Ethanol and Mr Vinod Khosla

Markets Work

Economists have a mantra which says “Markets work” and mumble under their breath the disclaimer “subject to a bunch of conditions, of course.” By “markets work” they mean that when a whole lot of buyers and sellers get together and buy and sell stuff, magic happens through Adam Smith’s invisible hand, and everyone ends up better off than they were before the trades took place. Each market participant has to be concerned with only his objective (maximizing utility in the case of consumers, and maximizing profits in the case of producers) and the maximization of social welfare is assured.

When you go to buy, say, fuel for your home, you check out the alternatives and buy what suits your purpose cheapest. Basically, subject to the thickness of your wallet, you demand a quantity based on the price which you take as a given and which you cannot alter. You really don’t care how the fuel was produced or mined, how it was transported, how it was stored, and a million other things that went it to the process of getting that fuel to the store. All you care about is the price, and rightly so, because the price encapsulates within itself all the information you need to make the decision.

The question then is “how is that price discovered or determined?” If there are a lot of suppliers, then the price is determined through the interaction of the aggregate supply and the aggregate demand: the market-clearing, or equilibrium, price is that which equates the quantities supplied and demanded. If the demand exceeds the supply at a particular price, that is not the equilibrium price. When there are lots of suppliers and demanders, no one of them can determine the price; they are all “price takers.” And the market price is pretty close to the average cost of production. But if there is a very small number of suppliers, each controlling a large part of the supply, then they can set the price. In the extreme case of a monopoly, a market with only one supplier, the price is entirely up to that firm. The price has essentially nothing to do with the costs of production. (This is being typed on a machine running Windows™. See what I mean?)

One basic fact of the universe is that fortunately there are substitutes for pretty much anything we need. Rice and wheat are substitutes; chicken and lamb; so are cars and buses; electricity and gas; capital and labor – the list is endless. Take any category of consumables and you will find a collection of substitutes. Consider the category: liquid fuels. There is gas (also called petrol), diesel, kerosene, vegetable oils, biodiesel, etc. Oh, let’s not forget ethanol produced from biological sources. It is the ethanol issue that sparked the foregoing brief Econ 101 lecture.

Khosla versus Rapier

What happened was my friend Suhit Anantula alerted me to an online debate between Robert Rapier who works for a fossil fuel company and blogs at The Oil Drum, and Vinod Khosla.

Vinod Khosla is an Indian American venture capitalist who is considered one of the most successful and influential personalities in Silicon Valley. He was one of the co-founders of Sun Microsystems and became a general partner of the venture capital firm Kleiner, Perkins, Caufield & Byers in 1986. In 2004 he formed Khosla Ventures.

Vinod was featured on Dateline NBC on Sunday, May 7, 2006. He was discussing the practicality of the use of ethanol as a gasoline substitute. He is known to have invested heavily in ethanol companies, in hopes of widespread adoption. He cites Brazil as an example of a country who has totally ended their dependence on foreign oil.[Source.]

Rapier’s quarrel with Khosla is outlined in his post Khosla Debunked: Ethanol is NOT the Answer. His concern is that Khosla has the power to influence the US energy policy and people may be making a mistake in not scrutinizing the claims that Khosla is making about ethanol. Rapier fears that people—including policy makers—will be lulled into complacency and therefore not pay attention to the looming crisis following peak oil. He believes conservation (among other things) will be neglected.

Rapier’s thrust was to convince Khosla that corn ethanol is a bad thing. Khosla, on the contrary, believes that corn ethanol is a good thing and is an important intermediate step (which will demonstrate the feasibility of ethanol as a substitute for liquid fossil fuels) towards the ultimate goal of cellulosic ethanol which will reduce the US dependence on foreign oil. Khosla is putting some of his considerable wealth where his mouth is.

Khosla agreed to take up the challenges Rapier made in his debunking and debated Rapier on the phone. The details of that conversation are available in A Conversation with Vinod Khosla. Rapier acknowledges that, like himself, Khosla deeply cares about energy independence. They are also in agreement that current energy policy needs a dramatic overhaul. This include (among other things) the imposition of a carbon tax and the elimination of grain ethanol subsidies. They both support funding for research in energy storage devices.

Each of Rapier’s two posts generated around 300 comments. I was somewhat surprised that none of the comments addressed the issue from an economic viewpoint. The debate of course is between an oil man and a very successful venture capitalist, neither of whom is an economist. My aim here is to inject some basic economic reasoning to the issues they discussed.

When Markets Fail

The first question is: Is ethanol, whether grain or cellulosic, the answer to the matter of energy independence? The answer from an economist’s point of view is simple: The market knows. Essentially, consumers will vote with their dollars. If at the pump, they find ethanol to be a cheaper substitute, they will buy it. It all depends on the price. And if the price is “right,” ethanol will win and consumer behavior will assure a socially optimal outcome. But—and here is the clincher—will a socially optimal be assured if the price is tampered with?

