N. Gregory Mankiw may have been Chairman of the President Bush’s Council of Economic Advisers but he is seriously confused about the relative economics of fuel economy and carbon taxes. In today’s New York Times, he repeats a variety of standard myths and makes one classic blunder.
Why does he prefer taxes to raising fuel economy standards?
A carbon tax would provide incentives for people to use less fuel in a multitude of ways. By contrast, merely having more efficient cars encourages more driving [the "rebound effect"]. Increased driving not only produces more carbon, but also exacerbates other problems, like accidents and road congestion.
Uhh, nice try. But if having more efficient cars encourages more driving, then why has driving — vehicle miles traveled (VMT) — soared in the past two decades while the average fuel economy of US vehicles has actually declined? The answer is that people drive more mainly because they have gotten wealthier. It is a myth that the rebound effect is significant, as this recent study makes clear.
But won’t a carbon tax cut gasoline consumption? Not likely. The only carbon tax that Mankiw cites is $15 per metric ton of carbon dioxide ($55 per metric ton of carbon). That would add a whopping 14 cents to the price of gasoline.
How high would gasoline prices have to be increased through a carbon charge to significantly change the average fuel economy of U.S. cars, which currently averages some 20 miles per gallon for all vehicles and 27 mpg for new cars?
Consider that European countries have taxes of more than two dollars per gallon, which is five times more than the U.S. tax. Yet, as of 2002, the average fuel economy of European Union vehicles was 37 miles per gallon. Moreover, some of that fuel economy improvement was achieved not just with high fuel prices, but with strong tax incentives to promote diesel vehicles, which are typically more fuel efficient.
A carbon tax equal to at least $1 per gallon would probably be required to achieve the same oil savings as the kind of increases in fuel economy currently being considered by Congress. Yet that would require a tax of more than $100 per metric ton of carbon dioxide – far more than anyone considers plausible today. In Bingaman’s climate bill, for instance, the safety valve starts at $12 per metric ton of carbon dioxide.
The notion that a carbon tax — or even a cap-and-trade system — can possibly substitute for higher fuel economy standards is a major blunder for an economist to make.
Mankiw, of course, like all classical economists, believes the economy must be humming along at its maximum efficiency today. He repeats the tired myth:
And don’t expect savings on gas to compensate consumers in a meaningful way: Any truly cost-effective increase in fuel efficiency would already have been made.
Why? Because car companies have already designed their cars to optimize efficiency vs. price? Yeah, that’s what GM and Ford have been doing all these years of losing market share to more fuel-efficient Toyota. Of course there are many cost-effective technologies to boost vehicle fuel economy.
With this guy running the CEA, no wonder Bush’s energy policy has been such a dud. Now Mankiw is advising Romney. Good luck, Mitt!
Previous in TP Climate Progress
Language Intelligence: Lessons on persuasion from Jesus, Shakespeare, Lincoln, and Lady Gaga
The economic explanation for why carbon taxes do not curb gasoline consumption is that gasoline demand is relatively inelastic to price (http://en.wikipedia.org/wiki/Price_elasticity_of_demand). That means as costs go up, demand remains relatively constant. It is unclear exactly what price would dissuade people from driving because we have not hit a price in the US where this has happened. But it would definitely be at an offset cost greater than $100/ton. That makes both a cap-and-trade system with offsets on the high-end at $100/ton CO2e and a carbon tax at $1 per gallon ineffective at reducting consumption, as you say. This is a great argument for both alternative fuels and for fuel economy standards.
I have blogged at Local Warming (http://local-warming.blogspot.com/2007/08/gas-price-demand-and-carbon-taxes.html) about how gasoline consumption is relatively inelastic to price. The short-term Price Elasticity of Demand for gasoline as stated by PROPONENTS of a Carbon Tax at http://www.carbontax.org is 0.1. That translates to: if prices increase by 14 cents from $3.00, you only decrease demand by about 0.04%. These types of taxes also disproportionately affect the poor, who spend a larger portion of their dollar on gasoline and products transported using gasoline (like food).
The factor that I’m wondering about in the elasticity and inelasticity of gasoline is time. People don’t change vehicles when there is a price increase in gasoline, but when the next new vehicle cycle is for themselves. Then they might pick a more efficient vehicle.
Actually I think both is the better answer. Both an increase in the carbon tax and regulation in better MPG.
There is a difference between the short-term price inelasticity and the long-term. Long-term elasticity is 0.4. I got these estimates from http://www.carbontax.org. My question is-if a tax is designed to be revenue-neutral, and it doesn’t change demand, then what is the point?
Joe, unfortunately, I don’t think this is one of your better pieces.
To begin with, Mankiw himself is calling for a gas tax of $1 per gallon, phased in at $0.10 per gallon per year. Although that particular is left out of today’s NYT article, he said this in the WSJ last October. See here: http://gregmankiw.blogspot.com/2006/10/pigou-club-manifesto.html
In that earlier WSJ article, he articulates that this tax could be implemented as part of a broader carbon tax, but in any event, his initial target for the gas tax is $1/gallon.
Most of today’s article’s thrust is regarding the implementational advantages of a carbon tax versus regulations or cap & trade, etc., the likelihood of achieving (somewhat harmonized) international accords, and other issues.
One of those advantages not stated in today’s article is that you can relatively easily adjust the level of the carbon tax “on-the-fly” to achieve the desired level of consumption (reduction) as results from the marketplace respond. That certainly seems more effective than negotiating a new set of country caps, etc.
There is of course nothing preventing CAFE standards or incentives to complement to policy objectives of a carbon tax, if you felt is was still required.
