Development
Is Cato turning around?
The Cato Institute is one of the leading Libertarian think tanks in the country, but they have long had a big problem. Their foremost writer on transportation and urbanism, Randal O'Toole, doesn't actually believe in Libertarianism.
Photo by dwaynehoov on Flickr.
Although he never uses these exact words, his basic position on all things urban is that 'a large portion of the market prefers auto-oriented suburbia, therefore the state should mandate and heavily subsidize auto-oriented suburbia' (here's a recent example).
It's a profoundly anti-Libertarian position, and it has tarnished Cato's reputation in the field for years. How can they be taken seriously in discussions about cities when their senior fellow on the subject is such an obvious hypocrite?
It is gratifying, then, to find other Cato writers speaking more reasonably about the subject. On Tuesday, Cato published a blog post by writer Timothy Lee titled Free Parking and the Geography of Cities, in which Lee makes the well-founded point that government regulations requiring large amounts of parking in every development inherently make walking impractical, which discourages people from walking, which encourages car use, and that therefore such regulations manipulate the free market.
Progressive blogger Matt Yglesias agrees, and notes that such manipulations instigate a "feedback loop" in which every car-oriented development increases the impracticality of walking, which in turn begets more car-oriented development.
These ideas are a key part of contemporary urban planning. It has long been a mystery to planners why, at least on this issue, Libertarian groups like Cato should be opponents rather than allies.
Lee's piece is just one blog post, but hopefully it is representative of a shift at Cato away from O'Toole-style reactionism against change, and towards a more intellectually honest assessment of what a genuine free market would actually mean for our built environment.
Hat tip to Ryan Avent for succinctly summing up O'Toole's position.
Cross-posted at BeyondDC.
 | Dan Malouff is a professional transportation planner for the Arlington County Department of Transportation. He has a degree in Urban Planning from the University of Colorado, and lives a car-free lifestyle in Northwest Washington. His posts are his own opinions and do not represent the views of his employer in any way. He runs the blog BeyondDC and also contributes to the Washington Post Local Opinions blog.   |
Comments
- Cyclists are special and do have their own rules
- M Street cycle track keeps improving, draws church anger
- O'Malley announces first projects using new gas tax money
- Can Loudoun grow while protecting its rural areas?
- ICC losing bus service in classic bait and switch
- Silver Spring mall could get massive facelift, new name
- WMATA launches "Short Trip" rail pass on SmarTrip
by Rob on Aug 19, 2010 2:06 pm • link • report
by Erik W on Aug 19, 2010 2:14 pm • link • report
by Cameron on Aug 19, 2010 2:42 pm • link • report
by andrew on Aug 19, 2010 2:58 pm • link • report
by aaa on Aug 19, 2010 3:43 pm • link • report
by Chris on Aug 19, 2010 3:50 pm • link • report
1) not really in a city (think Rhode Island NE)
2) your city is so down in the dumps that land isn't worth much
3) Your city doesn't have a workable transit system
Does Houston have minimum parking requirements? I know they didn't have zoning for a long time, and I don't think the city is that different from everyplace else in the country.
by charlie on Aug 19, 2010 4:03 pm • link • report
by Rob on Aug 19, 2010 4:06 pm • link • report
http://www.thinktankedblog.com/think-tanked/2010/08/more-libertarianism-confusionrand-pauls-not-helping-.html
by a2b4u2 on Aug 19, 2010 4:17 pm • link • report
"...if the state is going to pick a fetish to mandate and heavily subsidize, it might as well choose auto-oriented suburbia, as dictated by market preference..."
But then, I'm no transportation guru. I just know that the town Randal and I both are abundantly familiar with -- Portland -- is hardly the virtuous model of urban living that the typical, popular narrative would have people believe. The city, despite its considerable charms, is a bureaucratic abyss.
See Jack Bogdanski's excellent blog for hundreds of examples:
http://bojack.org/
by Iced Borscht on Aug 19, 2010 4:34 pm • link • report
by Bianchi on Aug 19, 2010 4:44 pm • link • report
by Tim on Aug 19, 2010 5:02 pm • link • report
by JustMe on Aug 19, 2010 5:05 pm • link • report
by Andy Peters on Aug 19, 2010 5:10 pm • link • report
by Andy Peters on Aug 19, 2010 5:18 pm • link • report
High density development costs more to supply. Land costs are higher, because higher density by definition means more demand per unit area of land. And construction costs are higher. Low-rise buildings can use inexpensive building materials (wood or block frame covered in stucco or siding), and cheap, low-skilled labor. High-rise buildings require more expensive materials (steel, reinforced concrete, glass), and more expensive labor and construction techniques. The stricter building codes for high-rise buildings due to the greater risks from fire and other hazards further raise construction costs.
by Johnson on Aug 19, 2010 6:26 pm • link • report
"The city, despite its considerable charms, is a bureaucratic abyss." you may not agree with the everything portland does (i dont either) but the city government is more effective and functional that any other government i've seen. portland likes input from its citizens, its citizens like to give their input. if portland is so bad why are young DIY independent people flocking there to open small businesses in formerly run down neighborhoods?
bojack & Jack Bogdanski is known around town as the blog for angry old portlanders who hate any change. think of walt kowalski of gran torino.
we used to have private railroads, transit companies and ferryboat companies. they were profitable and covered all their capital and operating costs and paid taxes including on their property. then the government threw tremendous amounts of public money at building an extensive road system to all corners of the country directly adjacent to existing rail, streetcar and interurban lines and building free public bridges to put out of business private ferries. for awhile the companies held their own, albeit with diminished profits but as more and more public government roads came online with faster roads and a more extensive network, the competition (cars, trucks and buses) with its almost free travel began to take over. not surprisingly the transit, interurban, passenger/freight rail and ferries companies went out of business. given that they were providing essential services and many cities were built around these services, the public found it important to have publicly owned systems to retain much of this. and only after much of these former transit, rail and ferry systems were destroyed did we see how getting around solely by road would not work. i find it odd that libertarians that follow transportation ignore this history because it would seem to support their arguments about government and the market.
it works out to this... subsidize roads, you have to subsidize transit and rail but when you dont subsidize roads, rail and transit can be profitable. of course when you have a trillion dollars worth of road infrastructure built in the last 100 years, its kind of hard to go back to the free market transportation companies. it is worth noting that freight rail is the only one that was able to for the most part survive as private transportation. about the only subsidies it got were some land grants 150 years ago.
by jon on Aug 19, 2010 6:35 pm • link • report
by Johnson on Aug 19, 2010 6:48 pm • link • report
It's possible that there is more demand for a new building in a developed area, but the demand to build up has to be there in the first place to justify the higher construction costs. Rental and condo units are not totally fungible. You're not just adding to the building, and demand is sticky because people are bound into contracts or to the equity of the house. Either way, most rental units in central DC are going for well more than their construction costs. It wouldn't surprise me if the margins are larger for developers too.
