When an analyst is "shocked", is that usually a good sign?

[TJ in Oregon]

Social science field models are nearly always awful at predicting future events.

[TJ in Oregon]

Government faces the same risks and rewards as anybody else. The fact that government suffers losses (Solyndra) doesn't make it any worse than any private sector actor that occasionally suffers losses. The main difference is that economics does not allow for consideration of distributional questions, while such questions are at the center of governmental behavior.

Yes, but anyone underwriting a loan to Solyndra would have realized they would not get their money back. It was that bad. In another instance, the feds lent money to an outfit that had never sold a product (and no prospect that they ever will), then they sold some product to their Canadian subsidiary, and got a subsidy for doing that.  It seems that their main "product" is their political connections to the government.

There is an underlying problem somewhat specific to the solar energy field. Solar technology is very much still in the experimental stages and one its biggest problems is durability, which is very difficult to test in a lab. If you want to see if a solar panel will last 1, 5, 10, or 20 years before the efficiency starts to fall apart, the best way to test it would be to wait 1, 5, 10, or 20 years, but that's not practical because people want to buy stuff now. So what ends up happening is the company (in this case Solyndra) slaps a 15 year warranty on the thing, sells it anyway and just accepts that if the panels don't last 15 years they'll be out of business anyway and it won't matter. Then the company declares bankruptcy, folds, and the people who pushed it leave for another company.

[TJ in Oregon]

There is an underlying problem somewhat specific to the solar energy field. Solar technology is very much still in the experimental stages.

This is not true at all. President Carter had solar panels installed on the White House more than 30 years ago. The technology is advancing. But this is true of all energy forms. Is natural gas still experimental because we've only been fracking for a few years? And contrary to your claims, Solyndra didn't fail because too many people wanted a warranty. It failed because, like with so many other businesses, they were undercut by the Chinese.

Natural gas is not experimental because while fracking is, we have been using natural gas as a byproduct from oil drilling for decades and using natural gas as a power source is well understood. Fracking itself is still in the infant stages in some ways, but it appears to be working very well as we have more natural gas on hand today than we have ever had..

Solyndra would have inevitably failed because they produced a type of thin film solar cell that is not economically profitable compared to conventional crystalline silicon cells. The reason why you would want to use thin film cells is that crystalline silicon is really expensive, so much so that conventional solar cells are unable to compete with fossil fuels. Thin film cells are being developed in an effort to reduce the huge capital costs and truly make the solar industry viable instead of just a subsidy.

But thin film cells degrade--badly, often as much as a 5% drop in efficiency within the first few months. Thin film cells in particular are experimental technologies beyond the degree that crystalline silicon cells are.