I could go on, but I think that your challenges to one of the better tools in the policy toolbox were not fully-informed nor persuasive.
(Big fan most of the time!)
Good catch, but I stand by my post. Had Mankiw said any of what you write in this article, it might have made some sense. As it is, the piece was devoid of economic consistency. If he were a blogger, he could have linked back to the first piece.
I still think Mankiw makes no sense. A $1.00 tax would likely modify behavior 3%, assuming that short term PED for gasoline is relatively constant. The amount of tax required to modify behavior would have deleterious effects on the economy and people at the margins of society. The tax does not set a limit on emissions. As we have seen, fuel economy has not been improving over time. At best a tax on its own could stimulate alternative energy and provide pork for someone’s politics.
Here are a few observations on the subject of gasoline price and demand.
The U.S. did an experiment almost 30 years ago on what gasoline price it took to affect demand in a small way. In inflation adjusted terms, we are approaching that price again. According to one website, the annual average inflation adjusted price in 1980 was $3.06 in 2007 dollars. However, inflation adjusted gasoline prices do not tell the whole story. The fraction of one’s income spent on gasoline is relevant too, and probably of greater significance, and by that measure, gasoline would have to be well over $4 per gallon to have an effect similar to 1980.
If Joe’s figures on European taxes and MPG are correct, it is interesting that the average U.S. driver and average European driver can each go about 740 miles on $100 of gasoline. The extra efficiency cited (37 MPG) almost exactly cancels the extra $2/gallon of tax. This is similar to the situation with electricity: Californians pay more but use it more efficiently than the rest of the U.S., with the end result that the average bill is similar,
but the average annual greenhouse gas emissions are dramatically lower (about 3.7x lower per person).
How to handle the question of the poor.
In my state we have a sales tax on vehicles, used and new. That just raises the cost of vehicles for the poor also. Lets reduce the sales tax by exempting the first $500 of the tax for every vehicle purchaser. The money saved in sales tax, if it went for a more efficient car, could save fuel as well. I read somewhere that if $2000 more was put into the fuel efficiency of a Taurus, it would give the car 10 more MPG.
The point would be to reduce other taxes that the poor would pay. Reduce their property, sales, social security or income taxes. What would be the point of any of these taxes for the poor? What problem is to be solved? Carbon taxes should be the first taxes that anybody pays around the world, because this is an around the world problem.
Even in poor countries, there is still a need for taxes at some level to pay for government and government services. In a selection of all the taxes that can be made by a government, a carbon tax should be the first one to impose on people.
Shannon: I agree on the tax.
Earl: Nice analysis.
Ronald: I agree one must design policies not to penalize the poor, no matter how the price of carbon dioxide is increased.
Shannon: Thanks for citing my Carbon Tax Center Web site. But I’m afraid some of your estimates of the effects of different-level gasoline or carbon taxes on gasoline consumption are off considerably. Using price-elasticities of -0.1 (short-term) and -0.4 (long-term) from the Small-Van Dender paper that Joe and I both rely on, a 14 cent/gal tax would reduce usage by 0.5% (short) and 2% (long), while a $1/gal hike would reduce usage by 2.8% (short) and 11% (long). Contact me via my Web site if you want the derivations. In any event, it’s the long-term results that matter.
Joe: Okay, you’ve hoisted Mankiw on his own petard (a too-modest carbon tax), but now what? Let’s focus on Tidal’s point that the tax level can be adjusted once it’s been proven to be both effective and equitable. The principle of taxing carbon matters more at this stage than the level.
Joe again: I’m not sure of the point of your US-Euro comparison, and besides, the sponginess of the mpg figures (are they “sticker” or “real”) makes the comparison dicey. The more important point is that Europeans consume half as much motor fuel per capita as Americans, and that’s due to a considerable extent to their higher pump prices. Indeed, Europeans’ lesser VMT contributes mightily to their higher quality of life.
“Uhh, nice try. But if having more efficient cars encourages more driving, then why has driving — vehicle miles traveled (VMT) — soared in the past two decades while the average fuel economy of US vehicles has actually declined? The answer is that people drive more mainly because they have gotten wealthier.”
Uhh, nice try. The fact that people drove more while fuel economy declined does not rule out increased economy as “an” additional incentive to drive. People drive more because:
- they have more money or, rather, gas takes less of their disposable income and one part of that is fuel efficiency
- good highways make it easier
- cars are more fun (ever take a long trip in a 1950 Ford?)
- it’s a status thing
Nah. That’s why I put in a link with real numbers. Fuel efficiency has a small effect. Wealth has a big effect. Not sure how driving more is a status thing.
Status? Being seen behind the wheel is an important social statement to many people. Reversing that perception might play a significant role in reducing emissions.
Yes. In Europe the two dollar gas tax has had a limited effect on gas mileage statistics. It may be important to note that Europe has less sprawl (requiring shorter commute times), better public transportation, fewer cars per household and significantly less carbon use per person than in the US. You might have to raise gas taxes a lot in the US to get a high net drop in usage and increased demand for efficient vehicles. If the taxes are coming back to citizens in the form of reductions to payroll taxes I don’t understand why this is off the table. Will taxing poor people who commute long distances in fuel-inefficient vehicles affect them negatively? Yes. But so will forcing them to buy a new hybrid car. Any legislation designed to reduce carbon is going to cost Americans money and be especially hard on the poor. Personally, I like the carbon tax because it attacks the problem we are trying to solve head on and it lets each person decide how best to change their behavior to get there. I think that a US wide tax on carbon would be a huge incentive for companies to sell hybrid cars which they are beginning to do already without the tax.