I tend to emphasize that while walkable living is probably thinning, one shouldn’t underestimate the extent to which the correlation between obesity and driving is a result of self-selection of the obese into auto-oriented lifestyles. But hey, walkable living is probably thinning:

Increasing the availability of public transit systems is one among a number of modifications to the built environment that offers opportunities for increasing physical activity and reducing the prevalence of obesity and its associated problems. In a study published in the August issue of the American Journal of Preventive Medicine, researchers from the University of Pennsylvania, Drexel University and the RAND Corporation found that construction of a light-rail system (LRT) resulted in increased physical activity (walking) and subsequent weight loss by people served by the LRT. These findings suggest that improving neighborhood environments and increasing the public’s use of LRT systems could improve health outcomes and potentially impact millions of individuals.

That’s via Freakonomics. I don’t support improvements in urban policy because they’re healthy; I support them because they make good economic sense. But for people making decisions about where to live, this is a data point worth keeping in mind.

Tyler Cowen asks why there are so few cheap restaurants in Anacostia. By Anacostia, we are to assume that he means the neighborhoods east of the river. It’s annoying that he failed to make this distinction, but let’s set that aside and move on.

Wards 7 and 8 are not, for now, big destination areas. They’re not significant employment centers, and they’re not home to major retail attractions. Retail demand therefore comes overwhelmingly from the residential population. And the residential demographics tell us a great deal about the question Tyler asks.

Wards 7 and 8 are the poorest in the city. Somewhere between a third and a half of residents in those areas are on food stamps (for lots of great information on the District, visit here). A large share of the population is very poor and carless and simply doesn’t do much dining in sit-down establishments. Another large share of the population is less poor and mobile — car-reliant. For this group, existing retail concentrations across the border in Maryland’s Prince George’s County are close and often more accessible, and because they’re well established, they offer a large and diverse array of dining and consumption options that would be difficult for any individual enterprising restaurant to compete against.

As an example, consider — a couple living in a single-family neighborhood in eastern DC and thinking about driving to a restaurant can simply go a mile or two east and find several shopping centers with multiple dining options, some with movie theaters, all with other shops, and so on. These concentrations allow potential diners to choose what they want when they get there, adjust on the fly if one place is too busy, and do complementary activities like shopping and movie watching. A restaurateur opening a place in a District neighborhood, can offer only his own establishment. A retail ecosystem has yet to emerge. Now, shopping centers are hardly organic retail ecosystems, but they don’t have to be; they’re usually controlled by a single landlord that can price spaces to attract a complementary array of businesses. In District neighborhoods, there is no such entity.

The neighborhoods east of the river are also home to the young professionals that populate every District neighborhood, but walkable developments with rich retail options take time to emerge. The population density of the professional class east of the river is not yet high enough to support a critical mass of new establishments. This has also been true of other neighborhoods in DC that have been more successful, but the contrast is telling. H Street emerged as a destination thanks largely to the efforts of a single entrepreneur, Joe Englert, who coordinated with other entities and arranged to open a cluster of establishments all at once to provide the kernel around which other places could grow.

But the neighborhoods east of the river are attracting new residents and businesses. In some sense, that’s part of the problem, in the short-term at least. Real estate values east of the river incorporate, to some extent, the likely effects of the area’s expected growth. Land is relatively pricey, in other words, because five and ten years down it’s a good bet that there will be a market for higher end residences and retail options. But that market isn’t yet developed. You can’t run a profitable business catering to the existing market while paying land prices that incorporate the value of running a business catering to the market of the future.

Landowners should nonetheless be willing to rent space at a loss to current restaurateurs; that’s a better option than allowing a place to sit empty. But in fact that may not always be the case. They probably can’t get businesses to take short-term leases, and longer-term leases rule out the possibility of a better paying option down the road.

But these things will change. The land east of the river has a lot of advantages and will soon have a lot more employment and better transportation infrastructure. New developments are planned that will increase residential density in walkable areas. Before too long people will be asking how there came to be so many restaurants in Anacostia, when just a little while ago there were almost none.