Additionally, 5-6 stories can be built still with wood and block structures, and townhouses usually cost only a little more per square foot of land than an equivalently-sized freestanding house. There are actually efficiencies in building townhouses, such as shared load-bearing walls, but I can't say how much they save. Either way, high-density does not have to mean 15-160-story buildings, but anti-urbanists have been exploiting this misconception for years.
by Neil Flanagan on Aug 19, 2010 7:03 pm • link • report
Irrespective, a subsidy is a subsidy. Why should the government subsidize something that consumers already gravitate towards?
by Neil Flanagan on Aug 19, 2010 7:05 pm • link • report
by Rich on Aug 19, 2010 7:24 pm • link • report
Yes it does. Unless you're talking about the density of construction rather than the density of people. Of course, higher construction density without higher people density would mean more empty buildings. The point is that, in general, the more people there are competing for each square foot of land, the higher the unit price of that land. It's economics 101. Higher demand per unit of supply means a higher unit price. Thus, land costs tend to rise with density, so housing costs tend to rise with density.
As for construction costs, I'm not sure what the height limit is for wood- and block-framed buildings, but again, in general, construction costs tend to rise with building height, for the reasons I explained (among others). That is why tall buildings are rare where land is cheap.
by Johnson on Aug 19, 2010 7:28 pm • link • report
I'd be happy to eliminate both road and transit subsidies. The effect on motor vehicle usage would be trivial, because the subsidy is such a tiny share of the total costs of driving. But the effect on transit usage would be huge, because fares would have to skyrocket to pay the costs previously covered by the subsidies.
by Johnson on Aug 19, 2010 7:33 pm • link • report
That's a great point, in the linked post below, 'The Transit Trust Fund: a 21st century solution' from 'August 13, 2010 10:45 am', I challenged the on-going GGW assertion that cars are more heavily subsidized than mass transit. Rock_n_rent took the challenge, did the numbers, and lo and behold:
@Lance:
O.k., I see your point. To put it quantitatively, the government subsidy for roads is about $60 billion per year now. We have 3 trillion vehicle miles traveled per year now, so the government subsidy is around 2 cents per mile. Given that, for income tax purposes, the government accepts an average cost of 50 cents per mile for vehicle travel (similar to AAA calculations of "true cost" of driving), the subsidy amounts to about 4% of total costs.
Per the American Public Transportation Association's 2010 Fact Book (at http://apta.com/resources/statistics/Documents/FactBook/APTA_2010_Fact_Book.pdf), total operating and capital costs for all 7,700 public transit operations in the country were about $54 billion for 2008. Fare collections were about $12 billion, implying a government subsidy of $42 billion, or 78%.
So, there you have it ... cut off transportation subsidies for cars AND for mass transit, and the car drivers will barely feel the pinch, while the mass transit system collapses.
http://greatergreaterwashington.org/post.cgi?id=6763
by Lance on Aug 19, 2010 10:18 pm • link • report
by Andy Peters on Aug 19, 2010 10:33 pm • link • report
I would agree with your general point regarding mass transit, although some of the specific numbers appear a bit off (if you look at the Federal Highway Statistics for 2008, which is the most recent report available, you will see that the road subsidy across all levels of government now amounts to 2 cents rather than 1 cent per vehicle-mile). While this is indeed still trivial compared with the overall costs of owning, maintaining, and operating a private motor vehicle, an increase of 2 cents per vehicle-mile in gasoline excise taxes to eliminate the subsidy would mean an increase of about 34 cents per gallon, given an average fuel efficiency of 17 mpg across the nation's entire fleet. I have my doubts that this would be politically accomplishable in the U.S. today. In addition, the calculation does not take into account that the government subsidy for roads might be so low because the government is systematically underinvesting in new road capacity; an increase in road capacity to match increases in vehicle-miles traveled would require significantly more outlay for roads.
It is also true that, economically speaking, Americans perceive utility in single-vehicle mobility over mass transit. We can even put a rough number on this. The subsidized cost for most mass transit is around 20 cents per passenger-mile; the cost to drivers is around 39 cents per passenger-mile (lower than per vehicle-mile, because the abundance of dating couples and soccer moms means that, on average, a car carries 1.64 people). Thus (ignoring the costs of parking) the vast majority of Americans (about 99%, given nationwide transit mode share of 1%) value the utility of private vehicle travel at at least 19 cents per passenger-mile. I happen to be one of those who do not value this utility at 19 cents per passenger-mile, and thus I see no point in paying more to own a car rather than taking mass transit, but I am obviously in a very small minority.
This economic reasoning also explains why transit mode share is higher in dense urban areas, and highest in New York City. For one thing, NYC has the nation's highest parking rates. Including the cost of parking makes owning a car even more expensive than taking mass transit, and this increased cost differential exceeds the utility of owning a car for more people, so that more people decide that it is better to go with the low-cost option of taking mass transit.
The other cost you have in New York City is congestion-cost - it is simply far more difficult to get anywhere in NYC traffic than it is, say, in small-town Iowa. While some European and Asian cities have gone to explicit congestion pricing to unclog their streets, this is not yet the case in the U.S. (Mayor Bloomberg's efforts in this direction were shot down). But one can argue (as Lance did in an earlier thread this past week) that mass transit subsidies are a form of congestion pricing - private vehicle owners are willing to subsidize mass transit in order to get some drivers off clogged streets.
In fact, I think we can put some economic valuation on congestion cost. In 2008, according to the Public Transportation Fact Book, the total direct subsidy for mass transit was about $42 billion. About $15 billion of that came from the federal gasoline excise taxes. This means the remaining $27 billion came from state and local sources. Given that there were about 55 billion passenger-miles on transit, this means that the state and local subsidy was about 50 cents per passenger-mile. In the New York City urban area, overall mass transit mode share on a passenger-mile basis was right around 10% (this is not the trips-to-work mode share, which is much higher at above 50%). One way of looking at this is to say that all travelers together (including single-vehicle owners) were willing to subsidize mass transit at 5 cents per overall passenger-mile. In other words, the generalized congestion cost for the NYC urban area is 5 cents per passenger-mile (or roughly 8 cents per vehicle-mile). In the Washington urban area, where overall mass transit mode share is only about 4%, this implies a generalized congestion cost of about 2 cents per passenger-mile. And of course, in small-town Iowa, there is no willingness to subsidize mass transit, which implies that there is no congestion cost.
