People have enjoyed ethyl alcohol for 10,000 years, but today it holds a new fascination. It carries energy, and it works like this. First, capture solar energy and CO₂ in an ear of corn. Then, ferment the corn. Put it in your gas tank and drive around thumbing your nose at the rich oil sheiks and terrorists. Now that's an energy policy.

In 2006, the US produced 4.9 billion gallons of ethanol, and subsidized every one with a federal 51¢ blender's credit. On top of that $2.5 billion, the USDA paid $0.9 billion in subsidies for corn used to make ethanol. But drivers outdid the federal government by paying $3.9 billion extra at the gas pump, on an energy equivalent basis, to let their cars sip corn whisky. In return, we are 1.1% more energy independent, and GHGs were reduced 1/19 of 1%.

Is it worth it? Surprisingly, this question has been preempted by a controversy over whether ethanol has a "negative net energy balance." The controversy stems from the fact that it takes fossil energy, as well as corn, to make ethanol. So replacing a gallon of gasoline with ethanol both saves and consumes fossil fuel. If the production process uses more energy than the ethanol saves, the net energy balance of ethanol is negative.

While net energy is an interesting question, it's not the question that needs answering. Even if it were negative, ethanol could be produced with American coal and would replace foreign oil. This would still help make the US energy independent. And if the energy balance is positive, ethanol might still be a bad idea. If it saves too little and costs too much, it's just not worth it. There are many better possibilities.

Just for the record, an article in the July 2006 issue of the Proceedings of the National Academy of Science (Hill, 2006) puts ethanol's net energy balance at +25%. It takes 1 GGE of non-corn energy to make 1.25 GGE of ethanol. This clears the way for the two basic energy-policy questions: (1) how much does it cost to reduce GHG emissions with ethanol? (2) How much does it cost to reduce energy imports with ethanol?

Reducing Energy Imports

Ethanol faces two challenges as it replaces fossil imports: First, ethanol contains less than 2/3 the energy of gasoline on a gallon-for-gallon basis. Second, producing ethanol uses a significant amount of natural gas, which increase fossil imports. This is a problem because imports now account for 30% of total US energy use.

In January 2007, testifying before Congress, Keith Collins, the chief economist of the US Department of Agriculture testified that "In 2006, ethanol production, on an energy content basis, was equivalent to only 1.5 percent of U.S. crude oil imports." (USDA, 2007) In March he presented a graph predicting ethanol output would increase rapidly through 2010 and then gradually increase to 12 billion gallons in 2017 (USDA, 2007b), which would save 3.7% of oil imports if those stayed constant over the next ten years.

But his figure is for oil independence, not energy Independence. The US is also dependent on natural gas imports. "Imports of natural gas are projected to rise to meet an increasing share of domestic consumption," according to DOE (2007b). That increase is already happening and it's LNG, liquefied natural gas, from overseas, not Canadian gas, which is declining. LNG imports will have all the same problems as oil imports. Because making nitrogen fertilizer for corn uses a lot of natural gas, corn ethanol scores worse on energy independence than normally assumed. Correcting for natural gas imports reduces the USDA's 1.5% to 1.1% for 2006, and reduces the impact from 3.7% in 2017 to only 2.8%.

USDA projections indicate corn ethanol will contribute only 2.8% to energy independence by 2017, with little increase thereafter.

Although higher estimates for ethanol's contribution can be found, they are based on one of two errors. Some count only imports used to make gasoline, and some ignore the fact that a gallon of ethanol contains less energy than a gallon of gasoline.

Reducing GHG Emissions

In theory, use of ethanol releases no net CO₂, because all of the carbon in corn ethanol was taken out of the atmosphere by the growing corn. The corn plant sequesters even more CO₂ in its stalk, leaves and roots, but this has no net effect because these plants are not stored, they are ploughed under or fed to animals. Within a year or two they have decomposed and the carbon has returned to the atmosphere. The overall zero net effect is the meaning of "carbon neutral." In practice, however, several steps in the production of ethanol are energy intensive, and thus result in the net production of GHG emissions, even when compared to the production of other fossil fuels. The energy-intensive steps in ethanol production include processes such as making nitrogen fertilizer and burning coal for heat in ethanol plants.

The National Academy of Sciences' paper cited above includes an estimate of net impact on GHG emissions from growing corn and producing ethanol.

Relative to the fossil fuels they displace, greenhouse gas emissions are reduced only 12% by the production and combustion of ethanol. (Hill, 2006)

One unexpected reason for this meager result is that nitrogen fertilizer and soil microbes can work together to release N₂0, a potent GHG. UC Berkeley's Renewable and Appropriate Energy Laboratory published a slightly higher GHG reduction rate, but soon corrected it down to 8% (Farrell, 2006). Using Hill's 12% values, the impact of ethanol on GHGs in 2006 was less than 0.054% (about one-twentieth of 1%). By 2017, this will grow to just over one-tenth of 1%.