It’s pretty clear that long commutes are bad for you. Accidents kill 40,000 people a year, driving is stressful, and so on. At the same time, it’s worth being a little cautious when dealing with figures like those Richard Florida presents here. He posts charts summarizing data from a Gallup-Healthways survey, tracking the incidence of various health problems sorted by commuting time. So those with hour-long commutes are more likely to have neck and back pain, high cholesterol, stress, and so on.

It would be silly to say that there is no causation running from the driving to the health problems. But there are two other things to keep in mind. One is that people with certain health problems are likely to select into lifestyles that involve long commutes. Research indicates that the obese select into suburban, auto-oriented lifestyles. Similarly, someone with neck and back pain might not be particularly excited about life in a walkable, bike-able community.

The other thing to note is that much of the increase in the variables presented takes place at the long end of the distribution — those with one-way commutes in excess of an hour. Commutes of that length are relatively rare, and they’re also associated with declining incomes. Someone who has a two-hour commute experiences much more stress than someone with a 46-60 minute commute (who is roughly as stressed as someone with a 21-30 minute commute). They might be stressed by the commute, or they might be stressed by the set of circumstances that led them to live so far away from their job — low income levels in an expensive city, economically-induced household immobility, and so on.

At lot of times when we talk about location decisions, we’re really talking about income. And of course those things are related — bad housing policies make homes with good access to employment expensive. It’s just worth being careful about attributions of causation.

Tyler Cowen wrote a very nice column for the New York Times noting the many ways in which government policy leads to the underpricing and over-provision of parking, generating all kinds of nasty effects. One of the results of the piece was a barrage of perplexing responses from people who normally agree with Tyler (in O’Toole’s case the response isn’t perplexing; getting these things wrong is literally his job).

One thing that surprises me is that libertarian economists wouldn’t immediately adopt the default assumption that mandated parking minimums are bad. What does it mean to be a libertarian if that’s not your default position? Ditto for below-market pricing of scarce resources. You’d expect progressive writers to make a strong case that goods a, b, and c should be affordable to everyone and government subsidized as a matter of basic decency. It’s bizarre that libertarians leap to this position when driving-oriented policies are up for discussion.

Another thing I found curious is the suspicion that unused parking spaces are evidence of a policy that’s not welfare-maximizing.

But the main point is that it’s very difficult to make a positive case for government provision of parking spaces or mandated parking minimums. Given the existence of government provided spaces, it’s harder still to argue against market parking pricing. We have many examples of private firms building and operating parking lots or decks, charging positive prices, and doing a lovely business that seems to work well for operator and driver alike. How does one justify government intervention?

Now you might argue that there are public good considerations involved; that parking spots are like other bits of transportation infrastructure in that there is a role for government provision. Personally, I think parking spots are more like gas stations than roads, and meanwhile roads should be congestion priced (as many transit systems already are — and then some, in some cases). You’d think that libertarians making the public good argument would have no problem defending government provision of and subsidy for transit, but of course they don’t. They get around this by arguing that people want to drive and they don’t want to ride transit. This is strange in that in few other cases would a libertarian claim to know what markets want, and while they might refer to mode shares, those shares are themselves determined by decades of heavy subsidies for all things auto.

Anyway, I appreciate Tyler making the case against government parking subsidies for a wide audience.

Commenter JRoth says (among other things) that I blew it on the issue of saving the automakers:

To my non-surprise, he’s written nothing. Back in the winter of ‘08-’09, he couldn’t say enough about the need to destroy GM: they would never be profitable again, it was a waste of money to save them, and they were an inherently wicked company. Three pages of GM-related posts come up on the site search, virtually all from that timeframe, and not a single one in the last year.

That would be the year that has proven Avent completely fucking wrong. His analysis, and his understanding of the facts, were completely mistaken, driven (I guess) by emotion and ideology rather than any kind of grounding in the history of industry, industrial policy, and deindustrialization.

For the record, I didn’t advocate that anyone “destroy” GM. My recommendation was that the government provide a delicate way to allow the company to fail (GM had no trouble bleeding cash all by itself). But how about it? Was I wrong?

GM did just announce a quarterly profit of $1.3 billion, which seems nice. I may have overestimated how much money they’d need from the government before turning a profit. But let’s be clear about this: as of March of this year, the government had $76 billion committed to the automakers. So it’s perhaps a bit soon to be declaring the rescue a success.