The above analysis of course implies that mass transit subsidies are not some political creation that operates completely outside the economic realm. Instead, I would argue that frequent, contentious, and highly localized political battles that lead to the setting of local and state transit subsidies are a form of market price seeking that leads to the discovery of local cost-utility functions and results in an economically reasonable distribution - in general, the more dense and congested an area is, the more the general population is willing to subsidize mass transit as a form of congestion pricing, thereby monetizing the congestion cost.
by rock_n_rent on Aug 19, 2010 10:54 pm • link • report
Not counting that tiny issue of the cars not being able to move because of the massive spike in traffic.
Look: Cato gets 40% of its money from the oil and gas industry. They love cars. They're going to advocate everything they can in order to support zoning laws and highway policies that make driving an absolute necessity.
They are simply a group that favors a certain lifestyle/identity, one in line with what you find upper middle class suburban whites will prefer.
by Tyro on Aug 19, 2010 10:59 pm • link • report
Hard to say, exactly. Depends on what you count as a "subsidy." But under any reasonable accounting, government subsidies to commercial aviation are also vastly lower per passenger-mile than government subsidies to mass transit -- or railroad (that is, interurban rail - basically, Amtrak), which is the most heavily subsidized mode of transportation of all. See Table 4 at the end of this document to give you an idea of just how heavily subsidized mass transit is compared to motor vehicles and commercial aviation: Federal Subsidies to Passenger Transportation. The most recent report is from 2004, but the subsidy ratios have not changed dramatically since then. As you can see, in 2002 (the most recent year), commercial aviation received about $6 per thousand passenger-miles in federal subsidies, while transit received $150 per thousand passenger-miles.
You can quibble about the exact numbers, but by any serious analysis transit (and Amtrak) users are making out like bandits compared to drivers and fliers.
by Johnson on Aug 19, 2010 11:10 pm • link • report
Exactly. You're proving the point. There is higher unmet demand for densely built urban areas than there is for suburban areas. If enough dense city blocks were built so that all the people who marginally prefer to live in higher density cities were accommodated then prices would tend to equalize between urban and suburban. But since there isn't enough to go around the people who want it more are willing to pay more.
And it's important to note that this doesn't mean more people prefer urban than suburban just that there is some demand that isn't being met, which suggests we should build more.
by Mike on Aug 19, 2010 11:14 pm • link • report
As to your point about the high cost of housing in dense urban areas, I believe this is lacking in economic analysis.
As you should have learned in Econ 101, price is set not just by the supply curve, but by the intersection of supply and demand curves. Sure, it may cost more to build in dense areas, but the fact that there are still people who are willing to pay more to live in dense downtown rather than cheap sprawling suburbia means that there are still quite a few people for whom urban living has an increased utility when compared to suburban living. If this were not the case, then people would keep moving out of the city until urban house prices equaled suburban house prices -at which point, given your supply considerations, urban density would be the same as suburban density.
So, the fact that at equilibrium urban house prices tend to be on average higher than suburban house prices clearly indicates that, overall, there is an economic utility to urban living for which Americans are willing to pay.
Suburban flight at a time of high urban house prices does NOT mean that people value suburban living over urban living. It only means that the higher cost of urban living outweighs the utility from living in the city (just as, once you make driving more expensive relative to mass transit, more people will shift to mass transit even though, in general, everyone values the utility of private-vehicle driving much more highly than that of mass transit).
The fact that, during the mid-to-late 20th century, more people moved to suburbia than to urban areas (and, I presume, as a result urban house prices at equilibrium decreased relative to suburban house prices - housing prices in the burnt-out urban ghettoes after 1968 were probably quite low), suggests that the relative utility of urban living decreased during this time (urban living may still have been more desirable than suburban living, but not by as much, so that more people were willing to go to cheaper suburban areas, which as a result became less cheap).
Since about 2006, equilibrium prices in the exurbs have dropped considerably, while equilibrium prices in core cities have stayed steady or even increased (see "gentrification"). This means not only that Americans perceive higher utility in urban living, but that the perceived relative utility has increased.
This all is qualitative reasoning that should be evident from Econ 101. Extracting quantitative information about the relative utility of urban versus suburban living and its changes over times will probably require advanced econometrics.
by rock_n_rent on Aug 19, 2010 11:19 pm • link • report
Your back-of-an-envelope calculation of road subsidies is in the ballpark, but scholarly estimates of the subsidy are closer to 20 cents per gallon (about 1 cent per vehicle-mile on average for the passenger car and light truck fleet). See, for example, this study from the Institute of Transportation Studies at UC Davis. (Since the average occupancy of passenger cars and light trucks is about 1.6 passengers, the per-passenger-mile subsidy is even lower -- about 0.7 cents per passenger-mile).
But the exact number doesn't really matter. The important point, as you seem to agree, is that, per unit of transportation provided, subsidies to drivers are vastly lower than subsidies to mass transit users. Government subsidies distort the urban transportation market in favor of mass transit, and against driving, not the reverse. It is also worth noting that the flat fare structure of most mass transit systems also encourages overconsumption. If fares were higher for longer trips, mass transit usage would probably decline.
As for the claim that mass transit subsidies are justified on the grounds that transit relieves unpriced road congestion, I have never seen any serious analysis to support this claim. Most roads are rarely, if ever, congested. There is little evidence that transit has a significant effect on road congestion, except during rush hour on a few dense travel routes in a few dense urban areas.