One can tamper with the price in a number of ways. For instance, you could impose a tax (or a subsidy) on the thing itself or on its substitutes. Or you could tax (or subsidize) some of the inputs that are required along the long production process of the thing or its substitutes. If you do so, then the price does not reflect the full cost of production: it can either be higher (in case of taxes) or lower (if subsidized). Then the market outcome may not be socially optimal. In effect, you have not met one of the conditions that are necessary for markets to work. (It is the “no externalities” condition.)

Let’s confine ourselves to ethanol and gasoline (the stuff that gushes out of oil wells.) The price you pay at the pump for gas has, you would suspect, some relationship to the cost of discovering crude, mining, transporting, refining, storage, and distribution. It is not entirely clear though whether the price you pay is equal, higher, or lower than the total cost. What if in the supply chain, some inputs are not included in computing the costs? For instance, there is a cost associated with assuring supply of Middle East oil and the protection of the oil interests of the US. Wars have to be fought and blood split. The wars can be considered an “externality” associated with the supply of Middle East oil. Is a certain amount (tax) added to internalize the “defense expenditure” to the cost of the gas you pump at the station? If not, then it could then well be that the price you pay is lower than the cost and in effect, the more gas you use, the more subsidy you enjoy.

Then there is the externality associated with the consumption of the gas. How much is the cost of the greenhouse gas emissions associated with the burning of a gallon of gas? It is not at all a trivial exercise to arrive at that number. Yet, strictly, that cost has to be added if we are to claim that we are pricing gas at full cost. One instrument for internalizing the cost of that consumption externality is the carbon tax.

Finally, and not the least, is the question of how much is the cost of mining the oil, which is but the first step in the production process of gasoline. Note that crude oil is “mined” and not “produced.” There is a finite (perhaps unknown) amount of crude oil in the ground. The net present value of that oil stock depends on the present (and future) availability of substitutes. Imagine if we were to find a source of very cheap energy—cold fusion, for instance. Then the value of the stock of crude will go down substantially. Contrariwise, if no substitutes are found, then the price of fossil fuels will sky-rocket.

Similar concerns are associated with ethanol. Is there a subsidy (or tax) associated with it along its supply chain? Are the farmers getting a subsidy for growing corn? That’s a direct subsidy. Is the fossil fuel which goes into the production of the corn itself somewhat subsidized? That’s an indirect subsidy. What about the “blender’s subsidy”? That drives a wedge between the cost of production and the price.

It appears that it is a very, very complicated matter to figure out the true full cost of either of the fuels. So the market may fail to grind out the socially optimal result. It is of course conceivable that all those numerous taxes and subsidies and un-internalized externalities will somehow cancel each other out and the prices will reflect the true costs of production. In which case, the market will give us a winner. But I somehow doubt it.

Having recognized that the price may be “inaccurate” given the convoluted taxes, subsidies, and externalities, it is still true that at the pump the consumers will simply compare the prices of ethanol and gasoline and go for whichever gives them more bang for the buck. As it happens, the historical rack price of ethanol has been consistently higher than that of gasoline for the period 1982-2004 (see Rapier). For the record it must be mentioned that there is a $0.60 per gallon “blender’s subsidy” associated with ethanol. Does this subsidy go along some way to rectify the mis-pricing of gas? Would it be better to remove that subsidy and instead add on an additional tax on gas? This will have distributional consequences. Fuel companies, both gas as well as its substitutes, will see their profits decline, the consumer will pay more, and the public treasuries will gain. These are policy issues which are generally resolved in favor of whichever lobby has the most political power.

Fixed Costs

One of the guiding principles that Khosla follows in his recommendation is that the solution to energy independence has to be practical and therefore must suit the current fleet of vehicles. He believes that the auto companies will oppose any solution which requires re-tooling of their existing manufacturing facilities. In other words, there are higher fixed costs associated with some substitutes of gasoline, and Khosla’s position is that these are lower in the case of ethanol. Aside from the fixed cost of using ethanol, there is also the fixed cost of distribution of ethanol. The infrastructure of pipes which delivers gasoline cannot deliver ethanol for technical reasons. Setting up a parallel delivery system is a major fixed cost and will be a barrier to the adoption of ethanol.

The existence of fixed costs negates one of the conditions necessary for the market to grind out the social optimal. That is, if firms have to recover their fixed costs as part of the price of ethanol, then they will be at a disadvantage relative to firms which supply gasoline and whose investment in the distribution system for gasoline is a sunk cost and thus do not affect their bottom line. One way to level to playing field is to pay for the ethanol distribution system through public funds. However, the use of public funds have opportunity costs which need to be evaluated. More about this later.

