A closer look at the benefit cost analysis of the proposed CAFE standards

The WSJ's article on the proposed CAFE standards summarizes the draft Regulatory Impact Analysis:

The administration estimates that complying with the rules would cost manufacturers about $6.3 billion over four years and add $54 to the cost of a 2008 truck and $275 to a 2011 truck. The benefits for the same period are estimated at $6.5 billion to $7.5 billion, depending on how quickly auto makers switch to the new system.

The net benefits are only $500 million to $1.2 billion, which isn't that big an improvement in social welfare. The low range is kind-of close to a bad deal. The methodologies used to value benefits, I'm guessing "off-the-shelf" estimates (aks, benefits transfer, we'll have a future post on this) are used, leading to much uncertainty about the true range of estimates. So, I decided to take a closer look at the benefit cost analysis.

It's a longish read so what I find will trickle out slowly. First, let's deal with the benefits of the proposed CAFE standards.

Right off the bat, one problem with the WSJ summary: I'm not sure where the WSJ is getting their aggregate benefit numbers. Table VIII-12 on page VIII-43 of the RIA summarizes the "present value of lifetime social benefits" of "reformed CAFE"[1] for model year 2008-2011. At a 7% discount rate the benefits are $7.5 billion. So far so good. At a discount rate of 3%, however, the benefits are $9 billion. So where do they get the $6.5 billion?

I dunno. But the RIA does a comprehensive job of figuring the types of benefits from CAFE. Consumers will:

• spend less money on fuel
• drive more (a benefit because we enjoy driving and going places)
• refuel less, saving time

On the other hand the increased driving will increase congestion costs, traffic noise and decrease safety. The pollutant volatile organic compounds (VOC) will increase.

Other pollutants, NOX, PM, and SOX, will decrease.

Society will also enjoy a reduction in oil market externalities. The idea is that the U.S. is such a large buyer of oil that our demand drives the world oil price upwards. Any reduction in our demand will reduce the world oil price on every barrel that we purchase. These reduced expenditures are a benefit of CAFE.

Having done a complete job of estimating a wide range of net benefits, the analysis finds that the reduction in fuel expenditures is not much different than 100% of total benefits. Everything else cancels out. A bit of an anti-climax, eh?

Oh, and one other thing[2]. The fuel expenditure benefits are based on a gas price of $1.50/gallon. The reformed CAFE standards are projected to save about 10 billion gallons of gasoline over 20+ years. The nominal savings are $15 billion. If gas is $3/gallon the nominal savings are $30 billion. A good guess is that the discounted benefits of reformed CAFE will also about double with higher gas prices making the net benefits of CAFE larger than the $500 million to $1.2 billion that I worried about above.

Next, sometime next week I hope, I'll take a closer look at the cost estimates. After that, we'll put the costs and benefits together and look at their uncertainty analysis.

[1] "Reformed CAFE" refers to the proposed rules relative to what will happen anyway in model year 2008-2010. I was unaware that "unreformed CAFE" will be applied to model year 2008-2010 light trucks. So, is this a new CAFE applied to light trucks or the status quo? I dunno? Any help out there?

The benefits of unreformed CAFE are $4.8 and $4.0 billion discounted at 3% and 7%. The incremental benefits of reformed CAFE are only $4.2 and $3.6 billion. Most of these benefits, about 75%, are due to the 2011 model year. If unreformed CAFE is extended to 2011 there is not much difference between the two. That means that there is not much difference between reformed CAFE and the status quo. This lends some credibility to those who say that the proposed CAFE standards will actually worsen overall fuel economy by ignoring Hummers and other mammoths.

[2] Oh, and still one more thing. The benefits don't include the value of reducing greenhouse gas emissions. The environmental assessment (Table 4-3) finds that CO2 emissions will fall by 34 million tons (2.6% of the current light truck output). The value of this is positive and would increase net benefits. Since they were very conscientious about develop monetary impacts of the other factors, I wonder why they didn't try to develop monetary estimates of this impact? Maybe we're still waiting for that zero-emissions technological fix, developed via subsidies from the energy bill, to kick in and solve global warming rending the valuation of CO2 impacts for light trucks meaningless (sarc).

Comments

[allen claxton]

drive more (a benefit because we enjoy driving and going places)

Even though this makes total sense, I've never thought of this as an outcome of fuel economy standards. I'm totally converting to your gas tax monomania!

Out of curiosity, does the analysis make a distinction between the derived demand for transportation, and driving for its own sake?

The benefits don't include the value of reducing greenhouse gas emissions. The value of this is positive and would increase net benefits.

Doesn't this depend on the extent of the increased travel due to lowered cost of travel? That is, couldn't efficiency reductions easily be overwhelmed by increasing miles traveled? It's a pity the report doesn't address the topic at all.

[joshua corning]

holy crud a graph...yeah i want more graphs...graphs are cool. :)

"add $54 to the cost of a 2008 truck and $275 to a 2011 truck."

John didn't we have a discussion about how small the penalties were for not meeting the CAFE standards. What I think is going on here is that much of the heavy lifting (R$D) has already been achived in thier development of fuel efficiant cars and hybrids...and now it is fairly easy to implement the same technology into thier smaller trucks and SUVs.

[Tim Haab]

"yeah i want more graphs...graphs are cool."

Based on my classroom experience, you are in the vast minority.

[Robert McLeod]

I know you've talked about pay as you drive insurance before. What about the feebate? I don't see any addressing of it here, at least through the Google search engine.