Estimates of corn ethanol's contribution to GHG abatement, published by the National Academy of Sciences, place its total contribution in 2017 at just over 1/10^th of 1%, with little increase thereafter.

Who Gets the Subsidies?

Direct Subsidies

Direct subsidies are paid out of taxes. As mentioned earlier, in 2006 ethanol received $2.5 billion in direct subsidies paid to blenders. The 51¢ per gallon "blender's credit" is paid to those who blend ethanol with gasoline before selling it to gas stations. Although it's paid to blenders, this is a competitive business and the market prevents them from profiting from it. Instead, the credit is passed on to producers or consumers depending on the relative price of gasoline and ethanol. Since markets often play tricks on regulators, this is worth a closer look.

Suppose oil prices are high enough that ethanol and gasoline cost the same to produce; say that's $2.00 per gallon. Ethanol producers will take advantage of the fact that blenders want the 51¢ credit, so they will push the wholesale price of ethanol all the way to $2.51, at which point blenders will still see it as just as costly as gasoline in spite of the subsidy. The USDA notes that over the past 25 years, the ethanol producers did in fact push the price this high, in fact, they pushed it about 3¢ higher. The result is that producers have captured all of the subsidy paid to blenders. This was probably the intention, but notice that with gasoline costing just as much as ethanol, the producers don't need any subsidy—the subsidy simply provides them with windfall profits.

Now suppose the producers did need the full subsidy because ethanol actually cost $2.51 to produce. They would still charge $2.51 and the blenders would still get the 51¢ credit so the ethanol would have a net cost to blenders of $2.00 ($2.51 less the $0.51 credit) just as before. But this time the benefit goes to the consumer. Without the subsidy, the consumer would have to pay $2.51 (plus a markup) for the ethanol, the full wholesale price. With the subsidy, the consumer only pays $2.00 (plus a markup), so the consumer gets the full benefit of the subsidy.

• If ethanol costs the same as gasoline, the subsidy actually is not needed; it goes to producers as windfall profits.
  
• If ethanol costs more to produce than gasoline, and the full subsidy is needed, the ethanol subsidy goes to the consumer.

When ethanol costs more to produce than gasoline, but the consumer captures the blender's credit, the producers still do just fine. The demand for ethanol, partially the result of state requirements, will assure that the price ethanol producers are paid will be driven up to a level that covers their costs.

Corn subsidies, like other agricultural subsidies, are enormously complicated. Collins (USDA, 2007) values them at about $8 billion in 2006 and $4.5 billion in 2005. But only 20% of the 2006 corn harvest was used for ethanol (USDA, 2007), and less was used in 2005. So the subsidy for ethanol corn fell from $1.3 billion in 2005 to under $0.9 billion in 2006. Part of the corn subsidy goes away when corn prices are high, as they were in 2006.

Indirect Subsidies

Indirect subsidies are paid by consumers through high prices at the pump. The USDA found that between 1982 and 2006 the wholesale price of ethanol averaged 57¢ more than the wholesale price of gasoline (USDA 2007c). The average subsidy during that period was 54¢. From the blender's perspective, after the subsidy, the price of ethanol was just 3¢ more than the cost of gasoline. Apparently there is a strong tendency for the price of ethanol to equal the price of gasoline, but something is giving the ethanol price a little boost.

Although it may seem obvious that ethanol and gasoline would sell for the same price per gallon, the USDA's chief economist points out that with more supply, the premium could "decline toward ethanol’s energy equivalent with gasoline." (USDA, 2007) In other words, Collins is saying that in a balanced market, ethanol and gasoline will tend to sell for the same price per energy, not the same price per gallon. That seems reasonable since it's energy that powers a car, and not just gallons. There is already some downward price pressure on the E85 market where people are noticing that they get fewer miles per gallon, because with 85% ethanol, its a big effect.

But the USDA found that for the last 25 years there was no tendency towards this balance. Ethanol and gasoline have sold at the same price per gallon (after subtracting the subsidy) and not at the same price per energy. Ethanol selling at the same price per energy as $2.00 gasoline would not sell at $2.00 per gallon but at $1.33 per gallon. At $2.00 per gallon, ethanol is getting a 67¢ price subsidy simply because its lack of energy is ignored by the market.

Why does the market make this mistake? First, consumers often don't know they are buying the ethanol--it's just blended in for smog control. Second, they have no choice because all gasoline has it included. Third, very few people know it has less energy. These market anomalies are the source of the mis-pricing and the primary source of ethanol's indirect subsidy. In 2006, because of the MTBE phaseout and the switch to ethanol as an additive, wholesale ethanol sold (after subtracting the subsidy) for 13¢ more per gallon than gasoline. This added to the subsidy from mis-pricing.

