About a year ago, Paul Romer pointed out that

In the next century we’re going to be moving back and forth, experimenting with where to draw the line between institutions of science and institutions of the market.

In this context, there has been considerable criticism of President Bush’s proposal to spend $1.2 billion on hydrogen fuel cells.
Paul Georgia says that the technology is over-hyped.

Since hydrogen does not exist in geological reservoirs it must be extracted from fossil-fuel feedstocks or water. The process of extracting hydrogen uses energy, which means that using hydrogen is less efficient that burning fossil fuels…

The other option is to extract hydrogen from water using renewable-energy sources…by the time you use the energy to extract hydrogen from water, transport that hydrogen to where car owners can get to it and then recombine it with oxygen to re-extract the energy the cost becomes astronomical.

If hydrogen power is economical, then theWall Street Journal editorializes (subscription required) that

[this is] all the more reason taxpayers shouldn’t be forking more over to the effort. Granted, there are areas where government research funding serves a purpose — particularly in fields where there is little or no commercial market to drive development. But private companies clearly think they stand a chance to make money from fuel cells; if a market exists, they’ll find it, without Washington throwing money at them.

Lynne Kiesling weighs in

Such technology subsidies, particularly to get renewable energy technologies to the point where they are commercially viable, overlook an important point: the fact that these technologies are not commercially viable may mean that they are not economically efficient. If entrepreneurs and investors do not think that these forward-looking investments make economic sense, then for the government to override this decision with subsidies is almost certain to be wasteful and more costly than need be.