E.g.

DL Greene et al., Feebates, rebates and gas-guzzler taxes: a study of incentives for increased fuel economy, Energy Policy 33 (2005), pp. 767-775.

http://dx.doi.org/10.1016/j.enpol.2003.10.003

[M1EK]

Sign me up to the "gas tax uber alles; dump CAFE" movement too. It bugs me how many people blithely claim that gas taxes are regressive when the (few) studies out there seem to indicate otherwise.

http://mdahmus.thebaba.com/blog/archives/000186.html

[richard]

This is a number game. I think it has no value. It assumes the situation to remain more or less what it is. Only in the US people think you have to drive an SUV. The rest of the world thinks you don't.

Let's say gas prices go up. GM can't provide smaller cars, because they have no money, so does Ford. Toyota can, and in very large quatities.

The best selling car of Toyota is the Vitz, 1.3l engine, 40mpg. Pretty cool car, says I (6 foot 6)

You don't need new technology to increase CAFE, only GM/Ford need it. The rest of the world already has it.

Please factor in what happens if GM or Ford go bust. Because good bet that is what is going to happen.

Just my 2c's, delete it if it makes you laugh ;-)

[allen claxton]

gas taxes are regressive when the (few) studies out there seem to indicate otherwise.

Can you recommend a study or two that demonstrates that?

[M1EK]

Allen,

Read my links - I link to two in one of those two articles.

[joshua corning]

Richard,

"Only in the US people think you have to drive an SUV. The rest of the world thinks you don't."

Few things on this...the US is sparcly populated compared to most of the developed world and even compared to much of undeveloped world...this is a fact and in order to move people and goods around such a large area you have to use gasoline...the is no other way.

SUV's are not responsible for all of the US oil consuption. In fact 40-50% of all oil consumed in the US goes to non-transportation uses...It is nice that France Germany and Japan use nuclear power for electrical energy production it is to bad that Enviornmental groups in the US have stopped such facilities being developed in the US.

M1EK,

good to see you. :)...I think I am part of the dump the CAFE standards and end the gas tax crowd... i think here it is a crowd of one. :)...you are always posting on reason.com...shouldn't you be in my crowd?

[John Whitehead]

Some replies:

Allen, the report uses estimates of the "rebound effect," the increased miles driven, and does not distinguish for the purpose of the additional miles. The greenhouse gas emissions account for these additional miles (I'm sortof sure).

Joshua, $54 and $275 are price increases, not penalties.

Robert, I'll have to get back to you on the feebates, I have no idea, maybe someday.

Several, gas taxes are regressive. So are sales taxes. Environmental (and international trade) policy is a terrible place to indirectly handle equity. Much easier to consider this via the income tax.

Several, I'll try to remain agnostic on SUVs (at least today). They use more fuel than is entirely necessary in many parts of the country (i.e., flat parts). Full disclosure: I drive a little one with AWD, a Forester, up big hills every day. I kinda need it, especially in the winter. Everyone in town needs one. In Wilmington NC, flat city, where I used to live, not so much (we had 2 wheel drive vehicles).

[M1EK]

"Several, gas taxes are regressive."

No, they aren't. If you have studies that show so, post them. I've already provided links to the contrary.

Ref: SUV's:

"Everyone in town needs one."

What did they do before SUV's? Die every winter and come back like perennials?

[John Whitehead]

Here is one study that finds that gas taxes are regressive:

Estimates from a Consumer Demand System: Implications for the Incidence of Environmental Taxes, by West and Williams, Journal of Environmental Economics and Management, May 2004.

And the abstract:

"Most studies suggest that environmental taxes are regressive, making them less attractive policy options. We consider the distributional effects of a gasoline tax increase using four incidence measures and under three scenarios for gas tax revenue use. To incorporate behavioral responses we use Consumer Expenditure Survey data to estimate a consumer demand system that includes gasoline, other goods, and leisure. Our estimates confirm that when revenues are not recycled, a gasoline tax is regressive. Use of incidence measures that ignore demand responses, however, will substantially overstate this regressivity. In contrast, the differences between the equivalent variation and easier-to-implement consumer surplus measures are relatively small. In addition, our results suggest that using the additional gas tax revenue to fund labor tax cuts makes the policy substantially less regressive while using the revenue to fund lump-sum transfers actually makes it progressive."

[M1EK]

John,

Your citation is fairly moderate in its assertion that gas taxes are regressive - pointing out as many others have that almost any other tax (payroll taxes, sales taxes, property taxes - anything except federal income tax as far as I can tell) is worse. I come to this conclusion because most of them say "yes, higher gas taxes with no other changes may have a small regressive effect, which we can counteract with reductions in (those other taxes)". If the other taxes were LESS regressive, that wouldn't work.

So even the guys who agree that the gas tax is somewhat regressive view it as one of the better of the lot. Certainly less regressive than paying for roadways out of sales and property taxes, which we do all over the place in this country.

http://www.bepress.com/fullertonwest/

"A tax on gasoline is not regressive across the lowest incomes but is regressive from middle to high incomes."

[Rob Williams]

Thanks for plugging Sarah West's and my paper. It's always nice to see someone is actually reading my articles.

By the way, I disagree about the effect of higher gas prices. You said "If gas is $3/gallon the nominal savings are $30 billion. A good guess is that the discounted benefits of reformed CAFE will also about double." That statement implicitly assumes that without CAFE, a doubling of gas prices would have no effect at all on consumer behavior, even in the long run--and that clearly isn't correct.