I didn't realize that the mess of a financial reform bill just passed still contained the provision that the credit ratings agencies can be held legally liable for their ratings. This is ridiculous, because nobody can predict with 100% accuracy and forcing people to just makes them unwilling to make predictions, and whatever you feel about the quality of ratings predictions, they do provide useful information to back up their ratings (as an analog, I may not agree with a research analyst's buy recommendation for a stock, but I'm usually interested in why THEY think it's a buy).
The other problem surfaced in the journal this morning: http://online.wsj.com/article/SB10001424052748704723604575379650414337676.html
Agencies refusing to rate loans (or, to a lesser extent, refusing to accurately rate a loan by not giving any good ratings for fear of liability) means that it's harder to sell those loans. If the issuer needs to keep them because they can't sell them, that increases risk for the issuer. They need to raise interest rates to compensate for that additional risk. In other words, it raises the cost of capital substantially at a time when we really want to stimulate consumption by lowering interest rates and reducing the cost of capital. I noticed Mankiw picked up on this as well - it's brutally anti-stimulatory at a time when we're desperately looking for cost-effective stimulus.
More and more pieces of the financial reform bill look like they may stall a recovery. They cripple banks' earnings potential for a while, and they will almost certainly result in less lending. They will also almost certainly result in bigger bank fees, which presumably will shrink deposit bases and thus lending as more people either stash cash under the mattress or turn to other methods of saving or investing. Hopefully, de-banking people would result in more spending, but de-banking has the ability to stimulate a lot of illegal activity (theft, as well as the immediate temptations of drugs and prostitution - the third world suffers from this de-banking problem significantly) and result in more disparate and worse socioeconomic outcomes for low-income people.
In short, Congress really does seem unable to think about second-order consequences. Something to ponder for the "more government intervention" crowd.