">This study found that even the New York subway costs more in public subsidies than it produces in congestion-relief benefits. It seems to me virtually inconceivable that mass transit services in the vast majority of the country could produce congestion-relief benefits even remotely close to their costs in public subsidies.
by Johnson on Aug 20, 2010 12:02 am • link • report
I think you missed the point. The higher price of high-density housing compared to low-density housing does not imply that there is higher unmet demand for high-density housing than for low-density housing (or, indeed, that there is any unmet demand for either type). The market clearing price -- that is, the price at which supply and demand are in balance -- of high-density housing is higher because high-density housing costs more to supply as a result of higher land costs and higher construction costs.
by Johnson on Aug 20, 2010 12:13 am • link • report
Yes, of course there is still a substantial market for dense urban housing. But there is a much larger market for sprawling suburban housing. The suburbs have been gaining population share at the expense of dense urban areas for decades. And there is no serious evidence of any dramatic change in this long-standing trend, let alone a reversal. The trend has slowed somewhat in recent years as a result of the housing bubble and the recession, but the suburbs are still gowing faster than the cities. See the latest Brookings Institute report here: The State of Metropolitan America
If this were not the case, then people would keep moving out of the city until urban house prices equaled suburban house prices
No, that doesn't follow at all. The market floor for dense urban housing prices is the cost of supply, not the price of suburban housing. If demand for dense urban housing falls below the level needed to cover the cost of supply, builders will simply stop supplying it.
by Johnson on Aug 20, 2010 12:32 am • link • report
Any why exactly do you think that there are higher land costs? Because people will pay more to be close to amenities and work.
by Brian White on Aug 20, 2010 12:36 am • link • report
The suburbs are a result of government subsidization, starting with the GI Bills after WWII, which could only be used for new housing. In a system where government subsidies make the playing field lopsided, you can read little into the choices made in the market.
by Brian White on Aug 20, 2010 12:39 am • link • report
http://real.wharton.upenn.edu/~sinai/papers/urban_housing_demand_030807.pdf
"On net, residence patterns suggest that most people want to live in or near cities,
and that desire is increasing over time. In fact, by 1999, 75 per cent of US households lived in
cities (Rosenthal and Strange, 2003). Today, urban America is where housing demand is most
likely to exceed housing supply and generate rising house prices"
...
"Since the 1960s, the patterns of where people live have begun to shift back to cities, even
though people are now less likely to work in the downtown areas."
http://www.theatlantic.com/magazine/archive/2010/06/here-comes-the-neighborhood/8093/
"Last year housing starts were lower by half than in any year since 1959....But housing hasn’t cratered everywhere. According to Stan Humphries, the chief economist of Zillow, an online housing-research firm, if you plot changes in home values within a typical metro region on a satellite map, the result “looks like an archery target, with the outlying areas having experienced substantially higher total declines in home values” than areas closer to the central city. "
Seriously, where are you getting your information? Everything you are saying seems to be wrong.
by Brian White on Aug 20, 2010 12:46 am • link • report
I think you've misunderstood your own source. The population is becoming more "urban" in the sense of urban vs. rural, but not in the sense of high-density cities vs. low-density suburbs. As used by government agencies and in academic reports, "urban" generally means merely "not rural." It refers to urban development in general, which includes traditional dense urban cores, suburbs and exurbs. The population is certainly becoming more urban in this sense (i.e., less rural), but it is also becoming more suburban and exurban.
See the Brookings report I cited earlier for the data. Here's a relevant quote from the Executive Summary:
by Johnson on Aug 20, 2010 1:06 am • link • report
I already explained this. Higher (population) density by definition means more people per unit area of land. The more people there are competing for each square foot of land, the higher the price of each square foot is likely to be. Higher demand per unit of supply means a higher unit price.
The suburbs are a result of government subsidization, starting with the GI Bills after WWII, which could only be used for new housing.
Huh? What GI Bills that could only be used for housing? And how would that be a subsidy of suburbs, anyway? A housing subsidy could be used for housing either in a suburb, or in a dense city, or in a small town, or somewhere else.
The huge growth of suburbs and the relative decline of dense urban cores over the past 60 years or more is not limited to the United States. It's happened in virtually every industrialized democracy throughout the world. As incomes have risen and cars have become more affordable, people throughout the developed world have gravitated increasingly towards car-based suburban lifestyles.
by Johnson on Aug 20, 2010 1:25 am • link • report
by SEO Traffic Spider on Aug 20, 2010 2:28 am • link • report
What are your numbers? Where are they from? Do they account for only federal dollars spent, or also local dollars spent on roadways? Or is it just the amount of income taxed under the 16th amendment. Do state or local income and property taxes figure?
And what industrialized culture are you speaking about in 1:25? Are these places where the government has subsidized highways, like France? You're asserting truths in countries more socialized than ours.
by Neil Flanagan on Aug 20, 2010 2:46 am • link • report
by Neil Flanagan on Aug 20, 2010 2:53 am • link • report
I want to thank you for providing the links, and especially that to Mark Delucchi's study. I am quite aware that I'm merely an intellectually overreaching accountant wading into waters that have already been examined at length by the professionals, and so I've been looking for an entry into the peer-reviewed scientific literature (as opposed to think tank/lobby group opinion pieces).
I also note that Delucchi's study is based on 2002 data. By doing my back-of-the-envelope calculation using 2002 FHWA highway statistics, I also come up with about 22 cents per gallon ($36 billion subsidy over 166 billion gallons). In the years since 2002, the road subsidy by the crude FHWA measure appears to have generally been in the $40 - $45 billion per year range, but then jumped to $60 billion in 2008, mostly because state capital spending for highways jumped by $10 billion in 2008 (and this was even before the serious funding of "shovel-ready" projects with the 2009 stimulus bill). See Table HF-10 in each year's statistics at http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.cfm. It will be interesting to see how Delucchi's analysis will be revised over time.
by rock_n_rent on Aug 20, 2010 6:13 am • link • report
They ignore all evidence that shows that the auto-centric and sprawl-oriented lifestyle in the U.S. is massively subsidized, regardless of how they try to twist the figures to show otherwise.
by Steven P. on Aug 20, 2010 9:07 am • link • report
Sorry, but no. DC had similar population density in the 1980s. And land prices and housing prices were dramatically lower then, when adjusted for inflation. People didn't want to live in the murder capital, so they paid less to do so. Now people do want to live in DC, so they pay more.
"Huh? What GI Bills that could only be used for housing? And how would that be a subsidy of suburbs, anyway? A housing subsidy could be used for housing either in a suburb, or in a dense city, or in a small town, or somewhere else."
No, the housing part of the bill could only be used for buying a new house, not an existing row house. Levittown and imitators sprung up to supply the demand for brand new houses bought with government loans.
by Brian White on Aug 20, 2010 10:29 am • link • report
by JustMe on Aug 20, 2010 11:41 am • link • report
I'm not sure how much it is costing tax-payers to subsidize these roads vs transit but my experience tells me that if we created denser areas for people to live a good quality of life and not have to sit in a car all day long then it is worth it. Cars just eat up money, time, and your well-being.
