure, land use, incomes, and land values. Rail-based systems concentrate the value gains around “destination” stations. Roads and auto-mobility are far more “democratic” in terms of “who benefits”.

    Land prices rise at “efficient” locations, including near railway stations, to the point that eventually the number of people who can afford to live there, finds its own level. Just because planners identify demand driving land prices higher at these locations, is NOT a “case” for expecting everybody to live that way. Duplicating tracks and rail routes in an endeavour to “democratise” the benefits like roads do, is simply economically impossible – it would cost more than the total value it creates.

    Having said that, I agree that roads and parking spaces SHOULD be made to pay under simple free market principles, such that the “return on investment” matches that of the surrounding real estate. This would certainly spell doom for most roading networks in the highest-density, highest-value CBD’s. But there is a big “UNLESS” here. That is, UNLESS the owners of surrounding high value property were prepared to “pay in” to continue to enable auto-access to their properties. THIS is the missing element in funding roading in high-congestion areas.

    I have no idea what the results of such pure free market ideology might be; but I am prepared to trust it as “the most efficient solution”. The value of real estate is largely determined by the “access” there is to that real estate; if the values of the road space and the real estate space had to balance each other, I doubt that the real estate values would go anywhere near as high as they currently do due to de facto subsidised roads and parking spaces. But somewhere along the line I am convinced that the CBD property owners would see a strong case for “paying in” to the cost of the roads. THIS “subsidy” would be entirely just.

    Yes, rail transit systems might be discovered to be more viable after all, but I insist that these systems must be paid for primarily by those who will benefit, otherwise (Colin Clark states this quite definitely) the transit becomes merely a wealth transfer from those paying the subsidies, to the owners of conveniently located property (which already happens to be worth a lot due to CBD location).

  • Phil from NZ

    There is just a little more that needs to be said about the cost-benefit of public transport versus private autos. When the government subsidises roads, it is getting a bargain in that a high proportion of the cost of economy-boosting auto-mobility is in fact already paid by the buyer-owner of the auto; who also pays for its fuel and pays for its repairs, and replaces it when it has worn out.

    Public transport subsidies, however, are frequently expected to cover a substantial share of the equivalent costs of rolling stock as well as rails; and repairs and maintenance; and energy costs. The increasing labour cost element in public transport continues to tell against it to a greater and greater extent.

    One of the beauties of “auto-mobility” is that everyone has considerable choice of how much cost to outlay; extremely low cost running can be had from small Japanese cars that are a few years old. Much of the “total cost” in the economy, of private persons auto-mobility, is discretionary and based on income. Basing policy decisions on this, will hurt the lower income groups the most, because higher income groups can absorb the higher costs while lower income groups end up “priced out” of essential amenities.

    The single worst thing for low income groups, is the inflation in the price of all urban land when strict regulatory boundaries are imposed by one means or another. While “planners” have mental images of how they wish everybody to respond to the diminished land supply, “the market” behaves in quite different ways to that. What happens, is that “space” and “location” become more and more unevenly distributed along INCOME lines. The largest lots and the most efficient and convenient locations become increasingly concentrated into the hands of the highest income earners, while the distribution of land according to incomes, forces reduced SPACE as WELL as less efficiency and convenience, on people of lower and lower incomes.

    The result of this is actually lower urban efficiency, because greater numbers of households and businesses end up in inconvenient locations compared to the “lower cost land” scenario. This is because the “cost of land” is no longer purely related to its REAL wealth creating capacity, and the “cost of land” swamps “cost of transport” in location decisions instead of there being a logical trade-off between the two. Paul Cheshire of the LSE (and colleagues) estimates the “inequality” effect to be greater than that of income disparities themselves; they have 5 decades of data from British experience with anti-sprawl planning to go by. They estimate the net public welfare effect to be equivalent to a 4% income TAX. i.e. it is not a benefit at all.

    Planners pointing to the “benefits” they think they have achieved, of higher urban densities etc; need to realise that without inflated land prices, there would have been MORE density in the RIGHT places. THIS is why some analysts are starting to notice the phenomenon of “dysfunctional density”. The reason is that the “centre of gravity” of the density has been forced in INEFFICIENT directions within the city. Alain Bertaud’s studies on “The Spatial Distribution of Density” should have won him a Nobel prize; but instead, nobody has taken any notice.

    By far the most rapid improvements in efficiency of urban form, will be achieved in any city that leaves its “limits” freely open to development and hence keeps land prices low throughout the metro; AND encourages re-development at high density in the RIGHT places. Land prices tend to slope up by a factor of about 10, from the fringe to the inner suburbs (and slope up further again into the CBD). People need to be able to make the MOST efficient trade-offs between space and location, instead of being forced to make INEFFICIENT trade-offs.

    Large minimum lot sizes absolutely should be abolished, and height restrictions. The closer to the CBD, the more “incentive” there should be for developers and prospective residents. If land prices have successfully been kept from inflating at the fringe, the results in inner suburbs and CBDs are spectacular. Small apartments should be less than $200,000 instead of over a million dollars. High land prices are actually the worst enemy of the planner attempting to increase urban density for desired reasons of efficiency.

  • Phil from NZ

    GaryG, you would be very interested to read Alain Bertaud’s papers “Cities Without Land Markets”, “The Spatial Organisation of Cities”, and “The Costs of Utopia”.