I also argued that economic transition in the Midwest was an inevitability — that American auto sales and auto industry employment were unlikely to ever recover their pre-recession levels. That continues to look like a pretty safe bet. American auto sales remain well below the typical levels before 2007: around 11 million annually compared to 16 million in 2006. The same is true for employment in automobile manufacturing. Around 700,000 people are currently at work building cars in America (for all manufacturers, not just the Big Three). That’s fewer than at the point the rescue was initiated, and well below pre-recession levels. In 2006, nearly 1.1 million people were building automobiles. In 2000, the figure was over 1.3 million. The long-term health of the Midwest will depend on the strength of industries other than the automakers. That was true in 2008. It’s true now.

The other obvious point to make is that we need to know what the counterfactual is here. I never proposed that we tell workers in the Midwest to go take a hike. My view was that rather than throwing money at companies that would employ steadily fewer Midwesterners, no matter what, it would be a good idea to invest in the people and infrastructure of the Midwest. Sure, GM had a quarterly profit of $1.3 billion. So what? Is that a better or worse outcome for the Midwest than we would have seen had the $85 billion committed to the automakers been used to retrain workers, support credit to businesses generally, and invest in the physical infrastructure of the region? That’s clearly unknowable, but if we could run a controlled experiment, I know which strategy I’d put my money on.

There’s nothing miraculous or special about the fact that companies propped up by billions in government money continue to exist. And if that’s the measuring stick you’re using, you’re using the wrong measuring stick. I don’t care about individual companies. I care about the broader economy and the people it employs. And my sense continues to be that the money provided to GM has done little to improve the long-run growth prospects for the Midwest and would have been better used on other investments.

Megan McArdle has weighed in on the debate over fuel taxes and innovation, in a post that I don’t find particularly persuasive. She begins by arguing that, hey, Europe is different:

I am not going to extensively rehash the ways in which the US is simply different from Europe:  the number of children and the length of time they have to spend in car seats that cannot be crammed three-to-a-bench-seat; the geographic dispersal; the extremes of weather which make air conditioning and heating necessary for much of the country.  I will note that though our population is majority urbanized, large swathes of it are living like my grandmother, in a place where the nearest “big city” is almost an hour away and has 150,000 in it, something that I don’t think is physically possible to achieve in most of Western Europe.

Europe is different. But so what? Fertility varies widely across Europe. The French are nearly as fertile as Americans, and vastly more so than the Italians. Are the French more transit averse than the Italians? Are the Scandinavians lousy urban planners? No, obviously. Neither is the density of the population the big factor here. Germany and Italy have population densities similar to that in Maryland, while France has a population density on a par with Ohio or Florida. Plenty of grands-mères live an hour away from “big” cities of 150,000 people or fewer.

Next, Megan notes that where urbanism is concerned, America has its work cut out for it:

I will also note that it doesn’t do much good to say that we shouldn’t have sponsored so much auto-heavy development, because the fact is that we did, and these things seem to be path dependent.  The only city in which a majority of the population find it convenient to move about entirely car-free is New York.  No city that I am aware of has managed to replicate auto-free development in an auto-abundant atmosphere no matter what tax incentives were applied.  DC is the best we’ve managed to do so far, and according to wikipedia, a whopping 32% commute by any form of public transit; almost half drive.  That’s of the people who work in the district.  Obviously, the numbers outside will be lower still.  That’s after decades of relatively heavy investment in the metro, and one of the worst traffic congestion problems in the country.  I am very skeptical that changing our living modes will get us from here to there.

Well, ok. But America has a low gas tax rate. Isn’t the discussion here concerning the pace and quality of adjustment in response to a price signal? Personally, I think Metro has been remarkably successful given that for most of its history real oil prices have been declining (along with the real value of the federal gas tax). Since oil prices began rising again just over a decade ago, the District has enjoyed an impressive renaissance, centered on transit-oriented development. The city has moved from population decline to rapid population growth. In the District, most people do not commute in a single-occupancy vehicle, and SOV mode share is declining. And per capita transportation emissions in the District are low by American standards, even among those who don’t take transit to work. So that’s all good then.