Otherwise, lets all just buy an RV and drive around all day long...just get it over with.
by Tom in DC on Aug 20, 2010 12:11 pm • link • report
Houston still has very bad parking regulations that many developers have cited as one of the primary reason we have not had as much transit-oriented development along light rail as was expected.
The City Planning department is finally going through an assessment of the parking policy and is expected to change the rules by December, but it is clear from their expectations that they will not move more toward a free market in general, but instead carve out special districts which will be allowed to develop their own parking plans, while most of the city will continue to require suburban parking.
Also, by the way, there are all kinds of other codes in Houston that make urbanism illegal. There's a report by There's a famous case where CVS build a suburban style store surrounded by parking in the heart of Midtown a couple blocks from light rail. The Mayor tried to fight the plans and failed, as they were in fact following the city code exactly, whereas the Post Apartments complex built across the street, which is Houston's best recent mixed use transit oriented development, had to go through a long series of variance requests because walkable urbanism is illegal in Houston.
Michael Lewyn's paper "How overregulation creates sprawl (even in a city without zoning) argues that Houston's Chapter 42 development codes are in fact more cumbersome than traditional zoning.
So don't listen to Randall O'Toole when he tells you that Houston is a magical free market. Its a lie. And many of us in Houston hate that he keeps misrepresenting Houston.
by Jay Blazek Crossley on Aug 20, 2010 12:37 pm • link • report
As I said, virtually every industrialized democracy in the world. Western Europe has been sprawling and suburbanizing dramatically since the end of World War II, just like the U.S. Europe's dense inner cities have either grown much more slowly than the surrounding suburbs, or have actually lost population. Mass transit usage in Europe has essentially remained flat, while driving has skyrocketed. Ditto for Canada, Australia, etc. The growth of suburbs and sprawl is not an artifact of the GI Bill, or U.S. zoning laws, or any other particular set of public policies. It happens whenever a free society becomes wealthy enough for mass ownership of private automobiles. In some countries (e.g., Singapore, Japan, Netherlands), sprawl is severely constrained by the lack of land, but whereever land is available and most people can afford cars, sprawl seems to be the virtually inevitable outcome.
by Johnson on Aug 20, 2010 12:42 pm • link • report
Then its beneficiaries could have used it to buy new high-density housing instead of new low-density housing, if that is what they had wanted. I'm not sure why you have so much trouble understanding why the suburbs are more attractive to most people than dense inner cities. Including to young men returning home from the war, getting married, settling down, and starting a family.
by Johnson on Aug 20, 2010 12:51 pm • link • report
by bing!bing! on Aug 20, 2010 1:00 pm • link • report
by Nat on Aug 20, 2010 1:45 pm • link • report
I would argue that there are significant differences in both mass transit usage and suburbanization between at least Central Europe and the U.S. (as a bit of background, I grew up in West Germany in the 1970s; the townhouse I grew up in, which back then was at the edge of town - a knife-edge separation between residential and agricultural areas - is still at the edge of town today).
To control for the effects of density, let's pick areas that are matched on a population density basis. Some time ago I was playing with some numbers, and discovered that the German state of Hessen (which is where Frankfurt is located) is almost the exact same size (on a land-area basis) as Massachusetts, with Massachusetts having slightly more inhabitants and thus a slightly higher population density. Moreover, state GDP per capita for Hessen and Massachusetts are within 10% of each other, so economically they are also well matched. But whereas overall mass transit mode share in Massachusetts, on a passenger-mile basis, in 2008 was only about 3% (which in itself is three times better than the U.S. average), in Hessen it was around 30%!
Some details on this: in Massachusetts in 2008, there were 5.4 million cars, buses, and trucks registered; at the same time, in Hessen there were 3.4 million such vehicles (one cannot compare just cars to cars, because in the U.S. SUVs, pickup trucks, and the like are counted as trucks rather than as cars, while in Germany they are included in "person vehicles" as opposed to "goods vehicles"). Moreover, in Massachusetts, each vehicle traveled about 10,000 miles per year, while in Hessen they only traveled about 6,000 miles per year.
On the other hand, in Hessen, commuter rail and subways transported 1 billion passengers for 17 billion passenger-miles, while in Massachusetts, commuter rail and MBTA heavy rail only transported 200 million passengers for 1.4 billion passenger-miles. Use of buses (once you include school buses in the U.S., as school transportation in Germany is handled using public transit rather than a separate bus system) and of streetcars is similar between the two states, so the main difference in mass transit use is the use of rail for relatively long-distance (average 16 miles per unlinked trip) travel - I presume mostly work commuting.
Of course cars are still the dominant mode in Germany - but what these data show is that they are much less dominant than in the U.S., even in the densely populated and land-poor "megalopolis" region of the Northeast.
As to suburbanization, yes, of course there was massive flow out of dense inner cities into suburbs in Germany after WWII (although this seems to have slowed down significantly over the past several decades, as overall population growth has nearly halted). But I would argue that suburbanization has been far more "clumpy" in Germany than here.
As a random example, take Pinneberg, which is about 10 miles from downtown Hamburg, and which grew from about 13,000 inhabitants in 1939 to about 42,000 in 2008 (for these figures I am relying on the German wikipedia site). The town as a legal entity has a size of 8.34 square miles, so population density is around 5,000 per square mile. German census data are wonderful, in that they also give you detailed land-use data, that is, how much area in each town or township is used for agriculture, woods, water, and what is called "settlement area", which includes houses, roads, parking lots, commercial areas, and cemeteries. In Pinneberg, the settlement area is 4.78 square miles, which means that 43% of the town's area is open space (which you can easily verify by looking at Pinneberg, Germany in maps.google and admiring all those fields coming right up to the edge of dense neighborhoods - and which is typical of suburban areas and towns across Germany). It also means that, in terms of actual built-up area, the population density is close to 9,000 per square mile.
Contrast this with, say, Falls Church, VA. This town grew from around 2,500 inhabitants in 1940 to slightly more than 11,000 today, and thus experienced very similar post-WWII population growth. Overall population density is also right around 5,000 per square mile. Unfortunately, U.S. census data do not give you township-level information on land use, but I would dare say that there is very little open space left in the Town of Falls Church, so that the built-up area also has a density of only slightly more than 5,000 per square mile.
