on July 28, 2008.
At home and abroad, bashing ethanol has become an increasingly popular sport. Earlier this year, Time ran a cover story lambasting politicians and big business for pushing ethanol ("The Clean Energy Myth," Apr. 7, 2008). And at the G-8 summit earlier this month in Hokkaido, Japan, World Bank President Robert Zoellick called on leaders to reduce and reform their ethanol subsidies ("US and EU Urged to Cut Biofuels," BBC News, July 7, 2008).
Yes, it's true that the 54-cent-per-gallon tariff is blatant protectionism and that it only deepens our energy crisis by cutting our access to cheap, clean Brazilian ethanol. Yes, there is little environmental justification for the 51-cent-per-gallon credit, because the reduction in greenhouse gas emissions from burning corn ethanol instead of gasoline is small or even, some say, nonexistent. Yes, compared with U.S. energy needs, useful cropland is limited, and ethanol refiners' voracious appetite for corn puts pressure on food prices.
Those unpleasant facts take the luster off ethanol's once-golden image. But we should not get carried away. There is still an ironclad case for a permanent subsidy for ethanol.
Imagine if all U.S. cars could get 60 miles on a gallon of gas. In effect, that's the promise of ethanol in terms of America's energy security — if our vehicles could be converted to run on E85 (85 percent ethanol, 15 percent gasoline) and we could produce enough ethanol to fuel them. U.S. petroleum consumption would be halved, and the billions of the dollars we send to the Middle East would instead end up in the Midwest. Helping to make this vision a reality is an easy decision for Capitol Hill politicians. And so they give U.S. ethanol a 51-cent tax credit for every 1 of the 6.5 billion gallons produced in 2008.
But let's be clear: Injecting money into downtrodden farm communities is not by itself a justification that economists can endorse on principle. It is a welcome byproduct that happens to coincide with compassionate politics. It is the political icing on the policy cake. But if there were no improvements for the overall economy from ethanol tax subsidies, they would be nothing more than a handout.
The economic justification for subsidizing ethanol is the necessity for government intervention in the market to promote energy security. Petroleum use imposes costs on our economy and national security that are not taken into account in the free market. The U.S. policy of tilting the playing field in favor of petroleum substitutes is one approach to correcting the problem.
How much energy security does the government get for its money? Table 1 on the next page provides part of the answer.
Panel A of the table shows how much E85 is needed to replace one gallon of gas. Ethanol has less energy than gasoline, so the volume of fuel must increase. Total ethanol (in the 1.38 gallons of E85 needed to equal one gallon of pure gasoline) is 1.17 gallons. Gasoline consumption out of the fuel tank is reduced to 0.21 gallons. Taking into account petroleum used in the extraction, refining, and transportation of the fuels, total gasoline from the fuel switch declines from 1.12 gallons to 0.29 gallons — a 0.83-gallon cut in the use of gasoline.
Panel B shows the financial benefit of switching to E85 when there is a 51-cent-per-gallon ethanol subsidy. The 1.17 gallons of ethanol provide 59.8 cents of tax credit. And note that the credit per gallon of gas saved is 72.1 cents.
Panel C shows the financial benefit of switching to E85 when there is a 72.1-cent-per-gallon gas tax. The switch reduces the gas tax from 80.6 cents to 20.8 cents — a saving of 59.8 cents — the same as under the ethanol subsidy.
Because the dollar savings per gallon is the same for both the 51- cent subsidy and the 72.1-cent tax, we can think about them as being equivalent.
Imposing an excise tax on gasoline is the textbook answer and the right answer to the question of what is the best policy response to energy insecurity. (Note to aspiring economists: We are imposing a Pigouvian tax on a negative externality.) But even though we say the subsidy is "equivalent," it is not the same as a tax. And the differences between the subsidy and the tax eat away at the economic benefits that could be achieved under the economically preferable gasoline tax.