Megan closes:

Nonetheless, I think that Manzi’s argument is a serious challenge to those of us who are in favor of a carbon tax.  It’s not enough to say that it will change our marginal propensity to live in high rise apartments.  The Chinese are already living in high-rise apartments, and they’re going to rapidly outstrip us in greenhouse emissions if something isn’t done.

Let me reiterate that my argument was not merely that price signals would trigger substitution, but that price signals would lead to innovations which improve the quality of the substitutes. Take the District. Living in a relatively dense environment with good transit access is a substitute to a driving-intensive lifestyle, and one a number of people have opted to accept as the cost of the driving life has risen. But the quality of that substitute has steadily improved. Bike and car-sharing services have sprung up in the District, making urban life easier. So have food carts and home delivery businesses. Smart phone apps have enriched urban life. The quality of urban consumption amenities has soared. And the end result is that lots of people want to move into the city for reasons that have nothing to do with the cost of transportation. And this is in a very short time-frame.

Two other points. First, Chinese per capita emissions are well below ours and are unlikely to ever surpass ours. And second, transportation accounts for a far, far smaller share of Chinese emissions than it does of American emissions — around 5%, compared to about 33% in the US. This suggests that it will be innovation in other areas — industry and power generation above all — that will prove crucial in limiting Chinese emission growth (and growth globally).

And it is here where Jim Manzi’s use of the fuel tax example is most problematic. For a few reasons. First, a given carbon tax would have a much larger effect on coal prices than it would on gasoline prices. A carbon tax of $30 per ton would increase gasoline prices by about 3%, and coal prices by something like 10%. Second, there is good reason to believe that energy demand in non-transportation sectors is more responsive to price changes than demand in the transportation sector. Demand for gasoline is generally estimated to be fairly inelastic (especially in the short run) relative to demand for electricity (and particularly relative to commercial demand for electricity). Finally, the record of innovation in response to changes in energy markets is impressive and dramatic relative to the record in the transportation sector.

In other words, it’s more important to encourage innovation in electricity generation and consumption than it is in the transportation sector, and a carbon tax is more likely to generate innovation in electricity generation and consumption than it is to change the transportation sector. Which is why a price on carbon is so important to have. Confident as I am in the ability of higher fuel taxes to generate transportation innovation, that’s not the main reason to favor a carbon price.

Let’s address this comment:

It’s disingenuous to keep referring to government pricing with reference to the market such as “the government wasn’t using market prices.”

Requiring people to pay for road space might be a good idea but it has little or nothing to do with markets or creating a market. It’s pricing by government decision — not pricing by individual consumers bidding on their perception of value in real time.

The institutional structures will (my surmise) never allow or be able to do dynamic pricing. How would the bid process work? Park by the side of a road until the flow of cars was low enough so that the agency would lower the rate to encourage more capacity? In real time? Every 30 seconds?

Making people pay to drive cars may be necessary but let’s not play with words and call it “market pricing.”

So, a congestion charge would obviously be set by a government, and if your definition of “market price” excludes any price set by the government, then you’ll obviously conclude that a congestion charge is not a market price. But I’ll disagree with you.

First, it’s clear that dynamic pricing would be difficult to impossible to achieve, nor would drivers be able to bid in any sense on what a trip is worth to them. But when I walk into my local grocery store, I don’t announce bids or hang around waiting for a food item to drop to a level I’m willing to accept. I pay the same thing every morning for a cup of coffee. But I don’t think anyone would argue that the six-pack of beer I get at the grocery or the morning coffee aren’t sold at market prices.

Or we can look at this another way. I think most people would agree that the price of carbon in a cap and trade system is a market price. A congestion charge isn’t that different. The government is basically adjusting the cost of a negative externality until it finds the point at which congestion is “capped” at zero. And basically we’re talking about matching consumer demand with a fixed supply of road space.

Are prices charged by monopolists market prices? Price may not equal marginal cost, but really, how often does it?

Nothing to add to this at this time, except that it suggests pricing will be easier to adopt in places where there is already a functioning transit system (or systems). Those hoping to use congestion pricing to pay for transit improvements would be wise to build the alternatives first, then move for pricing.