    The reason the former USSR’s cities were so inefficient, is simply that a population that has to travel an average of 1 hour by train is NOT more efficient than one that travels an average of 15 minutes by auto, and is capable of covering a lot more destinations at its discretion if it wishes to do so.

  • Gonewest

    @ Phil from NZ

    I am obviously not an economist… but I see you might actually be, so take this for what it’s worth.

    Regarding density for density’s sake versus density in the right places, I assume you’re referring to places like Los Angeles which has dysfunctional density. A flat distribution of people throughout a huge metro region which creates a situation that has neither the characteristics of a suburb nor that of a dense inner city like Manhattan.

    You seem to be referring to a choice between auto-mobility or rail-mobility. But a combination of auto, bus, subway/trolley, and rail is what really occurs in most cities I’ve visited. Can’t it be said that a combination of various transit modes in a sufficiently dense urban setting creates the ‘democratic’ distribution of benefits that you describe?

    In particular, for many people public transit enables the human interactions that are necessary for a services based economy to thrive. It provides for face to face meetings and traditional office workplaces. This must work, else the success of banks and law firms and advertising and publishing in Manhattan must be explained some other way.

    Also, in cities like Manhattan the roads are still available for cars, taxis, and buses. Also delivery trucks. So in a dense urban setting it intuitively seems that roads support the physical transactions that are necessary for a goods based economy, and also afford additional transit options for those who can’t do it on rail alone.

    What’s really interesting to think about is the potential impact of teleconferencing and telepresence systems from companies like HP or Cisco or Telaris. I’ve had the opportunity to work at an office that uses such systems daily, and without a doubt it affects the way we think about travel. It won’t change the xfer of physical goods, but it does change everything else.

  • Phil from NZ

    @Gonewest,

    Thank you for your thoughtful response. You actually have a better grasp of what I am talking about than 90% of the people I try and discuss this subject with – including “economists”. Their track record on urban and transport issues is extremely poor, far too few of them have ever studied the necessary specialities. The whole financial crisis is connected with this failure in “land economics”.

    Colin Clark actually wrote decades ago, that the level of density that an urban core could sustain as 20th century development occurred, actually correlated strongly with, and depended on, the amount of roading lane-miles provided in the core. You are quite right to notice Manhattan as an outstanding example of this. While Manhattan has far higher transit person-mile use than anywhere else, it also has far higher numbers of vehicles per square mile than anywhere else. I doubt that the density required to support Manhattan’s uniquely viable Transit, would exist apart from the still-high level of vehicles.

    You ask, “Can’t it be said that a combination of combination of various transit modes in a sufficiently dense urban setting creates the ‘democratic’ distribution of benefits that you describe”? I have no problem with that as long as there are no subsidies from non-users and non-beneficiaries. The “beneficiaries” include the CBD property owners who primarly capture the benefit. Hong Kong basically paid for the construction of its subway system by levying the proceeds of the sale of land around the stations. Of course, this land was under-developed at the time. But the principle is the same wherever you want.

    Currently, these property owning beneficiaries are largely capturing a wealth transfer from the diverse base of payers of subsidies. You will buy yourself a massive fight if you suggest the kind of thing I am suggesting here; yet these people will know that what I am saying is the truth. Somehow, we missed opportunities decades ago, to get honesty in urban transport funding.

    There is an excellent book called called “Downtown: Its Rise and Fall.” by Jack Tager.

    As the 20th century progressed, and people attained greater mobility, CBD property owners attempted via political influence, to perpetuate the pre-eminent retail and office center of a metropolis. Rail networks focused on the core were a popular tool for this purpose.

    The book includes an editorial cartoon from the 1920s in opposition to a plan to build a system of subways in Chicago; the cartoon depicted a businessman wielding a “Loop Tube Magnet” that drew all the wealth from the surrounding region towards it.

    That is not too much of an oversimplification of what the fight was about in Chicago then, and what it is about all over the world in cities today. The vested interests who stand to gain from Transit projects do not care a jot about the “net cost to society”, all they care about is the “net gain to them” at the end of the wealth transfer from the hundreds of thousands of people who will not benefit.

    It should have been possible to pay for the “ARC” tunnel under the Hudson that Gov Christie rightly refused the tab for, by levying the right beneficiaries. One study calculated property capital gains at the Manhattan end, if the project went ahead; to be $36 billion dollars. Why should New Jersey taxpayers pay $10 billion (plus) and Manhattan property owners nil? If we had honesty in transport funding, CBD property owners would be pushing for projects in the full knowledge that they are going to pay AND benefit, with the benefit exceeding the cost. Currently, we have vested interests pushing for projects from which they will benefit, and not pay for; even if the cost exceeds the benefit.

    You are exactly right about “dysfunctional density”. Los Angeles has this, so does London, Vancouver, and Sydney. Alain Bertaud’s papers use the term “spatial distribution of density”. “Dysfunctional density” IS everywhere the consequence of inflated land prices plus regulations AGAINST increased density in the right places.

    The way to avoid “dysfunctional density” is to remove regulations AGAINST density in the right places (minimum lot sizes, height restrictions etc), AND keep land prices LOW throughout the metro by simply allowing “sprawl” at the fringe. The small amount of sprawl that will take place at the fringe, will be well and truly over-compensated for the “densification” that will take place in the right places.

