A compelling argument by Megan McArdle.
The whole post is worth reading, but here's a sample:
A number of people have claimed that the government had to make these loans because they're "too risky" or "too big" for the private sector, forcing the government to act as a VC firm. That riskiness means that yes, a non-insubstantial number of the loans will fail.
But this doesn't really make any sense. The private sector doesn't have any trouble dealing with risky ventures; it simply prices the capital accordingly, demanding high interest rates, or a larger equity chunk, in exchange for money. . . .
[A]t the company level, there's no difference between an optimal market outcome, and an optimal social outcome (from the DOE's point of view); both investors and society benefit if more solar cells are sold. If the solar cells are unlikely to be sold to many people, than the loan guarantee is not a good idea — it will not foster much environmental benefit. If the solar cells are likely to be sold to many people, than the loan guarantee should not be needed; private investors should be easily found to back the manufacturing.