So the big difference here is that, in the U.S., suburban development entailed the building of single-family homes on substantial plots of land, which each home having its own large garden. In contrast, suburban development in Germany has involved much more clustered development, with more townhouse developments and few or small individual gardens, leaving much more contiguous open space surrounding the towns. The closest to this paradigm I can think of in the U.S. is the Central Valley in California, where you also find rather dense neighborhoods with small yards, and then a sharp edge to agricultural land (see, for example, Davis, CA). Very different from what is happening, for example, in Loudoun County today.
I believe this is generalizable - that is, if you spend enough time on Google Maps looking at German towns compared to American suburbs, I think you will get a clear intuitive feel that residential patterns are far more clumpy in Germany (dense townhouse-style developments separated from large open space, with no 1-house-per-acre or, God forbid, 10-house-per-acre districts) than in the U.S., whose suburbs are far more diffuse and, in a word, "sprawling". Unfortunately, given the limits of U.S. census data collection, I can think of no easy way of quantifying this
The clustering or clumping of development of course creates more walkability and makes transit (especially suburban rail transit) so much more feasible in Germany - in effect, it is transit-oriented suburban development. And I am naturally very curious why it turned out this way in Germany - whether gas prices in Germany have always been high, so as to shape development in this direction, or whether mortgage policies were different, or whether planning and zoning boards had different attitudes there - or all of these.
by rock_n_rent on Aug 20, 2010 2:50 pm • link • report
It doesn't make much sense to try to evaluate differences between the transportation systems of the U.S. and Europe by cherry-picking pairs of cities or regions. There are all sorts of factors that may influence modal shares in a particular city or region.
The overall situation in Europe is described in Eurostat's Panorama of Transport. Yes, mass transit has a larger share of the transportation market in Europe than in the U.S. But it's still a tiny (and shrinking) share. Yes, urban development in Europe is not as sprawly and car-oriented as in the U.S. But the trend in Europe is still overwhelmingly in favor of low-density, car-oriented development. In fact, Europe is catching up with the U.S. The rate of growth in car ownership is much higher in Europe than in the U.S. Europe's motorway network is also growing much faster than the U.S. interstate/freeway network. In 1950, public transportation provided more transportation in Europe than private automobiles - a few hundred billion passenger-km. By 2000, use of public transportation in Europe had barely grown at all, but transportation by private automobile had skyrocketed to more than 3.5 TRILLION passenger-km. American proponents of urbanism generally seem to have a grossly exaggerated view of the importance of mass transit in Europe's transportation system.
Europe probably won't become quite as sprawly and car-oriented as the U.S. in the foreseeable future. But that's a reflection of resource constraints and historical contingencies, not consumer preferences. Most of Europe's central cities were laid out long before the rise of the automobile, and cannot realistically be adapted to car-based lifestyles. So they have to rely on mass transit to a much greater extent than most American cities. Incomes are still lower in Europe than in the U.S., and cars are still not as affordable to Europeans as they are to Americans. And land is much scarcer in Europe than in the U.S., so there are greater economic incentives to conserve it.
by Johnson on Aug 20, 2010 4:17 pm • link • report
This seems to be a matter of definition. You are defining walkable, transit oriented, denser than normal exurbs as suburban. So people moving from total sprawl to dense exurbs don't count in your book. I don't know why. When I lived in silver spring in a condo building on a street with rows of condo buildings, the density was far higher than it is in my capitol hill neighborhood.
by Brian White on Aug 20, 2010 4:39 pm • link • report
by bing!bing! on Aug 20, 2010 4:44 pm • link • report
by bing!bing! on Aug 20, 2010 4:48 pm • link • report
Thank you for the link to the Eurostat report, which makes for fascinating reading. However, I would draw somewhat different conclusions from it than you do.
From tables 4.71 through 4.78, one can calculate that the land surface transportation mode share (i.e., excluding ships and airplanes) for mass transit decreased from 19.2% in 1995 to 17.7% in 2006. But for the EU-15 (i.e., those western European countries that were already members of the EU in the 1990s), mass transit mode share remained stable - 16.6% in 1995 and 16.7% in 2006. For the 12 new countries that joined in the 2000s (all formerly Communist eastern European countries, except for Malta and Cyprus), mass transit mode share fell from 39.3% to 24.6%, and the corresponding car mode share increased from 60.7% to 75.4%.
What this means is that the increase in European mode share for cars is entirely attributable to eastern European countries finally undergoing the mass motorization that western Europe underwent from the 1950s to the 1980s. In western Europe, mass transit mode share has stabilized at 16% - even at a time of increasing incomes. So it is no longer declining - and as to whether 16% is "tiny", I don't know, but it is much larger than the 1% mode share mass transit has in the U.S.
Of course the U.S. has much more land than Europe. That is why we need to control the comparison for population density (as you repeatedly argued, land values are related to population density). This is not cherry picking, but rather making sure we are comparing apples to apples. The densest portion of the U.S. is the Northeast. Take the 13 states from Maine to Virginia (including West Virginia), and throw in DC. This region is about the same size and has the same density as France, and it also has roughly the same population density as the EU-15 as a whole. But whereas mass transit mode share for France, as well as for the EU-15, is right around 16%, for the Northeast U.S. it is only around 3% (maybe 5% if you include school bus transportation, which is not available to the general public) - at the same population density!
So western Europe is not converging on U.S.-style motorization or suburbanization, even when controlling for population densities. Of course there are historical contingencies. As to settlement patterns, I would argue the critical difference is not the density of inner cities (there was plenty of urban renewal in 1960s western Europe, especially bombed-out Germany), but rather the much more compact nature of the suburbs that were built after WWII. Even more important, I believe, is the difference in gas taxes between the U.S. and western Europe
A very interesting theoretical framework is provided by Hammar et al. in The Energy Journal, 25 (2004), no. 3, pp. 1-17 (referred to by Delucchi). As they argue, it is well known that gasoline consumption (and thus car mode share) responds to gasoline prices. But as they point out, gas prices (through gas taxes) are also a function of gas consumption. In those countries, such as the U.S. and Australia, where gas consumption per capita is very high, so that people have arranged their lives and their built environment around cars, there is great political resistance to gas taxes, which keeps gas prices low and thus reinforces high consumption. In contrast, in countries where gas consumption per capita is low, and people have alternatives to car travel, there is less resistance to high gas taxes (hey, in Germany the governments that jacked up gas taxes in both the early 90s and the early 2000s got reelected), which keeps gas prices high, thereby keeping gas consumption low. In Germany, gas taxes (often in the form of import duties) have been very high since the Great Depression, and the resulting high gas prices may very well have led Germans to design their post-WWII suburbs in a much more compact fashion than the Americans, who underwent the first stage of mass motorization even before fuel taxes were implemented in the 1920s and have kept the price of gas low ever since.