    Even Portland displays the wrong spatial distribution of density. If you do not keep land prices low by “allowing sprawl”, the inflation of land prices will defeat your best intentions for density in the right places. You will sell a lot more inner city apartments if they are under $200,000 like in Houston, than you will if they are over $1,000,000, as in Sydney, Australia. Australian cities are the classic example of the connection between anti-sprawl planning, inflated land prices, and “dysfunctional density”. They never had low interest rates, easy credit, Fannie Mae and Freddie Mac, Derivatives or Credit Default Swaps, or any of the other reasons commonly trotted out to explain “the US property bubble”. Australian cities are surrounded by spare land every bit as abundant as Texas. Their climate is just as repugnant.

  • Phil from NZ

    Gonewest,

    I really didn’t respond strongly enough about this definition:

    “…..dysfunctional density. A flat distribution of people throughout a huge metro region which creates a situation that has neither the characteristics of a suburb nor that of a dense inner city like Manhattan…..”

    That is a good definition, but it does not go far enough to explain what inflated land prices do. Portland’s OUTER suburbs now clearly have HIGHER density than its inner suburbs. Curitiba, Brazil, is even worse.

    It is not merely a matter of creating a flat distribution, but of creating “peaks” in the wrong places – even if they are low peaks. The fact is that free markets DO create urban form with a HIGH peak in the RIGHT place; regulatory interference that forces land prices up, not only stops this from happening, it creates a kind of de facto gated community in its place, with artificially LOW density in comparison; only the wealthiest people being able to afford the prices.

    The longer these bubble-creating policies are pursued, the worse we can expect the inequalities in society to get. At the bottom, we can expect rising numbers of people living in trailers, tents, subways, and sewers. In other words, we will regress to third world levels of housing inequality.

    Hernando DeSoto and Alain Bertaud have both written definitive studies on the way the poorest people in third world countries COULD be enabled to get a foot on the “asset” ladder if only they could get a LEGAL home within their income. It is morally wrong, even in these circumstances, to require standards of legal housing so high that hundreds of millions of people end up legally “homeless”. We in the “first world” are merely starting from a higher base, and regressing; with our newfound fashion for urban growth controls.

  • Al

    Phil: Thanks for the interesting thoughts. I found myself nodding in some places and confused in others. I have to spend a little more time on it in the future. For now, one or two comments:

    I like very much the idea of mobility as the enabler of wealth creation. WRT the public transit/private auto debate, I think it might be useful to compare the relative mobility and costs (both subsidized and direct) in a city such as New York, with extensive transit and somewhat limited auto-mobility (limited by parking and traffic costs) and a city such as Houston with the opposite. I suspect New York might come out ahead because of factors like vastly increased mobility on foot (or bike) outweigh the inefficiencies of transit.

    You write: “One of the beauties of “auto-mobility” is that everyone has considerable choice of how much cost to outlay; extremely low cost running can be had from small Japanese cars that are a few years old. Much of the “total cost” in the economy, of private persons auto-mobility, is discretionary and based on income.” I think you’re underestimating here the effect of traffic delays and subsidized parking costs. When large numbers of people moved to cheap housing in the suburbs, that put an enormous strain on the road network that didn’t exist before. The low-income person who switches to driving may not require the direct subsidy to public transit (though it’s not much reduced for him leaving it) but does cost a lot in other people’s time in traffic, and in subsidized parking (even here in the city, where parking is in such short supply that occasionally you read about someone getting killed over it, the political pressure to provide free parking is immense– and it is done.)

    I’ve read, approvingly, about London’s greenbelt scheme. I take it you’d be opposed to this, or other large urban parks? Or am I misreading you?

    I strongly agree with the ideas of reducing density regulations (though I have sympathies for historic preservation), and the method of funding transit by taxing nearby properties for added value seems logical. I think

    You will sell a lot more inner city apartments if they are under $200,000 like in Houston, than you will if they are over $1,000,000, as in Sydney, Australia. Australian cities are the classic example of the connection between anti-sprawl planning, inflated land prices, and “dysfunctional density”.

    It seems to me that the difference is less to do with anti-sprawl and more to do with anti-density limitations in Sydney limiting supply, and car dependence in Houston limiting demand (what’s the point of living in an apartment in inner Houston if you still end up driving everywhere?).

    From what you’re saying, I take it you mean that Houston is not an example of ‘dysfunctional density’. But I think it has one of the strongest anti-density regulations: parking minimums. For instance, in my neighborhood there are a number of barber shops, which are often in small storefronts of 500 square feet or less (and offer no parking). Reading Houston’s parking regulations, that salon would be required to have at least 10 spaces for employees and customers, which would take an area about 3000 square feet– a very effective way of discouraging density.

  • Phil from NZ

    Al,

    I am feeling very encouraged by the results so far, from participating here.

    It is great to engage with someone who works on grasping the complexities – and you have pointed out some details I was not aware of.

