Over at Credit Writedowns, Edward Harrison discusses Richard Koo's work on balance sheet recessions. Koo believes that monetary policy is ineffective in such settings. I disagree and have made the case before against Koo's views on balance sheet recession. Here is the comment (with some slight edits) I left for Harrison:
U.S. households during the 1920s acquired a vast amount of debt and began a deleveraging process during the Great Depression. Consequently, there was a "balance sheet" recession in the 1930s too. Monetary policy, however, was not impotent during this time. At least when it was done the right way. FDR's price level targeting from 1933-1936 sparked a robust recovery. In my view, this experience provides a great example of why Richard Koo's balance sheet recession views are wrong. Yes, deleveraging is a drag on the economy, but for every debtor deleveraging there is a creditor getting more payments. (And if the debtor is not making payments and defaulting then the debtor still has funds to spend.) In principle the creditor should increase their spending to offset the debtor's drop in spending. The reason they don't--creditors sit on their newly acquired funds from the debtor instead of spending them--is because they too are uncertain about the economy. There is a massive coordination failure, all the creditors are sitting on the sideline not wanting to be the first one to put money back to use. If something could simultaneously change the outlook of the creditors and get to them to all start using their money at the same time then a recovery would take hold. Enter monetary policy and its ability to shape nominal spending expectations. That is what FDR did from 1933-1936 when he forcefully communicated that he wanted the price level to return to its pre-crisis level. He backed up the message by devaluing the gold content of the dollar and not sterilizing gold inflows. (See Mike Konczal for more on this experience.) It could happen here too if the Fed would commit to a level (not growth rate) target, preferably a nominal GDP level target.
QE2's limited success was not because monetary policy is impotent in such situations, but because they failed to properly shape and anchor nominal expectations. Ryan Avent summarized this problem well: QE2 changed the direction of monetary policy, but didn't set the destination. I supported QE2 and hoped the best for it. I also acknowledged, however, from the start that in the absence of a well defined level target it was bound to be limited and politically polarizing. I believe Bernanke knows all this--as is suggested by his work on Japan--but he is faced by political constraints and has burned up most of his political capital on QE1 and QE2.