Clean energy subsidies

There has been some discussion on econblogs, originating, I think, with this Economist blog piece, attacking the idea of carbon offsets and particularly of subsidising clean energy as a way of reducing pollution from traditional energy sources.

The first pro-subsidy argument is that subsidising clean energy will cause clean energy to be substituted for dirty energy, which is good.

The reply, in the Economist piece, and also Arnold Kling's recent TCS article, is that a subsidy to clean energy is a subsidy to energy, and will increase overall consumption of energy, possibly producing more total pollution than would otherwise have been the case. Pigouvian taxes on dirty energy can be justified, but not subsidies.

In a comment on a Kling EconLog entry on the subject, I put forward a second pro-subsidy argument:

The purpose of subsidising clean energy is not simply to displace dirty energy in the immediate sense -- that is undoubtedly done better by taxing the dirty energy.

New technologies require a lot of development investment to bring to market. Once they are being successfully produced and sold in the market, experience and competition often produce rapid improvements, particularly in efficiency. The subsidy to clean energy use should not be seen so much as encouraging a certain quantity of clean energy to be produced, but more as a subsidy to the of clean energy technologies, to bring them more quickly to the stage where they can be rapidly and profitably developed.

The intervention in the market is then not aimed at correcting the direct negative externality of pollution, but at the positive externality of improving technological knowledge.

I believe that this is what the proponents of subsidised clean energy have in mind.

I brought the argument up not because I agree with it, but because I don't like to see important points being missed.

Note that because the point of the subsidy is to encourage industry to invest in the relevant technologies, the effects if it works correctly -- increasing the profits of businesses in the relevant sectors -- would end up looking an awful lot like pork, which is the major criticism. The profits might not be rents, but fair reward for valuable additions to the knowledge of clean energy technology.

There are, nonetheless, a number of problems with the argument.

First, there are other mechanisms which are intended to address this very general market failure (that is, the positive externality of R&D). Most relevant in this case are patents and direct state-funded or state-subsidised research. The existence of these alternatives do not automatically invalidate the use of energy subsidies -- one would have to show that it is a less effective approach. Both government research and patents are far from perfect, and should not be presumed to be the best solutions.

As an example, subsidising the research, rather than the energy, reduces the incentive on the producer to actually improve efficiency. If the subsidy directly targets the energy use, then there is competition between firms producing the subsidised energy, whereas if the subsidy is on a particular research project, it is a pure rent.

More significantly, there is no guarantee that the subsidies will go to the most promising technologies. I don't believe that making ethanol from corn will ever be a useful contribution to the energy industry, no matter how much more efficient the process becomes. It will be carried on only for as long as it is actively subsidised, and is pure deadweight loss. But as long as it is subsidised, it might be profitable for producers and consumers. Solar panels, on the other hand, while not really worthwhile at the moment, might well, if pushed into the market by subsidy, and improved in efficiency once there, develop to the point where they are a major portion of energy production at a genuinely low cost.

Since the policy is currently directed far more to biofuels than to solar energy, I think it is not working.

That's not to say it can't work. There are many other points that can be made pro and con. I am, as always, sceptical that any interventionist policy will be effective in the long run, but the argument has too much merit to be ignored or dismissed out of hand.