    Here is what I believe about a comparison between New York and Houston. Manhattan’s density is not something that we are sure can be replicated even in decades. But “automobility” CAN be a substitute here and now, for density, as a “demand enabler” and wealth creator. THIS assessment by Tory Gattis in 2006 was interesting:

    http://houstonstrategies.blogspot.com/2006/05/density-vibrancy-and-opportunity-zones.html

    (Do follow the embedded links to his earlier essays)

    Possibly the main point I am researching and disseminating in forums like these, is that land prices really matter. I do not beleive that any other city will ever increase the density of their core by any significant amount IF they have driven land prices on the fringe up. The “farmland price” of fringe land IS the essential initial ingredient in “urban development”. Manhattan also has a crucial location with centuries of history behind it.

    The economic history of cities is full of examples of CBD’s that have “imploded” long before density was anywhere near Manhattan’s. Colin Clark puts this down to congestion, especially road congestion; but other infrastructure matters too. Infrastructure for density really needs to have been in place ahead of time, because it is extremely expensive to “retrofit” it.

    I have come to suspect that a genuine free market in land, roads, and parking as I describe, would simply NEVER reach anything like Manhattan’s densities. These densities absolutely require subsidy of infrastructure. If we could “cost” the subsidy of Manhattan’s roads (and parking) in terms of the return on investment value of surrounding real estate, I think we would find it to be quite colossal. I also think that even subway Transit would require colossal subsidy to support the same kind of density.

    But the answer is simply to levy the capital gains of the surrounding properties; these gains are also colossal. Or, the infrastructure should be owned and operated by syndicates of adjacent property owners. At some point, “market values” should balance out and “optimum density” be reached. I still suspect that this would be below Manhattan’s. I also suspect that the “efficiency” of super high densities is being under-estimated because of the sheer amount of territory required to perfom “supporting roles” like food production (and affordable homes for a large proportion of the workforce). I do not beleive that it is possible to have a “city” like Manhattan without “affordable” homes somewhere around; and I do not beleieve it is possible to have a “city” like Manhattan with affordable homes incorporated into the monolith itself, so to speak.

    Hong Kong is an interesting case because the whole thing was a sovereignly independent, truly “international” city and an “economic free zone”.

    A real problem you have, is that the “move to the suburbs” in most of your cities, involved unnecessarily LOW densities due to regulatory mandates. It will take decades to get this genie back into its bottle, and you never will if you push land prices up. (My most important point). All you will get then, is “dysfunctional density” with the increases in density happening at the wrong places, and the inner core itself resisting densification, because land prices there will reach Manhattan levels long before Manhattan’s income levels are being provided.

    The “disconnect” between incomes/GDP and land values IS the central issue in the whole global financial crisis.

    I do not mind at all when affordable vehicles mean that low income people can create congestion that higher income people suffer from. The alternative is undemocratic, and increases rather than decreases inequality. It would be better to increase road capacity rather than price lower income people off the road. Lower income people then also suffer from lack of affordable homes near Transit as well.

    Even arterial routes cause capital gains at their “destination” end. It is little understood that this is the best source of capital for road capacity increases. Tolls and gas taxes and subsidies from general taxation and infrastructure levies on fringe development, all represent a wealth transfer to the owners of properties that capture the largest share of capital gain.

    It is easy to retain “Green space” and heritage areas, and not inflate land prices at all, if you keep the fringe free to develop. Some of the most sprawling cities have the cheapest land, the most green spaces, and of course people have the most “green space” of their own on their own properties.

    The problem with a “green BELT” is that it constitutes a barrier at which serious land price discontinuities occur; the entire area within them becomes increasingly expensive as a long term trend, with cyclical up and down volatility getting ever higher too. The papers of Paul Cheshire and colleagues at the LSE are the definitive analyses of all this. Green spaces need to be “discontinuous” so as to avoid this effect. Multiple discontinuous green spaces actually end up enjoyed by far greater numbers of people than a “belt” that ends up with adjacent suburbs monopolised by high income groups.

    I am interested that “free market” Houston has such non-free-market parking mandates. I certainly do not agree with them. This helps to explain to me why Houston under-performs compared to what I would have expected.

    I don’t exactly think car-dependence ever limits demand. Real Estate Agents will generally tell you they can sell more apartments with carparks almost anywhere, than without carparks. The places where so far, it makes sense to sell apartments without car parks, is very limited. Manhattan, of course. And replicating Manhattan will require decades of careful incentives combined with de-regulation to keep land prices low. Then apartments without car parks wil become a lot more common. Meanwhile, TOD provides most of the potential there IS. It is not controversial to say that you can’t have TOD everywhere, and use land efficiently. Transit is “radial” and does not lend itself to utilisation of land between the “spokes”. A “ring” of Transit is only viable in the innermost part of the highest density inner cities.

    The difference in apartment prices between Sydney and Houston (or Atlanta or Indianapolis or any city with a low median multiple in the Demographia survey) is mostly explained by the cost of fringe land. The cost of fringe land is like a “denominator” from which the price of all other land in the metro is derived. I strongly agree that LOW density regulations WITHIN urban growth boundaries will only intensify the inflation effect on inner city apartment prices; that is a good observation to note. It is completely incoherent to plan growth boundaries, and regulate AGAINST density ANYWHERE.

    I hope this remains of interest to you all.

  • Al

    Ok, this makes a lot of sense: “It is completely incoherent to plan growth boundaries, and regulate AGAINST density ANYWHERE.” This is a frustration I have with local politics– they spend billions to extend the subway to the suburbs, only to end at a parking lot. Then the ridership is anemic, and they wonder why.

