Shouldn't we be taxing gas more heavily?

I was lucky to be able to attend part of the NBER workshop on Environmental and Energy Economics at Stanford last week.

My favorite was a talk by Michael Anderson of UC Berkeley.  He spoke about a paper joint with Max Auffhammer, also of UC Berkeley:

"Vehicle Weight, Highway Safety, and Energy Policy"

Sorry, no link.  The  issue is one that's been talked about many times: an arms race in vehicle weight and safety.  The essential problem is that the heavier my vehicle, the safer it is for me and the more dangerous it is for you.  Now, if we could all commit to smaller lighter cars, we'd pay less for our cars, have better gas mileage, and be no little less safe [please excuse my exaggeration], since when it comes to car-on-car collisions, it's mainly relative size that matters.

This sets up a classic prisoner's dilemma in which it's smart for one and dumb for all to buy bigger, heavier vehicles.

That basic tension is pretty well known, I think. What Anderson and Auffhammer did was measure, with apparent extraordinary accuracy, the size of the external cost of extra vehicle weight.  That is, they estimated how much more likely someone is to die in a car accident if the opposing vehicle weighs a little more.  I'm going from memory here, but I recall the number was something like a 50% increase in the odds of fatality for a 1000 lb. increase in vehicle weight.  They estimated this using a huge database of actual vehicle-on-vehicle collisions and the estimate seemed amazingly robust.  (Still, I need to read the paper...)

Using EPAs measure for the value of a statistical life (something like $5.8 million/life) and information on vehicle mileage, there were able to convert that weight externality into a near-equivalent gasoline tax.  That tax didn't exactly match an appropriate tax on weight, but it turned out to be extremely close.

The take home number:  $1/gallon.  

That's a huge number. Before this study the conventional wisdom among transportation economists was that the largest driving-related externality was congestion, at something like $0.55/gallon.  Pollution externalities, including CO2, come in at about $0.33/gallon. These are rough numbers from my recollection.

Can we start taxing gas more heavily already?  It's not as if we don't need the revenue.


  1. Surely "no less safe" is an overstatement, since over half of automobile occupant deaths occur in single car (or truck, SUV) crashes. Naturally, given the arms race mentioned, of those who do die in trucks or SUVs, they are more likely to die in single-vehicle crashes.

    However, looking at the statistics, you can see that the death rate in single-vehicle crashes is much lower for SUVs and for large cars than for smaller cars.

    So I think that the evidence actually argues against a tax on weight. In any case, a paper that focuses solely on car-on-car collisions (and perhaps this one investigated single-car accidents as well) is woefully incomplete.

  2. Of course, a countervailing set of data would be to consider pedestrians, cyclists, and other non-occupants, who are worse off when cars are larger. My understanding of the statistics is that the number of fatalities among non-occupants is significantly lower (though it does affect behavior.)

  3. John: There's no externality with single car collisions. There is with car on car collisions. So, what matters is just the odds of a car-on-car collision. That's what the authors measure.

  4. With gasoline prices as high as they are now, this does not seem like a prudent thing to do. You can push the externality problem too far in my estimation. There are externalties all around us, both negative and positive. The idea that Pigouvian taxes/subsidies can take care of the problem is naive. It ignores the fact that we cannot really measure the externalities to any degree. There are so many problems we have to deal with as a country now this seems really incidental and a waste of time. The largest problem we have is reigning in an irresponsible government. The biggest problem we have is not market failure but government failure. This country is on the verge of bankruptcy and having the dollar's sovereignty being subverted and you want to talk about taxing gasoline more? Come on!

  5. Verge of bankruptcy when interest rates are 0%?

    You'll have to read the paper for yourself, but the numbers looked unusually solid to me.

    I don't think one can take basic economics too far. We all like income, right? So why do we tax something we all like and refuse to tax things we all dislike: death, congestion and pollution.

    Obviously it would make sense to adopt a tax like this gradually.

    I'm not expecting this kind of thing to happen. There is no political will, given most people have the same knee jerk reaction you do. I u
    Understand Greg Mankiw got more than an earful for suggesting a gas tax to former President Bush.... Oh well..

  6. Oh, and a gas tax probably would not increase prices nearly as much as the tax. Supply is fairly inelastic, so the Saudi's would pay a decent share of it.

  7. Have you seen this? Top safety ratings, low gas consumption and a little extra weight:

    I don't think this is a way around your proposed tax, but it does indicate that these more efficient cars might help reduce the weight gap from the all gas SUVs.


Post a Comment

Popular posts from this blog

Renewable energy not as costly as some think

Answering Matthew Kahn's questions about climate adaptation

Nonlinear Temperature Effects Indicate Severe Damages to U.S. Crop Yields Under Climate Change