In June Art Carden and I argued that subsidies for nuclear power were a bad idea, not because there's anything wrong with nuclear power, but because we cannot know the best way to generate cleaner energy--it takes a market with prices to figure that out.
Daniel Gross argues in this piece in Slate that people who oppose subsidies for electric cars misunderstand "the process of innovation, economic history, and the current macroeconomic situation". For someone who is writing with such confidence, I would say Gross is the one displaying staggering misunderstanding. Yes, costs of building electric cars will probably come down with economies of scale, and yes, that process has repeated many times through history. None of that suggests that subsidies are a good idea, and Gross doesn't even begin to justify subsidies. How can the government know that electric cars are the best solution? What about small turbocharged clean diesels, or hydrogen, or some other technology? How can the government know which is the best one into which to sink billions of dollars? It can't. With a subsidy, however, it is making that very expensive bet.
If we want cleaner cars, a gasoline tax is the way to go. Car producers will experiment with a variety of techniques in an attempt to find the cheapest way to reduce gasoline usage. We'll also get responses from consumers, who move closer to work, carpool, and find other ways to reduce fuel usage (thereby reducing pollution). Government subsidies for new car technologies make as much sense as subsidies for, say, new MP3 players or new vacuum cleaners--that is to say, they make no sense at all. Markets are great at technological innovation; rather than ramming one form of innovation down our throats, why not give everyone the incentive to innovate in a variety of ways until we find the best solutions?