    What I find somewhat baffling is your focus on low land prices: “replicating Manhattan requires keeping land prices low”??? (I don’t generally use multiple question marks but I’ll make an exception in this case.) It seems obvious to me that density goes hand in hand with rising land prices: no one in their right mind would build a skyscraper in the middle of nowhere (with the exception of stupid-rich and self-aggrandizing gamblers). You’d just build a sprawling building of a couple stories at a small fraction of the cost.

    Regarding urban growth boundaries: I recall a study that compared two potential Walmart projects – one on a ‘greenfield’ site, one on a reused industrial site. It concluded that though the municipality would collect approximately the same amount of additional tax revenue from both sites, in the case of the greenfield site that revenue would be entirely consumed by the costs of building and maintaining the additional infrastructure required (sewage, roads, and other services) to support it, while those services were mostly present in the reused site, and had to be maintained either way. Perhaps this is the situation growth boundaries are meant to prevent.

    Even arterial routes cause capital gains at their “destination” end. It is little understood that this is the best source of capital for road capacity increases. – Maybe. But the destination may well be already saturated with traffic. Not much of a benefit to the destination if the new arterial only delivers people who displace others. If there’s a congestion charging scheme it’s a different story, but then you’re basically back to tolls anyway.

    I don’t exactly think car-dependence ever limits demand. – What I mean is that if you live in NYC, it’s a tradeoff: you can live in a smallish apartment with no parking, but you can get milk three doors down, walk to most of your errands and ride the subway to work; or you can have a yard and a garage and a car, but you have to drive everywhere you go. In Houston you can live in a big house and drive everywhere, or you can live in an apartment and drive everywhere. No surprise that there’s less demand for apartments.

    I do not mind at all when affordable vehicles mean that low income people can create congestion that higher income people suffer from. The alternative is undemocratic, and increases rather than decreases inequality. It would be better to increase road capacity rather than price lower income people off the road. Lower income people then also suffer from lack of affordable homes near Transit as well. – Sure. But we’re basically talking about subsidizing low-income people’s access to transportation, for the same reason we subsidize education. And I agree that it’s a good thing to do so. I just think that public transit is a better, more sustainable way of doing it. The lack of affordable homes near transit, I would say, is entirely due to anti-density regulations–in most cities, apartments are required to have one or more parking spaces regardless of transit accessibility. Add a 300 sf parking spot to your 500 sf studio, and costs are bound to rise significantly, especially considering that transit-accessible means higher land value. Consider that, as expensive as New York is, large numbers of working poor and recent immigrants make their home there. When considering the cost of housing, as in that Demographia survey, it’s sloppy to not consider the cost of transportation, which can be as high or higher. Much of the cheapest housing is paired with high transportation costs. I think it’s notable that many of the “severely unaffordable” cities are known for their extensive transit networks, while few of the affordable ones are. It would be interesting to compare that chart with a similar chart of transportation affordability– perhaps it would be in reverse?

    Incidentally, here are Houston’s parking requirements: http://library.municode.com/HTML/10123/level4/COOR_CH26PA_ARTVIIIOREPALO_DIV2REPASP.html
    I think it’s noteworthy that they effectively ban just about any structure or business in Manhattan. Even in my “streetcar suburb” they would prohibit 95% of businesses, and 70% of housing.

  • Phil from NZ

    Al, I agree with almost everything you say. I would very much like to get my point about land values across to you, because you seem to have the right kind of understanding necessary to grasp it.

    You clearly understand that as a city grows, the value of land rises in “established” areas, relative to the fringe. Initially, the density that results is actually consequent on the prices. Agglomeration efficiencies can take over at some stage, so that the value of land is driven higher again, than the effect owing to its “location relative to the fringe”.

    But when the value of FRINGE land is driven up well above the norm that is set by its normal use as farmland or forestry, the “denominator” of the price of ALL land in the city is higher, and the price of all land is higher than it otherwise would have been set purely by locational advantage and use. These higher prices then become a BARRIER to densification, rather than an incentive for it.

    You can’t just say, “lets ration land so strictly that prices of land quickly go as high as Manhattan, then Manhattan densities will follow”. When you have Manhattan LAND prices and Chattanooga’s population, incomes and GDP, you will not densify Chattanooga, you will kill it. THIS is what is happening all over the world, and is the main cause of the GFC.

    The normal free market rise in land values is actually connected to the rising INCOMES (and the city’s rising GDP) that are the result of efficiencies of scale, agglomeration, and location. Land values rising way ahead of GDP, are a catastrophe for the economy on multiple levels.

    Britain is a classic case study. Their whole economic competitiveness has been eroded by high land prices. The cost of living has placed considerable strains on society; the effect of rationing the total quantity of land rather than rationing the amount per person (which would require totalitarian government anyway), is that the distribution of land becomes ever more unequal as the price rises and forces the lower incomes to economise disproportionally on land and floor space. Efficient “location” is also rationed more and more disproportionally by income, with the result that denser and denser zones form further and further away from efficient locations.