This means that the 16% mode share for mass transit in western Europe is not just stable, but also a valid expression (through political economy as well as the markets) of European consumer preferences.
by rock_n_rent on Aug 21, 2010 4:33 pm • link • report
So it is no longer declining - and as to whether 16% is "tiny", I don't know, but it is much larger than the 1% mode share mass transit has in the U.S.
Your comparison is specious. The set of tables you refer to include both intracity transportation ("tram & metro" and transit buses) and intercity transportation (railways and intercity buses, or "coaches" as they usually referred to in Europe). You cannot meaningfully compare this data to the U.S. without including air travel, because air travel accounts for so much intercity travel in the U.S.
So your comparison is apples-to-oranges. To meaningfully compare the relative shares of "mass transit" in the US and Europe, you need to isolate intracity travel only. Unfortunately, Eurostat does not disaggregate travel data for transit buses and "coaches." But for the urban rail transit category ("tram & metro"), the difference in shares between the US and Europe is tiny, about 1 percentage point. See Table 4.65.
I'm not sure why you picked the period 1995-2006 rather than the longer period 1990-2006. The longer period provides a better indicator of the long-term trend. And that trend shows a decline in all modes of public transportation relative to passenger cars, in both the EU15 and the EU27. For the EU15, travel by passenger cars grew at an average annual rate of 1.7%. This is a higher rate of growth than for any of the public modes. Buses and coaches grew at only 0.8%, railways at only 1.4%, and tram & metro at only 1.4%. The other indicators I mentioned (growth of motorway network, decline of rail network, growth in the number of passenger cars) also indicate a further shift towards car travel.
You're right that the shift towards cars has been greater in the newer, poorer EU member states. That illustrates another trend - as countries become richer, their transportation systems tend to shift increasingly towards cars. There's no reason to think this trend has stopped in either the EU27 or the EU15.
by Johnson on Aug 21, 2010 5:54 pm • link • report
Of course the U.S. has much more land than Europe. That is why we need to control the comparison for population density (as you repeatedly argued, land values are related to population density). This is not cherry picking, but rather making sure we are comparing apples to apples. The densest portion of the U.S. is the Northeast. Take the 13 states from Maine to Virginia (including West Virginia), and throw in DC. This region is about the same size and has the same density as France, and it also has roughly the same population density as the EU-15 as a whole. But whereas mass transit mode share for France, as well as for the EU-15, is right around 16%, for the Northeast U.S. it is only around 3% (maybe 5% if you include school bus transportation, which is not available to the general public) - at the same population density! So western Europe is not converging on U.S.-style motorization or suburbanization, even when controlling for population densities.
I've already explained why your "mass transit mode share" comparison is specious. You're mixing together long-distance and short-distance travel but omitting the largest non-car long-distance travel mode in the US (air). But a further problem with your comparison here is that you cannot meaningfully compare a region with a nation simply because they have approximately the same size and population density. For example, I suspect that the northeast US devotes a smaller share of its land to farming and agriculture than France, because the northeast is part of a nation with vast areas of agricultural land in other states. That means a larger share of land in the northeast US is available for urban development. That means urban development in the northeast US is likely to be lower density than urban development in France. And that means transportation in the northeast is likely to be more car-oriented than transportation in France. And that's just one relevant kind of difference.
Of course there are historical contingencies. As to settlement patterns, I would argue the critical difference is not the density of inner cities (there was plenty of urban renewal in 1960s western Europe, especially bombed-out Germany), but rather the much more compact nature of the suburbs that were built after WWII.
The basic spatial layout of virtually all major European central cities was established long before the mass affordability of passenger cars in Europe. Yes, there was extensive rebuilding in many cities after the war. But they couldn't realistically rebuild London or Paris or Rome or Berlin to look like Houston or Phoenix or Atlanta. European cities generally have narrow streets, dense construction, and little land available for parking. This means they have to rely on mass transit for transportation to a much greater degree than America's newer, more spacious cities. And this helps explain mass transit's (modestly) higher share of the transportation market in Europe compared to the US. It's not that Europeans prefer mass transit; it's that they don't have a choice but to use mass transit more if they are to retain the historical character and features of their major cities. New development in Europe -- new suburbs and exurbs on previously undeveloped land -- is generally much lower density and much more car-oriented. European land-use and transportation patterns increasingly resemble those of the US.
by Johnson on Aug 21, 2010 6:55 pm • link • report
You mention Australia. Land use and transportation patterns in Canada and Australia resemble those of the US much more than those of Europe. This again illustrates the central role of land availability. Oz and Canada have lots of land. So land is cheap and there is little economic incentive to conserve it. So urban development tends to be low-density and car-oriented, just as it is in the US. Also like the US, and unlike Europe, urban development in Australia and Canada is largely unconstrained by historical contingencies. Again, the clear implication of this is that Europe's denser and more transit-oriented urban forms are not the result of a preference for smaller, denser housing or of a preference for getting around on buses and trains rather than driving, but are instead the result of unchosen historical and geographical constraints. If Europeans had as much land as we do, and if their cities were not constrained by layouts established in the age of horse-drawn vehicles, they would probably do as much of their traveling by car as we do, and their houses would be as big and sprawly as ours.
by Johnson on Aug 21, 2010 7:30 pm • link • report
Even O'Toole has said roads are there regardless of economic conditions.
by Andrew Dawson on Aug 22, 2010 6:25 pm • link • report
Agree that Eurostat data are not good for isolating local mass transit. Not only bus/coaches have to be split up, but railways also include commuter rail (per detailed German stats, the German railway number is about 55% "Nahverkehr" or local commuter rail), and of course car travel includes long-distance as well as local.
Per Table 4.64, Europeans use "domestic" (intra-EU) airtravel almost as much as Americans (9% vs 11%). Air travel is by far most rapidly increasing mode, not just in Europe but also U.S. and Australia. Cannot demonstrate it for Europe with Eurostat data, but in U.S. and Australia air travel competes with long-distance car travel as well as rail/coach; if you include air, then car mode share is actually decreasing.
