This piece popped up in my RSS reader as being recommended by Reihan Salam, and I figure it’s worth a few comments. It’s basically a discussion of a book (which I haven’t read) on how the world may look with steadily rising oil prices. So, first of all, there’s this:

In Chapter $8 he foresees the bankruptcy and liquidation (not reorganization) of most of the airlines in the United States and Europe. This assertion is plausible. Most people do not realize that US domestic flight seats have already shrunk 21% since 2001.

For the time being, analysts agree that the airlines, by cutting routes and employees, grounding planes and imposing fees, can weather the downturn. In fact, when the latest round of capacity cuts takes effect in September, the seats on domestic flights will drop to 66.5 million — down from a peak of about 84 million in 2001 and the lowest September figure since 1984, according to OAG Aviation, which tracks flight schedules.

We will, in time, return to 1970s and 1960s levels of air passenger transportation. I do not expect technological advances to prevent this because the aircraft need liquid fuels - no electrically powered substitutes available. Boeing’s much ballyhooed 787 Dreamliner only boosts fuel efficiency 20%. Aircraft fuel efficiency would need to improve by multiples to compensate for Peak Oil and energy substitutes would have to be in the form of liquid fuels. Unless algae genetic engineering solves the problem we’ll do a lot more of our travel on the ground. Robert Rapier has explained better than I can the problems of algae biofuels.

I don’t know much about the potential alternative fuel or airline technologies, but I do know that last year’s oil price spike dealt a serious body blow to airlines. It seems to me that it will be much easier to handle rising oil prices on the ground than it will in the air. Intercity travel is very important, economically speaking, so this is a good reason, in addition to the environmental benefits, to invest in fast intercity rail. Highways are a very poor substitute for air travel.

Next, we have this:

Steiner’s lifestyle preferences are not my own. He celebrates what he sees as an expected return of the suburbanites to cities. Me, I think the people who left cities had good reasons for doing so. I like wide open spaces and I even like some suburbs. A retreat from the rural life and from exurbs doesn’t strike me as a positive development. I think SWPL writers who live in cities do not appreciate that, no, not all people will enjoy cities like they do. City dwellers can be pretty damned provincial. (and why didn’t he mention Vespa scooters?)

There’s a lot wrapped up in this passage. First, its important to recognize that we ought to try and take tastes out of the equation. Preferences will always vary, and given most any set of costs there will be those who prefer living in the country, those who like small towns, those who want to live in the suburbs, and those who prefer an urban life. What’s important is to try and focus on what policy is encouraging and discouraging and whether those policy impacts make any sense. It seems pretty clear that government policy — via things like highway subsidies, mortgage-interest deductions and FHA lending, zoning rules, and so on — has strongly encouraged suburban development over the past half century. The market has therefore provided too much suburban style housing and too little urban housing.

When urbanists like myself argue in favor of better policies, it is (generally) not with the belief that we all should or ever will live in very dense urban environments. Rather, I think that we should improve policy, and that the result will be slightly fewer living in low-density environments and slightly more people living in high-density environments. And it’s worth pointing out that “high-density” can mean many things — everything from Midtown Manhattan, to the walkable rowhouse neighborhoods in the District, to transit-oriented neighborhoods in places like Arlington where dense development around transit hubs rapidly gives way to detached but compact single-family homes. To the extent that any urbanist out there is arguing that everyone must live the condo and Whole Foods life, I’m ready to declare that they’re wrong.

Two other points. Adjustment to costs, including rising fuel prices, will also mean that a lot of traditional suburban areas find it advantageous to develop in smarter ways. We won’t necessarily see a mass migration to urban areas, in other words, but a slow metamorphosis of suburban areas into places that are less dependent on automobiles. And suburban development is increasingly anachronistic given demographic changes. There are clear advantages to having a big house with a yard if you have two or three kids, but the share of households with children is declining and will continue to decline. Part of any shift to a more urban life, in other words, will be a natural result of the aging of the boomers and the increase in the number of singles and childless couples.

But I have a more fundamental disagreement with him over the idea of a coming urbanism: I’m not sure that by his own logic it makes economic sense. In Chapter $16 (yes, he thinks gasoline will go that high - I’ll outline opposing arguments below) he argues that the cost of long range transportation will get so high that more food will be grown and consumed locally. Well, okay. But isn’t that an argument against large cities? New York City needs to bring its food in from longer distances because it has so many people. The amount of land needed to feed them all has to stretch many miles away from it. This is made all the more problematic because NYC borders on fairly built up urban and suburban areas around it that also need to have their food brought it from distant places. So isn’t the high transportation cost argument an argument for the spread of people out to places closer to where the food is grown? In the United States that would be places like Nebraska, Kansas, and the Dakotas.

No. For one thing, we can look back at what America was like when transportation costs were higher and observe that cities were far denser places. Secondly, there are tremendous economic advantages to urban agglomerations; otherwise there wouldn’t be so damn many of them and we wouldn’t live in them, because land out in the boonies is a lot cheaper than land in a major metropolitan area.

If you look at a typical model of economic geography, you see that for very high and very low transportation costs economic activity is widely dispersed, but for costs in between there are gains to clumping. Frankly, I don’t see any oil price that would destroy the advantages of urban agglomeration. Transport will shift to greener modes first, and people will shift more of their consumption toward food for a long, long time before that happens.

The piece then goes on to argue that gas probably won’t hit $20 per gallon, and I think that’s right. In the short term, rising gas prices would send the global economy into a tailspin long before gas went to $20 per gallon in America (as we saw, $4 per gallon was sufficient to wreak significant havoc), and in the long term I think prices far lower than that will be sufficient to push automobiles completely off gasoline. There is some sweet spot such that substitutes for gas or cars develop quickly enough to prevent a price-spike driven recession but slowly enough that a hefty share of the world remains dependent on petroleum where prices might find their way to $20, but I think one direction or the other is far more likely.

And while I certainly leave open the possibility that a clean, affordable biofuel might be developed sufficient to keep internal combustion engines humming for decades, it seems pretty clear to me that what needs to happen is that the transport network needs to be electrified — via increased use of rail and migration to hybrid and the electric vehicles, and the world’s generation capacity needs to be greened. And because high oil prices seem likely to be a persistent damper on economic activity for the foreseeable future, that shift should begin immediately.