It would be bad form to start a writeup as long as this one without a quick explanation of the topics within. This is a writeup that is mostly going to discuss American cities, but the precedents in it apply pretty well to the social patterns of most developed countries.

Peak oil is a thing that is going to happen. People are going to be arguing for a long time about just how significant its impact is going to be; a lot of people think it's the end of the world, and others (me included) think that it's just a force for strong change. The idea is that demand for gas is increasing parabolically, but the rate at which it can be produced is not increasing accordingly. Neither is the rate of discovery. "Peak oil" refers to the point at which demand begins to overtake supply. It's pretty simple. In the US, even the Bush administration is starting to get concerned about this; gas prices have been steadily climbing for months and are starting to accelerate at worrisome rates.

http://www.lifeaftertheoilcrash.net offers some good facts and sources on this, albeit with an extremely catastrophe-minded attitude.

On to the thesis.

When the supply of gas becomes and remains scarce, American cities will visibly compress, and many suburbs will experience rapid depopulation.

Not everyone is going to agree with this idea. For one thing, it has become pretty clear that the price of gas does not matter to most American drivers. People are going to pay whatever gas costs, until it becomes truly backbreaking. Even if gas were to rise to US$5 or $10/gallon, people would still pay it. Most people's driving habits would start to become conservative, but very few Americans would give up their cars at this point. What needs to be considered is the point of restriction where, as in the 1970s, the public cannot be supplied with an adequate amount of gas and becomes unable to purchase it.

Immediate effects

Public transit will obviously see some major gains at this point. An excellent opportunity to observe the consequences of the disabling of automotive traffic occured March 26, 2004, in Southern New England, when an oil tanker crashed and caught fire on Interstate 95 crossing the city of Bridgeport, Connecticut1 and caused such severe structural damage to the highway that a long segment of the freeway had to be closed entirely. The public response was poor in what was already an extremely car-dependent area, but services quickly became available. The lack of adequate alternate corridors (notably the already-congested Merritt Parkway some distance to the north, which could not accomodate large vehicles) led to an expansion of local rail service. Shore Line East already operated along Long Island Sound, but Amtrak also redirected several of its Northeast Corridor trains to make commuter stops in order to increase mobility.

The Bridgeport disaster only lasted for a little more than a week. If such an event occured on the long term, the development of transit would be more rapidly spurred. A decline in passenger vehicle traffic would allow better usage of existing road corridors by buses and vans, but a growth in rail transit could also be expected. As gas prices continued to rise (even with reduced automotive traffic, oil demand is well spoken-for in agriculture and industry), the residential areas near transit stations and corridors would quickly become the most desirable housing in suburban areas.

Long-term effects

The central city would remain largely unaffected by a spike in gas prices. Transit utilization would improve, but the general proximity of essential services and employment to residential areas in a dense city allows car-free existence. In New York City, 63.2%2 of the population uses transit or walks to work. In Boston, the equivalent number is 45.3%. Even in less dense cities, the potential for infill exists -- by contrast, 18.5% of Atlanta's population uses transit or walks to work. In Los Angeles, it's only 13.8%, due to the lack of transit infrastructure. Los Angeles would get hurt.

Essentially, what can be expected to occur is the gradual urbanization of formerly suburban areas. Transit lines connecting these suburbs essentially evolve into an intercity rail network. These smaller "transit villages" are often compact enough to allow pedestrian traffic to suffice for the vast majority of commuters.

Dependent towns

Essentially all American cities are surrounded by a web of suburbs and freeways or high-traffic arterial roads. The Boston metropolitan area provides a useful case study due to the fairly effective transit network and the dense urban core. Although only 50.7% of the city's residents drive to work, in the urban area3 78.7% drive, either alone or in a carpool. In fact, nearly every town within 10 miles of the city center (the exceptions being the beltway cities of Woburn and Waltham, and the city of Cambridge, arguably part of the same core economy) sends more workers to Boston than to destinations within the town. (One suburb, Belmont, sends more workers to Boston AND Cambridge, individually, than it keeps.)

These dependent towns are the best way to measure the suburban landscape created by a city, and, conversely, are the towns most likely to contract or be eliminated entirely by a long-term gasoline shortage. The inner core of suburbs is more likely to see positive development (as it has already begun to do, in most American cities).

Lifestyle benefits and New Urbanism

New Urbanism has been a rising philosophy in the United States and in Europe, essentially by extolling the benefits of urban living in terms of increased social interaction, cultural development, and the available of both essential services (food, utilities) and nonessentials (entertainment, etc.). There is a growing trend among Americans under thirty years old of all social classes to migrate to the city, one that would almost certainly encourage the development discussed here.

Urbanization provides the simplest solution to a gas crisis, and also creates the potential for agricultural and industrial production in vacated suburban tracts. Improved land use patterns are one pleasant consequence of this development, as is the infusion of commercial development into areas that are now strictly residential.

Preparation

It's tough to accept a doomsday theory, even with a crisis as disturbing as peak oil. Alternative energy could be capable of saving the day, as well; however, it would be in the nation's best interests to develop contingency plans should the growing voices in the scientific and economic communities be correct.

The question of suburban disinvestment is a difficult one -- as the demand for houses at long distances from employment centers decreases, the ability to sell those same properties diminishes as well. Pre-emptive government intervention offers one solution; subsidizing both new landlords and the construction of new, multiple-unit housing in less densely-populated areas of American cities would both ease this transition and improve the current lifestyles in these cities, as would keeping rights-of-way available for future transit development.

American cities have many, many issues left to deal with that would not be mitigated by sudden population growth. These include crime, poverty, and declining school systems, all problems that the nation has not addressed well in fifty years. Do some research. Get the information that might matter if changes start to happen.


1 Sources: 2 Source: 2000 US Decennial Census, factfinder.census.gov
3 As defined by the Census definition of Metropolitan Statistical Area.