May I suggest we look at the Australian Transport Statistics Yearbook 2009 (http://www.bitre.gov.au/publications/10/Files/BITRE_TRANSPORT_STATS_YEARBOOK_2009.pdf)? As you say, Australians have even more land than Americans, and their cities are even younger (no Boston or Philadelphia to bother with).
Table 5.1 presents total passenger transport (local and intercity). Including all modes, air mode share has increased from 3.9% in 1970 to 15.1% in 2007. Car (including "other", which is mostly nonbusiness use of trucks) was at 82.1% in 1970, increased to 86.6% in 1978, and since then has slid to 76.0% in 2007. Bus and rail mode share went from 14.0% in 1970 to 8.6% in 1978, and since then has bounced around between 8.4% and 10.9%; in 2007, it was at 8.9%.
Excluding air travel and looking only at surface transportation, bus and rail mode share dropped from 14.6% in 1970 to 9.0% in 1978, and since then has bounced around at generally slightly higher levels; in 2007, it was at 10.5%.
This all during a time when real GDP per capita was increasing around 2-3% per year in Australia. So once a certain income plateau was reached in 1978, it appears Australians decided not to put their income growth toward shifting modes even more toward cars.
Comparable U.S. data can be obtained from the Bureau of Transport Statistics; unfortunately, consistent data go back only to 1985. Air travel mode share increased from 8.7% in 1985 to 10.6% in 2008, while car mode share (including motorcycles and truck passenger-miles) went from 87.2% to 85.6% and transit mode share went from 4.2% to 3.8%. This latter figure includes not only the pathetically small contribution of Amtrak, but also what are now about 150 billion passenger-miles of "bus" not captured in APTA data. This likely includes Greyhound/Trailways, but I believe is mostly school-bus travel. Excluding this line item, transit mode share went from 1.33% in 1985 to 1.06% in 1994, and has remained fairly constant since then, reaching 1.09% in 2008 (these figures are consistent with other estimates I've seen of transit mode share).
Excluding air travel, bus and rail mode share of surface transport went from 4.6% in 1985 to 4.3% in 2008; bus and rail excluding non-APTA buses went from 1.5% to 1.2% over that time.
So, if you include school buses, nationwide transit mode share in Australia has stabilized at a level more than twice as high as in the U.S.; if you exclude school buses (and assume Greyhound/Trailways travel is of the same order of magnitude as Amtrak), Australian transit available to the general public has stabilized at a mode share about 4-5 times as high as in the U.S. And the Australian mode share (including intercity as well as local bus/rail, which means we have comparability to Eurostat data) is only about 40% lower than the level at which it has stabilized in the EU-15 over the past decade.
Table 5.3 shows passenger-kilometers for various modes for metropolitan travel, which would thus be equal to local transit. For all Australian capital cities taken together, the bus/rail/light rail mode share in 2007 was 10.5%, not much different from its mode share of 10.8% in 1978. The only American urban area with this kind of transit mode share is the New York City urbanized area, which of course has a population density twice as high as anything in Australia. If you look at regions like Washington or Boston with sizes and densities similar to Sydney or Melbourne, public transit mode share is more in the 3-4% range. Thus, again, it appears that, in the Australian metropolitan regions, public transit has stabilized over the past 3 decades at a mode share about 2.5 times higher than in the U.S.
Hey, I'm not disputing that cars are the dominant mode. Any country that grows through a per-capita GDP level of around $10-20,000 in todays dollars will undergo mass motorization - the U.S. did from the 20s to the 50s; Western Europe in the 50s and 60s; Japan in the 60s and 70s; Eastern Europe and South Korea over the past 2 decades (and not quite finished in Eastern Europe); China is undergoing the process now. Anyone who thinks you can as a society remain car-less, or have, say, 1% or 10% car mode share in a developed economy while we still have appreciable global production and reserves of liquid fuels, is not being realistic.
But the motorization process can be described with a logistic curve - at a certain point, it plateaus at some level; it never displaces all other modes; there will always be at least some minimal demand for other modes (hey, there are still tens of thousands of horse-and-buggy users in the U.S. among the Old Order Amish). What I am interested in is this plateau level, and thus what the stable or equilibrium mode share for bus/rail is. I think we have already argued enough about the European case, and won't come to an agreement on that. But the Australian case intrigues me - how have the Aussies managed to maintain a stable transit mode-share 2-3 times higher than in the U.S. despite having suburbs as sprawly as ours? Is their unconstrained consumer preference simply different than ours?
by rock_n_rent on Aug 22, 2010 9:52 pm • link • report
In your latest blizzard of numbers you seem to have lost sight of these basic points:
1. In virtually every wealthy democracy the growth in car travel has vastly exceeded the growth in bus and rail travel, indicating the vast superiority of cars over buses and trains.
2. The relatively small differences in the mode shares of car travel between the US and other industrialized democracies are easily explained by geographical, historical and economic differences. In particular, the shift to cars in Europe has been significantly constrained by the lack of available land for urban sprawl, and the dense historical layouts of Europe's central cities.
3. Your claim of a 16% vs 1% ratio of "mass transit" mode shares in Europe vs the US is completely bogus, for the reasons I explained in my previous comments. With respect to urban rail transit, the difference in mode shares is tiny -- about 1 percentage point. For bus transit, the difference is almost certainly similarly small, although the Eurostat data make direct comparisons of urban bus travel difficult.
Regarding your claim that the long-term decline in the mode shares of buses and trains has "stabilized" in Europe (or the US, or Australia, or anywhere else) there is no basis for this assumption. Between 1998 and 2008, the real price of crude oil quadrupled. Bus and train users were largely insulated from this price increase by government subsidies. But drivers suffered the effects in the form of a huge increase in gas prices. It is hardly surprising that this massive increase in the price of gasoline slowed the shift from buses and trains to cars over that period. Since 2008, oil prices have fallen. This has reduced the economic incentive to use mass transit instead of driving. In the US, the effect of the earlier runup in oil prices, combined with the loss of tax revenues due to the recession, has finally rippled through to the mass transit system, as transit agencies have been forced to raise fares and cut services to cope with the dramatic increase in costs. This has led to a substantial decline in mass transit ridership.
by Johnson on Aug 23, 2010 3:41 am • link • report
by Brian White on Aug 23, 2010 7:51 pm • link • report
Add a Comment