    Paul Cheshire at the London School of Economics, is the most advanced academic in the world on these issues, because of course he has 5 decades of British data to analyse. His latest paper shows that the “retail sector” in Britain is “located” on average, well below international norms of efficiency, because of inflated land values; without these inflated land values, the retail sector’s carbon footprint would be much smaller.

    In the Wal-Mart example you give, there would be a far LOWER “land cost” factor in the decision, in any Demographia low-cost city, than in LA. The higher the average/median land price, the higher the chance that the location decision will be the “fringe” one.

    It is the same for households. You have noted well, the role of transport costs in the total cost of living “location” decision trade-off. But there is a difference between cities where land values are derived PURELY from incomes and location and agglomeration efficiencies; and cities where the “denominator” is higher because of “rationing”. Where the “denominator” is higher, LAND price will assume a much greater “weight” in the trade-off decision between location cost and transport cost; this is why density increases FURTHER AWAY from efficient locations.

    This is also important to notice when looking at the success of TOD. The success of TOD will be lower, the higher the denominator i.e. the higher “average” or “median” land prices are. This is where I find the Demographia survey most useful. It is a strong “predictor” of FUTURE urban efficiency TRENDS.

    Having to do the TOD further and further away from the CBD defeats the whole purpose of it. People want a short transit trip, not a long one. The demand curves for long drives and long Transit commutes are probably different. The more time taken out of people’s day in travelling, the more they require the flexibility of autos to claw back some efficiency in their use of time (chaining trips to shop, etc).

    The bottom line in why the cities of the former USSR were so inefficient, is simply that they built taller and taller apartment blocks at the fringe as it grew. Ultimately, their population density curve was steeply negative. Even if they used Transit almost exclusively, the sheer inefficiency of their population distribution more than negated any advantages. Their lack of use of land between the Transit “spokes” was an almost total wastage, too.

    I used to think that dispersing employment, which is a powerful trend, was a simple claw-back of efficiency, as people are less and less able to afford locations close to the CBD; they supposedly end up travelling shorter distances to jobs in the suburbs instead. But Alain Bertaud’s papers on “the spatial distribution of density” debunked this notion: he calls it a “myth”. People’s travel patterns are almost totally random; there is apparently far too weak a trend to “local living and employment” to result in an efficiency gain from “decentralisation”. A “normal” distribution of density is ALWAYS more efficient; i.e. the steeper the curve up to the centre, the better.

    New York, as it has grown, has successfully kept average/median land prices low, up till more recently. Lower income people have successfully been able to house themselves and still enjoy employment opportunities. This will change, to Manhattan’s detriment, if the greater NY metro area cannot keep its FRINGE land prices as low as the historical norm. There are dozens of different municipalities involved here; the initiative has to come from the State level.

    By the way, I have often wondered why the free market didn’t simply “compensate” the various income groups for the higher costs of land and transport; surely employers would have to raise their pay offers, to find people willing to work at their expensive-to-access location? I have concluded that it is highly significant WHO PAYS for roads and transport infrastructure. Where it is mostly the “traveller” who pays, and “households” (through taxes and subsidies), the cost will end up borne disproportionally by them. This is all the more reason to shift the cost burden onto those who capture the most value; i.e. the owners of property in the biggest “destination” locations.

    (By the way, the inflation of prices of ALL urban land as I have described it above, has led to the biggest capital gains of all time for these people, especially if they sell out at the right times and read the cycles right. Has it occurred to anyone that George Soros funds conservation groups, and then rakes in the capital gains that come when anti sprawl regulations force the price of land up?)

    Lastly, that insight about Houston’s parking requirements has been a revelation to me, you don’t know how much that now explains to me. It has bothered me for a long time, that “free market” Houston did NOT densify according to “best” urban form (steep density distribution graph).

    I guarantee that the “most efficient cities” in the world, will be achieved by
    1) no obstacles to density
    2) raw fringe land prices NOT inflated above “alternative use” prices (farming, forestry)
    3) equitable pricing of infrastructure that “follows the money” instead of assuming that the driver/household/fringe home builder “is liable”.

    I do not know of any city in the world that fulfils the last requirement. Except perhaps Austin, Texas, at least partly in the case of roads (but fails to do so for Transit). Points 1 and 2 are the main reason for European cities generally historically out-performing US ones, but they have increasingly been failing to observe 2) in recent years – hence the international nature of the plague of bubbles (there are OECD papers on this). Most US cities have had 2) but not 1), which still gives Demographia “affordable” results but poor urban efficiency. Some cities have introduced 1) BUT unintelligently regulated against 2), which has been counter-productive to the intended outcomes and introduced housing affordability issues. The WORST have neither 1 nor 2. No amount of introduction of 1) will produce the desired results unless enacted in tandem with 2).

    If anyone “gets it”, I expect you will become as “driven” as I am about it. There are terrible consequences unwinding, as a result of getting all this so wrong.

  • Al

    I don’t know. Maybe it’s my US-centric perspective, but I just don’t see 2) as being a very pressing issue. Hell, we have places here where newly-constructed houses are being demolished because the owners (ie banks) don’t see the value in maintaining them. Most everywhere, including regulation-happy places like the SF bay area, have cheap housing on the fringe where the cost of transport outweighs the cost of the housing. Even if you removed all obstacles, it seems unlikely there would be much movement in most places, especially if they had to pay for their own infrastructure. I suspect, though I have no good sources right now, that much of the fringe development is driven by subsidies, where everything from water to roads to electricity and telecommunications, both private and public, is paid for by taxes and fees on the existing residents.