There is another very important cost associated with petroleum substitutes: discovering them. In the case of ethanol, research has to be funded and processes discovered for continually improving the efficiency of transforming grain or cellulose into ethanol. Again these are fixed costs. For market efficiency, the fixed costs have to be paid for from a source other than at the pump. And here is where choices have to be made at a policy level. Where should government and society place its bets and how much? Should it be ethanol? Or should it be solar? How about butanol? Should government fund a wide range of potential substitutes and let the chips fall as they may? But then, public funds are not unlimited. Research funding could be too thinly spread out for any of them to be successful. Choices have to be made. Granted these choices are not made in a vacuum of ignorance; many terribly knowledgeable brains can help out with assigning probabilities to the bets. And it is venture capital geniuses such as Khosla who influence to no small extent which way the government bets will be made, and eventually make a pile when the chips fall.

Scale Economies

Coming back to the matter of when market competition leads to socially optimal outcomes, there is another condition which must be satisfied, and that is, there must not be “scale economies.” More specifically, if there are increasing returns to scale, then markets may not work as advertised. Here is a hypothetical scenario. Imagine that the cost of production of gasoline is $1.50 a gallon when one billion gallons of gasoline is produced. Assume that the final price is $1.50, equal to the cost. Now assume that only one million gallons of ethanol is produced. Let’s say it costs $2 a gallon, and the final price is $2.00 a gallon. Clearly, at these levels of production, ethanol would lose out to gasoline. Ethanol will not substantially replace the use of gasoline.

Now assume that someone spends $100 million on research and develops a very efficient process for converting cellulose to ethanol which brings down the marginal cost of production of ethanol to $1 a gallon. Then if we produce two billion gallons of ethanol using the new process, then the average cost of a gallon of ethanol is $1.05. In other words, we have constructed a scenario where there are scale economies in the production of cellulosic ethanol, and these scale economies arise because of the $100 million fixed cost of research and development. Given this scenario, the availability of 2 billion gallons of ethanol priced at $1.05 a gallon competes very well with the one billion gallons of gas priced at $1.50 a gallon. In fact we will have up to $0.45 per gallon of ethanol to play around with to pay for the any additional cost of distribution and use. Consumers will substitute their fuel demand from gas to ethanol and thus reduce the dependence on gas.

Conclusion

Here is a recap of the story so far. Ordinarily, the market can be relied upon to decide whether ethanol will indeed reduce our dependence on fossil fuels. However, the required conditions for markets to work are not met in this case. They are so because first, the prices are distorted due to externalities, taxes, and subsidies. Second, there are scale economies and fixed costs. Therefore all the usual talk about the market grinding out the optimal do not hold. This calls for intervention in the market, though I should hasten to add that the intervention does not have to be governmental. The main intervention has to be in the funding of the fixed costs for the development of ethanol production processes and the removal of factors that distort the prices. The latter involves rational taxation policy.

Khosla’s position, in my considered opinion, can be justified on economic grounds: he is intervening in a market which has market imperfections. As a venture capitalist, he is positioning himself to gain from the possible success of ethanol as a potential substitute for fossil fuels by investing in ethanol production. He is, in effect, paying for some of the fixed costs associated with ethanol. His lobbying efforts also go towards getting the government to fund research and development. Furthermore, his promotion of ethanol will likely catalyze some research investment in ethanol production processes. He cannot be faulted for doing what a venture capitalist does.

It does not matter whether Khosla’s motives are purely altruistic or not. As far as I am concerned, self interested behavior and social welfare are not necessarily mutually antagonistic. His case for ethanol also does not depend on the extent to which Brazil has succeeded in the use of ethanol. His position is also immune to any EROI (energy return on investment) arguments. This is so because the quality of energy input differs from the quality of energy returned. Simply stated, if one is spending $10 to convert 10 units of energy to produce 7 units of energy which is valued at $2 a unit (total value $14), then even though the EROI is only 0.7 (less than unity), the economic gain is $4.

So is ethanol the answer? We don’t know for sure. The great challenge is to get the price right. And for that, we have to determine the full cost of the alternatives. Having gotten the price right, we can depend on the market to determine the optimal solution from various alternatives. The alternatives are also discovered though a process which depends on the market. By that I mean not just the market for ethanol and its substitutes, but also the market for ideas which is the more important market. Through informed debate and dialogue, based on solid research, it is possible that we will converge to a solution – if not the perfect solution, to a close approximation of it. The best thing about a liberal democratic market-based economy such as the US is that the process of making public policy is itself market based. You have to defend your ideas against competing ideas and the ones which can withstand scrutiny survive.

Rapier’s challenge to Khosla and Khosla’s response is an excellent illustration of the basic process which underlies the enormous success of the United States: debate and inquiry. It has been a learning experience for me.

Post script: I am postponing the “food versus fuel” issue for later. Basically it says that if resources are used in growing biomass for fuel, food production will suffer and adversely affect those who are already hungry and poor. Khosla rejects that trade-off and maintains that there is sufficient food but the ability to pay for it is lacking. In a future piece, I will explore why I agree with Khosla.