Coordination Failure and Energy Policy

Should the government have an energy policy? Shouldn’t the free market decide these things? Well there might be a reasonable economic argument for having some government leadership, but it might not be completely convincing. For those who believe that all that the government touches turns to gold, they’ll be easily sold. For those who believe that all that the government touches turns to mud, they’ll be very skeptical. But before I give the argument, let me give some examples of coordination.

A few weeks ago, I noted in this post that workers in India didn’t use shovels because they lacked boots. And no one bought boots because there was no market for shovels. Selling shovels in India was a tough task because you needed to coordinate the market for shovels with the market for boots. I would not say that it is impossible to sell both; it is just more difficult.

Another example of coordination failure comes from the computer industry. People who are familiar with software are probably aware that the Fortran programming language lasted about 20 or 30 years longer than it should have. It wasn’t that there were no alternatives, there were too many. Fortran had an established position. Many people knew how to use it. All computers could run it. There was a big body of canned software that could be inserted into your program. These three factors made Fortran hard to beat. Other languages could have been better if more people knew them and used them but not many knew them and used them.

I remember that Pascal was supposed to be the new language. People were split into various camps. Some believed that Pascal was the next new thing and learned it. Others believed one of the various competitors would be the next new thing. Many people felt the next new thing would be coming in a few years, (which reminds me of a cartoon that went, “I’ll buy a new computer when they stop changing them.”) And some people insisted that some version of Fortran was always going to be the standard. Without coordination, Pascal didn’t really stand a chance and the people who invested in learning it got burned. They produced code that was essentially hieroglyphs later.

Of course, the market eventually did shift. Now, I believe (I am not an expert on software) that Fortran has largely been supplanted by C++. So the market worked in the long run. But like Keynes said, “In the long run, we’re all dead.” The transition was slow and it might have been that the transition would have been better if everyone knew when and where the transition would occur. As it was, a lot of people invested in Fortran because they just didn’t know that it would become obsolete.

Suppose for the moment that there were no laws on which side of the road you must drive and people were free to drive on either side. This is kind of what happens in India (at least it seemed like that with the drivers we had). People will spontaneously figure out that they better stick to the most popular side of the road. But then it helps to have right hand drive if you are driving on the left side. And if everyone else in the world drives left hand drive, it might be unfortunate for your auto industry if you drive on the left side.

Well Sweden faced exactly this situation in the 1960’s. The government picked a day (February 3, 1963) and said from that day forward, “we’re driving on the right side.” And that is what happened. It probably was really good for their automobile manufacturer (Volvo). But is this heavy-handed government intervention good public policy? The wiki article explains that one motivation was to reduce the accident rate (Sweden drove left-hand-drive vehicles on the left hand side of the road). This produced a modest short-term benefit and no long term benefit.

We can model the coordination problem as n-player game where n is large. Suppose each play can choose either technology O (original) or technologies A, B, C, D, E, or F. O is old and gives a payoff of 5 units. The other technologies all seem equally promising (but each person has a personal favorite) but you really don’t know what you will get if you play it. If at least 50% play A, the payoff will be between 5 and 15. But the same is true for B, C, D, E, and F; and if you don’t get the 50% threshold, your investment is wasted and you get nothing. You can see that people will stick with the “devil they know” rather than risk getting burned. They all might be much better off if everyone knew to pick any one of these alternatives – but which one? If the government chooses then the risk is that not only will they back the wrong horse, the government will kill any chance to back the right one in later years.

Brazil decided to do something interesting in the 1970’s. Like other countries, they were hit hard by the oil shocks. But they didn’t just sit back. They invested heavily in the production of ethanol. Now it is easy to see the coordination problem associate with ethanol: they needed production plants, they need specialized distribution sites (ordinary gasoline stations had to be modified to prevent water absorption), and they needed to modify the automobiles as well.

In the 1980’s and 1990’s, Brazil’s ethanol policy was a big government boondoggle. Oil prices were well below the level needed to make it pay (roughly $50 to $60 a barrel). But now that oil is above $60 a barrel (I read it was up to $66 today) it looks like Brazil lucked out. Or maybe Brazil was smart all along. Or maybe Brazil is still stupid and will find this out in 20 years when they discover that they backed the wrong horse.

On the other hand, we see that Japanese auto makers have developed hybrid technology with their own investment money. This is exciting technology that might evolve into electric cars in ten years. This might be much better than using ethanol. Or maybe hybrids will go along with 8-track tape into the museum of “almost technologies”.

Question 1, “Is it better to say ‘hands off’ and let the market resolve any coordination problem or is it better for the government to show leadership, at least once in a while?”

Question 2, “Should India be trying to copy Brazil’s ethanol production system.” India has the climate for sugar cane – and it is much easier to adopt a proven technology.

My comment is that there is a real difference between ideal government policy and real government policy. In the real world, the government prevents the importation of Brazilian ethanol to protect its investment in corn-based ethanol. Basically, it shows that governments, once shown to have backed the wrong horse, will knee-cap the competition to guarantee that their horse wins anyway.

Other interesting articles: Marginal Revolution, Knowledge Problem, Coyote Blog, Wikipedia on ethanol.