Table 1. The Arithmetic of Ethanol Subsidies and
A. Substituting E85 for gasoline — physical aspects
Gasoline (gal.) 1.00 0.21
Ethanol (gal.) 0.00 1.17
Total fuel (gal.) driving 1.00 1.38
Gasoline for production (gal.) 0.12 0.08
Total gasoline (gal.) 1.12 0.29
Gas saved (gal.) 0.00 0.83
B. Effect of fuel-switching under a 51-cent ethanol subsidy
Ethanol subsidy (per gallon) 0 51 cents
Ethanol subsidy — total 0 59.8 cents
Subsidy per gallon of gas saved 0 72.1 cents
C. Effect of fuel-switching under equivalent 72.1-cent gasoline tax
Gas tax — total 80.6 cents 20.8 cents
Tax saving from switch to E85 — 59.8 cents
So what are we complaining about? What are the problems with the ethanol subsidy vis-à-vis a gas tax?
First of all, like all alternative energy subsidies, the ethanol tax credit, by reducing the cost of fuel consumption overall, discourages rather than encourages overall fuel conservation.
An example may be helpful in illustrating this too-often neglected point. Suppose the government wants to reduce consumption of alcohol because of the social costs of drinking. The textbook response is a tax on alcohol — plain and simple.
But instead of directly addressing the problem, politicians opt to subsidize beer in the hope that consumers will switch to beer from hard liquor. (We can construct a beer subsidy "equivalent" to an alcohol tax subsidy — a subsidy with the same benefit as switching from high-tax liquor to low-tax beer.) Yes, the government's new policy encourages switching from hard liquor to beer. But the new policy has not made drinking more expensive (as it would have with an alcohol tax). Instead, it has made it cheaper. By subsidizing low-alcohol drinks, it has put abstinence and less drinking out of the scope of subsidized activity. It is even possible that overall alcohol consumption could increase.
There's an important lesson in this little story: The best way for government to reduce bad things is to tax them. Providing a subsidy for good things in the hope that it will reduce bad things has side effects that diminish or even negate the policy's potential benefit.
The other big problem with ethanol is that it does not allow businesses and consumers to choose the most efficient technologies for achieving the underlying policy goal: less petroleum consumption. It provides no credit for purchasing a more efficient automobile or making your current automobile more efficient through better maintenance or slower driving — even though these actions can contribute to energy security just as well (and perhaps at a lower cost) as switching from gasoline to E85.
What About Global Warming?
Energy policy has two goals: (1) promoting energy security by reducing dependence on foreign petroleum and (2) reducing emissions of greenhouse gases and other pollutants. So far in this article we have focused on energy security. That's because ethanol's role in reducing greenhouse gas emissions is small or even negative. (See, for example, "Study: Ethanol May Add to Global Warming," USA Today, Feb. 2, 2008.)
How should we assess the role of the ethanol credit in cutting greenhouse gas emissions? To answer this question, you need to know two things. First, the amount of greenhouse gas emissions from ethanol. Second, the dollar value of reducing greenhouse gas emissions.
The bad news for us here is that there is a lot of uncertainty about both these numbers. But the good news is that under a wide range of reasonable assumptions, we reach the same conclusion: Reduction in global warming is at best a sideshow in our ethanol policy.
The first two columns of numbers in Table 2 show you why. These figures assume, based on government estimates, that ethanol reduces greenhouse gas emissions by 18 percent. They also assume that the social cost of carbon is $30 a ton. Both estimates are on the high side of those available (see notes for details).
Given those assumptions, the social cost of carbon dioxide per gallon of gasoline is 38.9 cents a gallon, and the social cost of carbon dioxide from E85 is 31.8 cents a gallon per gallon of gas equivalent. The differential between the two taxes is only 7.1 cents. That's equivalent to a 6.1-cent-per-gallon ethanol subsidy. In other words, reduction of greenhouse gas emissions justifies an ethanol subsidy of only about 12 percent of the current-law level of 51 cents.
