Peltzman Effect Bleg:

I recall seeing a paper at some point that found a Peltzman effect for higher automobile gas mileage--i.e., when automobile gas mileage rises, people live further from work and have longer commutes, thereby dampening some of the reduction in fuel consumption that might otherwise occur. I also seem recall a finding that this increase in commute time also leads to more accidents as a result of greater driver fatigue.

Note that I am not looking for the classic Peltzman effect example about car safety and more accidents, but rather any countervailing effect related to gas mileage and the effect on fuel consumption.

Calculated Risk:
The problem with this analysis is that it fails to see the ability to live farther from work (i.e. to have more housing options) as a benefit.

It is true that if you increase someone's consumption options that they might choose to consume more of the resource that lead to the efficiency. In some cases, you could imagine someone consuming more than 100% of the efficiency, because the efficiency made choices that were previously out of reach within reach. But even in those, probably rare cases, consumer welfare is still enhanced.

Finally, let us move away from theory land and into reality. For most people with long commutes, it is the traffic and the lost time itself, rather than the price of gasoline, that is the most valuable loss. After all, low paying jobs are fairly easy to get, and there usually is really no reason to incur a long commute for a low paying job since there usually are more convenient substitutes and the cost of commuting are likely to eat up much of the income derived from such a job. For people with higher paying jobs, it is quite likely that the lost time and the frustration of a long commute heavily outweigh the cost of gas in the calculus of their well-being. In other words, if they are incurring the cost of a long commute, they probably have a good reason for doing so and this is decision is not going to be influenced by small fluctuations in the cost of gasoline or the fuel efficiency of their vehicles. (i.e. there demand is relative inelastic)

Why do people commute long distances despite the evident sacrifice in terms of time and quality of life that such a commute entails? Often, it has to do with family responsibilities. For example, they want to raise their children in an affluent suburb where they can attend a higher quality school that will lead to greater opportunities for their children. In other words, who locate far from work often have good reasons for doing so and in most cases would do so regardless of fuel efficiency standards. Now, it is true that higher fuel efficiency may enable more people to make this sort of lifestyle choice. But I do not understand how that could anything other than an unqualified positive.

Another major point needs to be made. People do not make decisions about where to live or where to work very often. It would be somewhat rare case where someone decided to change a job or break a lease because their car got 40 mpg instead of 30 mpg. In most cases, people will stay put and in fact increases in full efficiency will be completely realized in lower consumption of gasoline. Also, decisions about where to work and where to live can be classes as major decisions. The difference between 30 mpg and 40 mpg is relatively minor. In most cases, that difference will not change the calculus of choosing a particular job or a particular living situation much.

None of this is to deny that there would be a countervailing effect. In fact, this is true of any efficiency involving a good that can be had at a lower price. Now, what is the real good here? The good is not gasoline, it is what you can do with gasoline. That is, the distance you can travel. A more efficient car means that the costs to travel a mile has decreased. Well, since traveling is unlikely to be a Giffen good for the vast majority of people, of course a lower cost will result in more travel.

But, on the whole, this effect is likely to be small. People tend not to travel for its own sake, but because they are going somewhere. They stop when they reach their destination, they do not continue driving merely because the cost of traveling is smaller. Finally, for the major factor that determines the consumption of travel in a car more than any other, commuting, the decisions where to live and where to work are major decisions, and are unlikely to be greatly affected in all that many cases by changes in fuel efficiency.

The bottom-line is this. If you are trying to argue against requirements for increased fuel efficiencies, which I suspect, by suggesting that benefit will largely be dissipated by more travel, I suggest you go back to the drawing board. The increased consumption frontier that results from increased fuel efficiencies will mostly be used to consume goods other than increased car travel, and there will be a large benefit in terms of decreased quantity of gas consumed from such efficiencies.
9.20.2008 5:10pm
A. Zarkov (mail):
I did a calculation and posted the results on the VC a while back. The price of real estate should decrease by $400 per mile per dollar increase in the price of gas. This assumes that the drivers put no value on the time for an increased commute. Let's do an example. Stockton to the Bay Area, which is a distance of approximately 85 miles. If gasoline goes from $2 per gallon to $7 per gallon, then Stockton residential home prices should drop by $170,000. This calculation assumes that cars get 20 miles per gallon. If cars got 30 miles per gallon then Stockton property should drop only $113,000. It should be easy to work out an approximate model for how gas mileage increases will increase commute distances. You need a relation between range and cost, which might be available. Or look at how land cost decreases as the radius from the commute point increases. I suspect it's an inverse square law of some sort.
9.20.2008 5:37pm
trad and anon (mail):
I don't have access to the full article, but could it be Greene, David L. ; Kahn, James R. ; Gibson, Robert C. Fuel economy rebound effect for U.S. household vehicles, from the July 1999 issue of The Energy Journal? Google also has a citation (but no direct link to the article) for Greene, D.L. (1992). "Vehicle Use and Fuel Economy: How Big is the Rebound Effect?" The Energy Journal, vol. 13, no. 1, pp. 117-143, April, which sounds from the title like it might be on-point.