    You bring up LA for criticism. To my knowledge, if there’s one thing LA has never suffered from, it’s an urban growth boundary. I suppose the national forest to the north is one, but it’s also on terrain that physically limits development, and would likely require very expensive infrastructure. Development typically goes right up to the mountains. Meanwhile, to the east, south and northwest, almost uniform development continues on for dozens of miles. I have difficulty believing that anything at the fringes 40 or 70 miles away would have much of an effect on overall land values, even if it were artificially restricted, which I don’t know that it is. Now, density limitations, they’ve got those in spades.

    “surely employers would have to raise their pay offers, to find people willing to work at their expensive-to-access location? I have concluded that it is highly significant WHO PAYS for roads and transport infrastructure. Where it is mostly the “traveller” who pays, and “households” (through taxes and subsidies), the cost will end up borne disproportionally by them. This is all the more reason to shift the cost burden onto those who capture the most value; i.e. the owners of property in the biggest “destination” locations.”

    If the road to the destination is tolled, and the traveler has to pay to get there, then the destination will have to offer a higher salary to entice the commuter to put up with the toll, no? Certainly there are big ‘cost-of-living’ differences between different parts of the country, which are reflected in the salaries offered. I think this is intrinsically better than arbitrarily judging that someone “benefits” and then taxing them on that ground. There are too many examples in history of someone coming up with the “brilliant idea” of building a freeway through the middle of a city, taxing the city residents to pay for it, demolishing entire sections, and leaving a decayed husk. Now, these projects were probably supported by the “downtown interests”, but the trouble was that they were politically powerful and managed to push through the idea that it “benefited everyone” to avoid paying for it. Better just pay for it through tolls– keeps it simple. (I’m less opposed when it comes to rail and transit projects– less potential for wholesale destruction, fewer externalities like increased traffic, and it’s more obvious who benefits.) Tolling is simply the best way to implement “following the money”.

    Ultimately there’s always going to be rationing of land, for the simple reason that land is actually limited. If the government imposes an additional, arbitrary limit, then sure it’ll change the input variables, but it won’t change the equation.

    This makes absolutely no sense to me: “When you have Manhattan LAND prices and Chattanooga’s population, incomes and GDP, you will not densify Chattanooga, you will kill it.” Simply put, land values are determined by the market. If you have Manhattan land prices, that means that people value living or working in Chattanooga so highly that they’re willing to pay that much. Building a tall apartment building would be chump change in comparison, and if living in Chattanooga is so highly prized, that’s what they’d do, no? And there’s your density. Of course, if you still have Chattanooga’s population and incomes, the land prices would not get that high in the first place, so it’s a moot point.

  • Phil from NZ

    Al,

    I will start from the end and work back. I used Chattanooga as an illustration by exaggeration. My point is still valid. If land prices ARE set by markets, they bear a predictable relationship with INCOMES. Certainly I exaggerated when I suggested that land prices in Chattanooga could be forced up to match Manhattan.

    But look at “median multiple” house prices. (I don’t know if there are studies that analyse “median multiple” LAND prices, if there was, I would expect the results to be even more glaring). Every city that has median multiples significantly higher than 3, has experienced a sustained inflation in land prices at some point FOLLOWING the introduction of an urban growth limit or a proxy for it. Yes, this might be an actual run-up against geographical obstacles. But the example is still useful because REGULATORY obstacles have the same effect. It is bad enough for the national economy is we have a few cities that have truly “run out of land”; it is disaster if we engineer similar inflations in land prices for a whole lot more cities, through regulations.

    Britain is the example of an economy after 5 decades of this as a universal practice. Maybe they WOULD “run out of land” eventually, but they are far from having done so; but the regulations have had the same result for their economy as if they HAD ACTUALLY run out of land physically. (By the way, there is an economic paper that even demonstrates that Japan’s famous land price bubble was not due to running out of land in the physical sense).

    If we really WERE “running out of land” in the world or in any one country, then the price of food would go up to the point that using land to grow food was just as valuable a use of that land, as using it for urban development. We are nowhere near that point today. This is really crucial to understanding all this. The “value of farmland” SHOULD be the “denominator” of the price of urban land. Where the “denominator” is “the price of farmland” TIMES TEN (or 20, or 100 plus as in parts of Britain) plus development cost; you have a major problem. Not just in economic competitive disadvantage, but in “LOCATION efficiency” disadvantages; overall POOR urban form.

    I believe that high “median multiple” house prices are a good indicator of a FUTURE TREND to economic decline in a city, even if local advantages currently overcome the effect, eg Boeing might keep Seattle going. But Boeing itself will ultimately suffer from the effects on its workforce, of ever-inflating house prices and poor location choices.

    I used to think exactly what you say here:

    “….if the road to the destination is tolled, and the traveler has to pay to get there, then the destination will have to offer a higher salary to entice the commuter to put up with the toll, no? Certainly there are big ‘cost-of-living’ differences between different parts of the country, which are reflected in the salaries offered. I think this is intrinsically better than arbitrarily judging that someone “benefits” and then taxing them on that ground……”

    There is nothing “arbitrary” about capital gains. No bureaucrat has to decide to put the levy up, on a property owner; if his property has gone up in value, he IS capturing benefit from the growth of the city and the provision of infrastructure within it. What MIGHT be arbitrary is the % rate of the levy; but this would not be difficult to calculate.

