Raising MPG Standards: The Second-best Solution to a Gas Tax Increase


It got surprisingly little press coverage given the degree to which it will affect our lives (thanks, pesky world economic meltdown), but in case you missed it, the Obama administration recently worked out a compromise with the major automakers that will dramatically raise the corporate average fuel economy (CAFE) standards.

The new regulations mandate that the mix of new cars sold in the year 2025 must achieve about 54.5 miles per gallon (though if you read the fine print you’ll see that credits for various other green innovations mean that actual fuel economy will be more like 40 MPG.) For reference, the auto fleet currently on the road gets about 27 MPG. It’s a well-done agreement that will help avoid well-done citizens as global warming accelerates.

Before proceeding, let me note that I am strongly in favor of this policy. The problem of excessive fossil fuel use in transportation is multidimensional: if the issue of global warming doesn’t move you, the thought of Hugo Chavez and Mahmoud Ahmadinejad using our own hard-earned dollars to tweak our geopolitical noses should.

However, it is worth noting that raising CAFE standards is what political scientists and economists call a “second-best” solution; we could be doing considerably better if we thought all of this through more clearly.

This is not because CAFE doesn’t work; it does. In 1975, a few years before CAFE was implemented, average MPG for new cars and light-duty trucks was 13.1. In 2010 it was 22.5. Can this be attributed to CAFE? To a large degree, yes, as this graph makes clear:

CAFE standards were aggressively increased from 1978 to 1984, and, as the chart above shows, fuel economy responded. However, from 1985 until 2007 CAFE standards were no longer raised meaningfully—and MPG flatlined. The table makes it pretty clear that the CAFE standards created a floor under MPG for a 25-year period, when low gas prices (remember those?) rendered consumers otherwise indifferent to fuel economy.

So what’s the problem with raising CAFE today?

There is a long history of debate on whether “command and control” regulations (like raising CAFE standards) are a good way to bring about change. The other option is the use of price signals—which in this case would be increased fuel taxes—to influence consumer behavior.

Regulations do have some attractive features. For example, we can directly target what, when, and how much improvement we are getting. If we want fuel economy of 55 MPG, we can decree and achieve it with greater certainty than if we try to monkey around with prices.

However, in theory at least, economists generally prefer to do things with price signals as opposed to regulatory standards. Why?

Price signals inflict pain on consumers, but let them figure out what form they want to take it in. They in turn force producers to respond to their (altered) demand, but allow producers leeway in how that demand is met. This allows consumers and producers to change behavior in the most efficient possible manner.

Instead of CAFE, why not just raise the gas tax and let drivers figure out whether they want smaller cars, lighter cars, less powerful cars, more expensive cars, shorter-range cars, or, crucially, cars that are just as heavy, powerful, and cheap—but which get driven less?

This raises the true problem with CAFE. It misses out on a potentially key part of the solution to reducing fuel use: driving less. In fact, ironically, increased CAFE standards will have a perverse and unwelcome effect; better fuel economy will increase the fixed cost of driving (i.e. vehicle prices) but will actually reduce the marginal cost (i.e. fuel expenditures). To a degree, less thirsty cars will actually cause people to increase the number of miles they drive (as I’ve written about here).

With increased gas taxes, on the other hand, less driving will be part of the consumer’s toolkit. Some who absolutely need vehicles with poor fuel economy will have the option of avoiding the tax by driving less instead. As long as their fuel use goes down, why not give them that choice? Greater economic efficiency would result. In fact, the Congressional Budget Office ran the numbers in 2004 and found that cutting fuel use through taxes was considerably cheaper in the long run than raising CAFE.

Reducing driving through a higher gas tax would have other important benefits that improving fuel economy does not, like congestion relief and accident reduction. I personally am more sympathetic to automobility than most of my colleagues in my field, and I have faith that technological ingenuity will deal a powerful and probably decisive blow to our emissions problems. But raising the price of driving above current levels is pretty much a no-brainer; it has support that stretches across ideological lines in the transportation field, even among those like me (and even among carmakers such as GM) who do not see exchanging cars for biodegradable pogo sticks as the only possible solution to our transportation problems.

Another advantage of a gas tax increase is that it would start working today. Since the car fleet takes so long to turn over (according to the US Department of Transportation, automobiles these days stay on the road an average of about 12 or 13 years), it will be a very long time before the new CAFE standards actually translate into meaningful changes in emissions. But increasing the gas tax would have immediate effects.

(Some might object that fuel taxes are regressive and would hurt the poor, and to an extent they would be right. However, the rich drive considerably more than the poor, taking some of the stink off. And paying for many of the new fuel-economy technologies CAFE will result in will be regressive too.)

Thus CAFE might be a second-best policy: good, but not as good as we could have. Then why are we using CAFE while gas taxes stay laughably low by developed-world standards? Obviously, and understandably, because voters hate taxes. If anything, the political winds are blowing towards a lower gas tax, not a higher one.

Spokane Al

I wonder about the additional impact on food and everything else that often travels on roads to reach our tables/stores etc if gas taxes are raised. By only measuring the impact on driving due to increased taxes, aren't we missing a large piece of the puzzle?


Fuel efficiency of US cars is terrible compared to those in Europe. I was just in England for a couple of weeks and drove one of the larger VW models (since I had 4 passengers and a lot of luggage). I averaged 39 mpg, driving mostly on small one lane roads. My Dad is currently driving a Citroen (diesel) and averages 52 mpg with very little highway driving. This totally kicks the butt of the similarly sized "eco-wonder" Prius which has a 42 mpg rating.