Table 2. Translating Social Cost of Carbon Dioxide
Into Fuel Taxes and Subsidies
(1) (2) (3)
E85 From E85 From
Gasoline Corn Cellulose
Total fuel (gal.) 1 1.38 1.38
Gasoline (or gas
equiv.) (gal.) 1 1 1
MPG (of gas or
gas equiv.) 21.3 21.3 21.3
Miles 21.3 21.3 21.3
GHG per mile
("well-to-wheel") 552 451 154
tion from gasoline 0% 18.3% 72.1%
(grams) 11,758 9,606 3,280
Total GHG (tons) 0.0130 0.0106 0.0036
(Social cost of CO)2
($/ton) $30 $30 $30
Gas tax (or gas tax
equivalent) 38.9 cents 31.8 cents 10.8 cents
ethanol subsidy 6.1 cents 23.9 cents
And just in case you are wondering whether the case for a larger subsidy for ethanol could be based on its reductions in other air pollutants (sulphur dioxide, carbon monoxide, etc.), you can forget it. A recent study that compared 10 pollutants found that when E85 substituted for gasoline, the levels of 5 pollutants increased, 2 remained essentially the same, and 3 declined. Across the 10 pollutants, the average change through use of E85 was an increase of 40 percent.
So far, all the numbers we've been talking about are for ethanol produced from corn. Currently, corn is the source of 97 percent of U.S. ethanol. But the hope for the future is that cellulose will become the primary feedstock.
Cellulose is the tough stuff in cell walls that gives plants their structure. All plants have cellulose, but the sources most discussed for commercial use are high-fiber plants and plant matter like corn stalks, switch grass, and woodchips. It is more expensive to make ethanol from cellulose than corn because it is more difficult to break it down into the sugars that begin the fermentation process.
Cost is a problem, but cellulosic ethanol has two major advantages over corn ethanol. First, in terms of volume, cellulosic ethanol has more than 10 times the capacity of corn ethanol because of more abundant feedstock. Given the limitations in the capacity of corn production, only the widespread production of cellulosic ethanol will allow ethanol to make major inroads into the market for automobile fuel.
Second, greenhouse gas emissions from cellulosic ethanol are just a fraction of those from corn-based ethanol. As shown in the data in Table 2, corn-based E85 reduces greenhouse gas emissions by 18 percent compared with gasoline. For cellulosic E85, the reduction is 72 percent.
To determine the amount of tax credit that may be justified by a reduction in greenhouse gas emissions, we repeat the calculations we did for corn using the value for cellulose. We find the justifiable subsidy is equivalent to a 28.1-cent-per-gallon cut in the gas tax. In terms of an ethanol subsidy, that's 23.9 cents per gallon — less than half of the 51 cents a gallon available under current law.
On May 22, 2008, Congress overrode the president's veto and passed the long-debated farm bill into law (P.L. 110-234). To help pay for the bushels of goodies included in the legislation, including an increase in the tax credit for cellulosic ethanol to $1.01 per gallon, lawmakers reduced corn ethanol's subsidy to 45 cents a gallon. Both of these provisions take effect on January 1.
So starting in 2009, cellulosic ethanol gets 56 cents a gallon more than corn ethanol. Cellulose may be the future of the ethanol industry, but this differential is not justified by the economics. Corn ethanol and cellulosic ethanol substitute for gasoline equally well, so the case for a bigger subsidy for cellulosic ethanol rests solely on environmental benefits. The calculations in Table 2 show the improved environmental benefits justify a differential of only 15 cents. Or, putting it in other terms, the new 56-cent differential could be justified only if the social cost of carbon dioxide had the improbable value of $110 a ton.
Despite ethanol's tarnished image, there is a good reason to keep subsidizing it: the tremendous economic and security benefits of reducing our dependence on foreign oil. To a lesser extent, the credit might also be justified for ethanol's contribution to reducing global warming. The ethanol credit is scheduled to expire at the end of 2010. Because there is little prospect of our energy security improving anytime soon, Congress should make the ethanol credit permanent.
On the other hand, although an ethanol credit is justifiable intervention, the economic costs of our energy policy and its public benefits could be substantially improved by repealing ethanol credits and replacing them with a new tax on petroleum. To make the new tax politically feasible, all the revenue from it should be targeted to individuals who suffer the most from a high gas tax. To give you a rough idea of what this would be like, imagine a new regime with an extra 50-cent-per-gallon gas tax along with consumers receiving a $100 check each quarter.