    I actually believe that if CBD property owners historically DID have to pay for infrastructure, they would willingly do so in ANTICIPATION of the capital gains they KNOW they are going to make. The problem is, the people who understand this all become property investors rather than policy makers; and those who do become policy makers get “bought off”.

    What I already said above was, “…..I have concluded that it is highly significant WHO PAYS for roads and transport infrastructure. Where it is mostly the “traveller” who pays, and “households” (through taxes and subsidies), the cost will end up borne disproportionally by them……” i.e. even if we would have EXPECTED wages and salaries offered by employers, to “compensate”. I am still working on the theoretical mechanisms behind this; one point to consider is that employees DO “compete” for jobs; the person who will do the job for the LEAST tends to set the going rate. This very obviously will be least fair to recent home buyers who could not afford recently-inflated house prices anywhere other than in inconvenient locations.

  • Phil from NZ

    I just must make clear that I fully understand that in many parts of the USA, very large “minimum lot sizes” immediately outside certain cities existing boundaries, usually under different local government altogether, constitute a similar “regulatory obstacle to fringe growth” to an actual Urban Growth Boundary. This is also the case in many places here in NZ.

    Often, this results in people in forums like this arguing past each other, because planners and their supporters say “but we don’t have urban growth boundaries, therefore urban growth boundaries can’t be the cause of our land price inflation”.

    One paper I have read somewhere, suggested that abolishing minimum lot sizes in Marin County would go a long way to restoring some measure of housing affordability in Los Angeles. It might have been Glaeser or Fischel.

    I remain sceptical that in any cases, eliminating the minimum lot sizes in the city itself, WITHOUT restoring “access to low cost land at the fringe”, will restore housing affordability. Restoring access to low cost land at the fringe WILL often, however, involve abolishing minimum lot sizes beyond the fringe. But “minimum lot sizes” WITHIN the fringe clearly do not, of themselves, result in inflated land prices. The inflation comes when the minimum lot sizes force the city fringes to sprawl up against “obstacles to growth” much sooner, whether these obstacles are adjoining local governments even LARGER minimum lot sizes (60 acres in Marin County?) or geographic obstacles.

    I am quite excited about recent political moves in Atlanta, Georgia, to encourage higher density re-development in strategic locations. I fully expect any city with low land prices, to achieve spectacular results if they incentivise densification without destroying their cheap land advantage. Georgia is doing it by waiving infrastructure levies in strategic locations. They show encouraging signs that they listened to the superb advice given to them by Alain Bertaud a few years ago.

  • seema

     nice……!

  • Nathanael

    Phil, the usual issue in the cities in the US are prohibitions on building UP. The city isn’t getting any wider — nor do we really want it to if we want to preserve agriculture -, which means the only option is to go vertical.

    But in most cities there are stringent “zoning” restrictions on building height!

  • Nathanael

    “When you have Manhattan LAND prices and Chattanooga’s population,
    incomes and GDP, you will not densify Chattanooga, you will kill it.”

    Not necessarily.

    Because it’s not land prices which matter.  It’s rental prices per square foot of floor area.
    It is important to prevent the ground rents from concentrating all wealth into a few landlords, but that’s true sprawl or no sprawl, and Henry George knew it a long time ago.

ALSO ON STREETSBLOG

New Report: Road Funding From Non-Road Users Doubled in 25 Years

|
(Image: Subsidyscope) The myth that U.S. roads "pay for themselves" thanks to user fees is a subject that’s likely familiar to many Streetsblog Capitol Hill readers — but just how much of the nation’s highway funding is provided by charging drivers? The answer may surprise even active critics of the current asphalt-centric transportation system. Between […]

It’s Time to Stop Pretending That Roads Pay for Themselves

|
If nothing else, the current round of federal transportation legislating should end the myth that highways are a uniquely self-sufficient form of infrastructure paid for by “user fees,” a.k.a. gas taxes and tolls. With all the general tax revenue that goes toward roads in America, car infrastructure has benefited from hefty subsidies for many years. […]

Transit’s Not Bleeding the Taxpayer Dry — Roads Are

|
We’ve said it before and we’ll say it again: Roads don’t pay for themselves. But maybe they should. “Taxpayers cover costs that should be borne by road users,” asserts the State Smart Transportation Initiative at the University of Wisconsin-Madison. “Road subsidies push up tax rates, squeeze government services, and skew the market for transportation.” SSTI, […]

Tell Congress: Don’t Waste Money on Highway Expansion

|
States should know better than to funnel more money into road expansion at the expense of maintenance. With President-elect Obama back in Washington, action is heating up again around the economic recovery package, which could total up to $850 billion over the next two years. As much as $100 billion may be at stake for […]

The Fuzzy Math in the Road Lobby’s Memo to Congress

|
Don’t know what to make of the news that U.S. driving rates have dropped for the ninth year in a row? Looking for guidance about whether your state or city should be wantonly expanding roads or investing in transit, biking, and walking? The road lobby thinks you should turn to them for independent, unbiased analysis […]