A few people have been asking me about this, so here is a summary statement of some points:
1. The liquidity trap argument implies that the money-short bonds margin doesn't, at current magnitudes, matter. I am more closely wedded to marginalism than that. The argument also sees all the action (or should I say, non-action) in one margin, the money-bonds margin. The money-goods margin matters too. In the liquidity trap argument, everything is decided by one irrelevance result at a single margin in a highly complex multi-trillion dollar economy.
2. Short-term interest rates being zero, and short-term interest rates being almost zero, are very different cases, especially for understanding nominal shocks and whether they can stimulate aggregate demand. Unless short-term rates are literally the same as the rate on cash, asset swaps still can succeed. And QEII isn't be the same as simply switching the term maturity of the debt, as Krugman has suggested. There will be nominal effects also.
3. Even Keynes didn't believe he had ever seen a liquidity trap, including in the Great Depression.
4. Most of the good predictions of the liquidity trap models are covered by recognizing there are lots of unemployed resources and weak business confidence.
5. Consumer spending just rose 2.6 percent and business profits are high, yet we are in a liquidity trap? What exactly is the story here? "We can get spending up by 2.6 percent but not a smidgen higher"?
6. The stock market responded positively to the announcement of QEII and the TIPS spread went negative; both are the opposite of what a liquidity trap model would predict. Markets don't seem to think the liquidity trap idea is a very useful one and that alone creates real positive effects from monetary policy. If markets don't believe in a liquidity trap, that's enough for the trap not to bind.
7. Monetary policy worked quite well in the Great Depression, when it was tried. And on Japan I am persuaded by Scott Sumner.
These points are a less fundamental, but they are still relevant to the debate, if not always to the substantive issue itself:
7. There are different liquidity trap models. For instance I often see the "short nominal rates are zero" model being confused with the more stringent "money demand is a bottomless sink" model. It's only the latter case — clearly not true today — which generates the extreme results of liquidity trap models. Otherwise the money-goods margin remains operative and monetary policy can succeed.
8. Some of the LT models imply upward-sloping AD curves and downward-sloping AS curves, yet I don't see anyone applying those assumptions consistently to other decisions, such as tax policy for instance.
9. If a liquidity trap is to persist beyond the short run, something must be preventing the marginal product of capital from adjusting upwards and solving the problem. In other words, a non-short-run version of the liquidity trap has to be combined with an account of problems on the real side.
10. I see liquidity trap proponents spending a lot of time criticizing naive versions of real business cycle theory, bond market vigilante arguments, and so on. They spend less time engaging the most serious criticisms of the liquidity trap argument. A study of the liquidity trap literature does not raise one's confidence in the idea, though it might sound OK if the relevant alternative seems sufficiently bad.
11. Whether or not we should use fiscal policy depends on how well our government can target and mobilize unemployed resources; you don't need a liquidity trap to address that argument.
12. Ultimately I view the liquidity trap idea as a kind of shaggy dog that's been pulled out of the closet. If it's going to be made convincing, it needs a lot more work than simply repeating that some short-term interest rates are near zero.