In this paper, we take a political economy approach to study the introduction of urban congestion tolls, using a simple majority voting model. Making users pay for external congestion costs is for an economist an obvious reform, but successful introductions of externality pricing in transport are rare. The two exceptions are London and Stockholm that are characterized by two salient facts. First, the toll revenues were tied to improvements of public transport. Second, although a majority was against road pricing before it was actually introduced, a majority was in favor of the policy reform after its introduction. This paper constructs a model to explain these two aspects. Using a stylized model with car and public transport, we show that it is easier to obtain a majority when the toll revenues are used to subsidize public transport than when they are used for a tax refund. Furthermore, introducing idiosyncratic uncertainty for car substitution costs, we can explain the presence of a majority that is ex ante against road pricing and ex post in favor. The ex ante majority against road pricing also implies that there is no majority for organizing an experiment that would take away the individual uncertainty.

The whole thing is available here.

He says:

This strikes me as, at best, a word game. I understand that innovation is not identical to invention. But this is like saying that in response to an increase in the price of peanut butter, I “innovated” by making smaller sandwiches and eating ham-and-cheese more often (while noting that I designed these new sandwiches very well, and am probably healthier anyway with less peanut butter in my diet). If by “innovation” in response to higher gas prices, we mean switching to smaller cars and taking the bus and riding bicycles more often, then I agree entirely that higher gas prices in the U.S. will induce innovation.

The point is not that Europeans have responded to higher gas tax rates by using substitutes. The point is that the quality of substitutes has improved considerably over time — indeed, faster than improvements in the same substitutes here in America. If it’s relatively cheap and easy to respond to higher gas prices by using and improving alternatives, then why go the hard way and spend billions trying to invent some magic new zero emission engine? And to reiterate, this is why carbon pricing is important: because it will provide us with lots of evidence screaming that we shouldn’t spend billions trying to invent some magic new zero emission engine, because it’s a waste of money and there are better alternatives available.

In the latest edition of the New Yorker, or at least the latest edition to arrive at my house, there is a piece by Keith Gessen on the epic congestion of Moscow. Gessen quizzes a number of urban planners, traffic engineers, and so on on the roots and meaning of traffic, and a variety of explanations pour forth. The regrettable design (central city streets five, ten, eighteen lanes wide), the lack of driving decorum, the oddly anachronistic traffic technology, and the anything-goes-for-rich-drivers set of special privileges. Design can obviously impact traffic to some extent, but it’s not as though any city has hit on the combination of road design and rules that has eliminated traffic. The closest Gessen gets to a real answer on congestion comes from University of Pennsylvania transportation expert Vukan Vuchic, who provides the generally understood, yet generally unheeded, truth: “No city has ever constructed itself out of congestion.”

And yet the author comes dangerously close to understanding the phenomenon, through simple reflection:

I recalled Vladimir Sorokin’s novella “The Queue,” from the era of the Brezhnev stagnation, which is also about a line — a line of people waiting to buy something (it’s never clear what, and they themselves do not know), the line so long, so complex, that they, too, begin to live in it.

We’ve been here before. the cars standing in endless lines on the crowded Moscow streets: they resemble nothing so much as the people who used to wait in endless lines outside the Moscow stores for Polish coats, Czech shoes, and, famously, toilet paper. Now, more comfortably, they wait for the light.

Exactly! This should be a eureka moment for Gessen, but he quickly moves on to other thoughts. But this is the heart of the issue. Why were old Soviet citizens forced to queue for hours? Because the government wasn’t using market prices to allocate scarce resources. And why are Russians doomed to interminable congestion today? Exactly the same reason.

Now, Moscow may lack the institutional strength to adopt congestion pricing. Gessen relates how Moscow attempted to charge for parking by deploying orange-vested men to collect parking fees. This led to a boom in the wearing of orange vests in pay-parking areas, and the diversion of parking fee money (by both the legitimate parking attendants and the frauds) to private ends.

Still, we should reflect on the lessons here. Most cities are better governed than Russia. The competent Swedes have made congestion pricing work, as have the British. America, land of market-worship, should be able to manage the whole market pricing thing.