Google Scholar is your friend.
9.20.2008 5:46pm
Glenn W. Bowen (mail):
Divergent of topic:

That might be longer by mileage but not necessarily time (?).

Read, half-heartedly, a book "Traffic" recently. In line with the topic, it was stated the ideal commute sought after is no less than 40 minutes, IIRC.

Also, within a living area (what a resident would consider his environs), the average time from it's center to the edge and back is usually about an hour.

An article on traffic by the author:

9.20.2008 7:21pm
John Thacker (mail):
The increased consumption frontier that results from increased fuel efficiencies will mostly be used to consume goods other than increased car travel, and there will be a large benefit in terms of decreased quantity of gas consumed from such efficiencies.

Calculated Risk:

Oh certainly, a priori if you assume the choice is costless between less efficient and more efficient technology and vehicles. Quite obviously if you can simply magically improve fuel efficiency for free you'd do it. (Some environmentalists somewhere might well disagree if they were focusing on one input alone, I suppose.) But TANSTAAFL applies here, and there are costs from imposing the regulation-- one does have to ask why the more fuel efficient vehicles don't exist, or why people aren't purchasing them if they'd truly be better off. Emissions are a different story-- no one benefits directly from their own car emitting less, but people capture a great deal of the gains from their own fuel efficiency directly, as you note.

CR, your point that people who choose to drive more when better fuel efficiency is imposed on them must be better off than if they had driven the same is quite true-- but at the same time that leads you to the conclusion that they're (very likely) better off with their current vehicle choices than with higher fuel efficient imposed on them. The lower fuel prices due to the slight decreased consumption by everyone is much smaller than the possible savings by changing one's own vehicle. The only people likely to be better off are those who already choose to use the most efficient available alternatives.
9.20.2008 7:39pm

I did a calculation and posted the results on the VC a while back.

One difference between the cost of housing and the cost of gas is that the former can easily be financed at reasonably low interest rates whereas the latter can only be financed by credit cards at very high interest rates. We see this very issue work out when drivers of expensive, luxury SUVs complain about the price of gas or otherwise adjust their car-buying habits. It seems odd for a person who owns a $50,000 H2 to try to sell his truck to save $2,000 a year on gas until you realize that the guy has financed the entire purchase price of the H2 but is forced to pay his gas bill out of his pocket.

To make a long story short, I think we have to adjust for the differences in the way people pay for gas and housing. It might just play a pretty significant role.
9.20.2008 11:05pm
Dave Hardy (mail) (www):
There are many imponderables here. Having living barely "inside the beltway" in DC ... gas prices are nothing.

Inside the beltway, evading rush by going in early and returning early, major issues were time of transit and cost of parking and having to break off every two hours to feed the meter. I had to balance that against hitting a subway spot early (because the parking filled up) and paying for that plus the subway.

Outside the beltway, with the major congestions that arise because of beltway commuters hitting it, it might have been a different equation, even tho it was only ten minutes and a few cents per drive.
9.20.2008 11:57pm
TruePath (mail) (www):
This is just one (not very good) example of why personal conservation efforts are nearly useless.

The only real way to reduce CO2 production is through a carbon tax. Otherwise reductions in fossil fuel use or increases in efficiency will primarily result in other uses for that fuel.

Why is it so hard for people to understand that if you want significantly less fossil fuel usage you need to increase the price of fossil fuels!?!
9.20.2008 11:58pm
TruePath (mail) (www):
Calculated Risk:

I disagree that the choice to live farther away as a result of increased fuel efficiency is necessarily or even likely a benefit.

I mean consider the following (reasonable) model. Most people like to live just outside high concentrations of population so they can enjoy the cultural benefits and proximity of friends while gaining a bit of space and an increased feeling of safety about their children. In other words the ideal place to live is in a suburb right on the edge of a dense city.

Alright so what happens if we increase the distance that everyone can commute. The people who previously couldn't afford to live outside the urban center move out from it to the near suburbs which increase in density and become part of the urban sprawl. Those people who used to live just on the edge of the city now move even farther out and so forth. Everyone then might end up being significantly worse off.

The point is that choice of living location may be a prisoner's dilemma type situation. We would all like everyone else to cooperate and live in the dense city so we can access them easily while letting us live right outside. Unfortunately everyone tends to defect and the greater the more ability they have to do so the worse things might be for everyone.

Also, don't forget that people are amazingly irrational about choosing what will make them happy. It would not surprise me if people insufficiently weighted the benefits from friendship and other interpersonal interactions from living closer in to the city.
9.21.2008 12:12am
All stable many-degree-of-freedom systems behave this way. It's a trivial consequence of the fact that the second derivative of your measurable (in this case commuting distance) with respect to your variable if interest (in this case cost of gasoline) must be positive if the system is stable to fluctuations.
9.21.2008 4:36am
Zywicki (mail):
Trad and anon:
Thanks—it looks like my problem was that this is referred to as the "rebound effect" in this context, which was terminology with which I was unfamiliar. So I couldn't come up with the right search terms. When I search under "rebound effect" I do find a whole bunch of stuff. So thanks for the pointers.
9.21.2008 11:27pm