The technology is obviously already available and without the added cost of electric/hybrid engines.

John B

A good point---but you are missing the big picture.

The reason European cars get better mileage is that they have fewer mandates and regulations than the US. You cannot legally drive the European cars in the US.

The US "Smart Car" gets much better mileage in Europe because the one sold here has to be modified to meet additional US requirements.

I am not saying the US modifications are not good--but there is a price. Much lower fuel mileage.

Allowing the cars made in Europe and the far east to be sold here would do more to increase gas mileage than any CAFE standard and any gas tax.


While raising the gas tax would result in less miles driven, that still doesn't change the political reality of the situation. Very few people in /any/ Legislative session (regardless of whether any particular party is in control) would be willing/able to raise the tax. Regardless of the logical/rational desires to do so.

On the other hand, CAFE standards can be passed with relatively less opposition. While it may be in your opinion the Second Best option, it's the option that will actually be accepted. And what this also means is that in the event that a compromise can be reached legislatively, you could also raise the fuel taxes. While it'd have a lessened impact than solely a tax increase, it would still change the math involved so that you could have a easier transition to more or equal mobility with less environmental harm.


I have another idea. A crazy one. But brilliant.

Reduce the national speed limit to 45 mph. Yep, you heard me.

Not only will this save a lot of fuel, a lot of money, and a lot of lives, but maybe, just maybe, slowing down and smelling the roses (so to speak) might serve us well.

I remember that I had gotten so used to Carter's 55 mph that I felt like I was breaking the sound barrier when I moved back up to 70 mph. We've likely wasted thousands of lives in our "need for speed."

Let's just slow down for a few years and see what happens. Going faster hasn't particularly helped us, has it?


The author seems to believe that people driving less is automatically good. The choices brought about by higher taxes impose real costs to people that are not monetary. If you raise taxes, people will have the anguish between paying gas prices and not driving. Say I didn't want to pay the taxes and so I chose not to visit my family. That makes me sad and is a cost to me that is not monetary.

Economists need to realize that making the most efficient choice does not necessarily mean that the utility of society is better off for it. There are costs to choosing and knowing what you are going without. Choice, even if it makes us more efficient, is not necessarily best for our society. (See: Paradox of Choice)


The gas tax approach won't work well when cars are not gas powered, i.e. electric. Various approaches to mileage-based taxes have been floated over the past couple years (in Oregon or Wash. St. I believe) and have been shot down, but eventually the mileage-based tax will probably win out. The gov't wins with it's CAFE approach because it (1) increases nannystatism while appearing to look serious, and (2) doesn't tie itself to the soon-to-die gas tax.


For it to be effective, the gas tax need not be any more than what it would take to cover all the external costs of gas consumption (pollution, foreign wars, damage to the environment ...). Other forms of energy would be taxed (or not) based on their own external costs (e.g. tax the fuel used to power a coal or oil based electric plant).

There are, of course, external costs of just having that vehicle on the road regardless of the energy source. So a perfect model would need to account for miles driven (and probably the weight of the vehicle and other factors that raise the external costs). Technology should be able to handle that efficiently - requiring each vehicle to have a GPS monitor would probably be cheaper than implementing and enforcing the somewhat arbitrary CAFE standards.

Mark H

Even as far as top-down regulations go, CAFE standards are extremely ugly: when people get better mileage they drive more, which increases traffic accidents (and fatalities), congestion, and even some emissions, like NOx and carbon monoxide. This policy trades one externality for several larger ones.

Contrary to what Morris is saying, if it’s not politically feasible for the government to enact a fuel tax (or cap & trade), CAFE is not a worthwhile second-best. Taking all externalities into consideration, cost-benefit analyses of CAFE come up extremely negative.

For more info, Andrew Kleit has done some great research on CAFE. Here’s one of his papers on the subject: http://www.freepatentsonline.com/article/Economic-Inquiry/115635786.html


Can this idea be tested by looking at countries where tax on fuel has actually been used as a means to lower fuel consumption. The Netherlands have been trying for quite some time. For your information, a gallon of petrol costs about 8.80$ a gallon. The difference between the price in the Netherlands and in the US is probably for over 90% caused by taxes.

Eric M. Jones.

....and when the population reaches 10 Billion?.....15 Billion?...20 Billion?


Good post, but I would switch : "the thought of Hugo Chavez and Mahmoud Ahmadinejad using our own hard-earned dollars to tweak our geopolitical noses should." to...

"the thought that the US toppled a secular democracy in Iran in the early 50's and instated a religious totalitarian regime should."

Justin Credible

Yes, lets tax and regulate until our overlords get us to act the way they want us to act....hey were not forcing you to do anything, we are just limiting your options...one wonders when they will just rename it the Peoples Republic of America.


Like most people who like government controls you lean toward the stick instead of the carrot. You can change behavior by penalizing people who do what you don't like. Another way is to reward behavior you do like. We subsidize farms for growing desired products, oil companies and airlines so why not subsidize consumers for making your desired choice?

Going to regulations to force manufacturers to build what you think is good or penalizing buyers for not making what you think is a good choice is the mark of the person who thinks they have the right to dictate to others. Make the price of your favorite choice less than the behavior you don't like allows people to spend more for their "objectionable" behavior but spend less to conform to your idea of what is "better".