In 2006, the effective wholesale price of ethanol, after subtracting the blender's subsidy, was 75¢ higher than the price of gasoline with the same energy content, hence the indirect subsidy to ethanol, paid at the pump, was 75¢ per gallon of ethanol. (Note: the 75¢ includes a 5¢ adjustment for possilbe octane value.)

Total Direct and Indirect Subsidies

The indirect subsidy to ethanol on the 4.9 billion gallons produced in 2006 comes to $3.6 billion. Together with the direct subsidies of $0.9 billion for corn and $2.5 billion for ethanol the grand total is $7.0 billion. That's $1.45 per gallon of ethanol, or $2.21 per gallon of gasoline replaced.

In 2006, the total subsidies for ethanol came to roughly $7 billion, which is $1.45 per gallon of ethanol produced or $2.21 per gallon of gasoline replaced.

The Cost of Subsidies

Subsidies are not the same as social cost. One man's tax payment is another man's profit. Social cost measures the use of the nation's resources, such as labor, capital, and energy. Collecting income tax and putting the money into the pockets of ethanol producers does not use up any resources, and is not a net cost to society, but it is a cost to taxpayers.

Unfortunately there are no accurate public data on the excess profits of ethanol producers. For the sake of discussion, suppose the chief economist of the USDA is right that ethanol costs $1.65 per gallon to produce (USDA, 2007b). Gasoline with the same energy cost 38¢ less in 2006, so the social cost of using ethanol instead of gasoline is 38¢ per gallon of ethanol. This does not include the costs and benefits of externalities; those must be compared against this market-based social cost. The rest of the $1.50 per gallon subsidy, or $5.4 billion in total, lines the pockets of real farmers, corporate farmers, and ethanol distillers. In the following table, the "Apparent Cost" refers to the total cost of subsidies and the "Real Social Cost" refers to the extra, market-determined cost of using ethanol in place of gasoline. The costs shown are based on the calculations of GHG and energy import reductions discussed above.

The reader may believe the USDA has misestimated the cost of ethanol production. In this case, the choice is between believing more than $5.4 billion went to windfall profits, or believing real social costs are higher than indicated in the table above. They cannot both be lower; the $7 billion in subsidies went somewhere.

The corn-ethanol subsidies create a net flow of money from the rest of the country to the corn states. Just as advertised, this creates wealth and jobs in these states. But spending more in corn states means spending less in other states, which reduces employment in the states that pay for the corn-ethanol market euphoria.

Policy Implications

The first rule of subsidies is to pay for performance. That means: pay for outcome not for input. The desired outcome is not ethanol. Ethanol is the input, the means of achieving the outcome. The principal outcomes we are interested in are reductions in GHG emissions and reduction in energy imports. If we pay for ethanol we will get ethanol, but it may be made with coal and actually increase the GHG problem. To get a reduction in GHGs, pay only for reductions. In fact there are many ways to grow corn and produce ethanol, some much better than others. Because the subsidies are not tied to GHG performance, billions are given away without providing the least incentive for better performance.

As an energy policy, ethanol subsidies are way off track and deeply entrenched, but there is one simple fix that would help a great deal—just stop the 51¢ blenders credit. Between the high price of gas and the state ethanol requirements, there will still be more ethanol produced than makes sense. Yes ethanol is a help, but it provides way too little help for too much cost. A little less ethanol will be produced—which would be a good thing for the environment and our pocketbooks—but mainly we would fork over several bilion dollars per year less to pay for windfall profits.

Ethanol from Cellulose

To end on an upbeat note, consider cellulosic ethanol, which is just ethanol made from cellulose, the part of plants we cannot digest because we are three stomachs short of the standard set by cows and because we are not in the same league with termites. Currently the best thing about cellulosic ethanol is that it is not made from corn, which means it does not require the most nitrogen fertilizer of any crop or heavy doses of insecticides. In fact, it requires no nitrogen fertilizer and no pesticides because it is made from the parts of crops, and non-crops that are not used.

For example the new plant being built by Celunol, in Jennings, Louisiana, will make ethanol from bagasse, which is what is left over when sugar cane is processed. Corn stems and leaves, switch grass, and wood chips, are just a few of the other good candidates. Although the Jennings plant is sure to work, just as a much smaller plant in Canada has been working for a few years, it is unlikely to be competitive with corn ethanol. The problem is that currently-available technologies for converting cellulose to ethanol are themselves not in the same league with termites, which know how to use a plethora of specialized gut bacteria to do an efficient job of breaking down cellulose in a way that our guts -- and our factories, so far -- cannot manage. But there is much current research aimed at replicating termites' trick; thus there is a good chance of getting cost effective GHG reductions from cellulosic ethanol in ten years.