In the meantime, as long as we are stuck with the inferior approach of subsidizing gasoline alternatives, we should try to make that approach as rational as possible. Congress moved in the opposite direction in the 2008 farm bill. It gave an overgenerous subsidy boost for cleaner ethanol. At the same time, it reduced the general subsidy for gasoline-saving ethanol at a time when the need to promote energy security is more pressing than ever.
Much of the technical data used in this article are from the study by Norman Brinkman, Michael Wang, Trudy Weber, and Thomas Darlington, Well-to-Wheels Analysis of Advanced Fuel/Vehicle Systems — A North American Study of Energy Use, Greenhouse Gas Emissions, and Criteria Pollutant Emissions, May 2005, available at /taxbasehttp://www.transportation.anl.gov/pdfs/TA/339.pdf (BWWD). The study uses the GREET (greenhouse gases, regulated emissions, and energy use in transportation) model developed by the Argonne National Laboratory with funding from the U.S. Department of Energy.
The amount of E85 equivalent to one gallon of gasoline is derived by computing the ratio of Btus in one gallon of gasoline (115,5000) to the Btus in one gallon of E85 (76,000). The Btu figures are from Table 2-2 of the BWWD study. The amount of gasoline is 0.15 of the amount of E85, and the amount of ethanol is 0.85 of the amount of E85.
The amount of petroleum used in the production, refining, and transportation of gasoline (the "well-to-tank" amount) is computed from information from Table C-1 of the BWWD report.
As in the BWWD report (Table 3-1), we assume a figure for miles per gallon of gas (or gas equivalent) of 21.3. Data on greenhouse gas emissions are from Table D-1 of the BWWD report. The units are grams per mile — 552 for gasoline, 451 for corn E85, and 154 for cellulosic E85. They include emissions from fuel use in vehicles and in the fuel production process (the "well-to-wheel" amount). Carbon dioxide is by far the most important greenhouse gas, but there are others, and those — expressed in terms of CO2 equivalence — are included in these figures. The table also includes data on emissions of pollutants that are not greenhouse gases.
Cellulosic ethanol. More information about the potential of cellulosic ethanol can be found in a 2005 study by the Department of Agriculture and the Department of Energy ("Biomass as Feedstock for a Bioenergy and Bioproducts Industry: The Technical Feasibility of a Billion Ton Annual Supply," 2005, available at /taxbasehttp://www.biomass.govtools.us/pdfs/billion_ton_vision.pdf).
Social cost of carbon. Prof. Richard Tol performed a meta- analysis of 211 estimates of the social cost of carbon. He calculated the average estimate to be about $6 per ton of carbon dioxide and concluded that there was only a 1 percent probability that the true value exceeded $21 per ton of carbon dioxide ("The Social Cost of Carbon: Trends, Outliers and Catastrophes. Economics Discussion Papers," No 2007-44, available at /taxbasehttp://www.economics-ejournal.org/economics/discussionpapers/2007-44). Environmental economist William Nordhaus of Yale puts the cost at about $10 per ton (The Challenge of Global Warming: Economic Models and Environmental Policy, July 24, 2007, available at /taxbasehttp://nordhaus.econ.yale.edu/dice_mss_072407_all.pdf). Richard Newell and William Pizer estimated the value closer to $20 per ton ("Discounting the Benefits of Climate Change Mitigation: How Much Do Uncertain Rates Increase Valuations?" Pew Center on Global Climate Change, Dec. 2001, available at /taxbasehttp://www.pewclimate.org/global-warming-in-depth/all_reports/discounting_the_benefits). All authors emphasize the uncertainty in their estimates. The widely publicized Stern Review put the cost at $85 per ton, but this estimate has encountered much criticism (Stern Review on the Economics of Climate Change, available at /taxbasehttp://www.hm-treasury.gov.uk/independent_reviews/stern_review_economics_climate_change/stern_review_Report.cfm). Our calculations in this article assume the social price of carbon dioxide emissions is $30 per ton. The $30 figure is our guess about what a reasonable figure would be for a fairly aggressive greenhouse gas reduction policy. Those interested could easily redo the calculations presented here for their own preferred value for the social cost of carbon. However, unless that figure is considerably larger than $30 per ton of CO2, most of our qualitative conclusions would remain the same.