Joseph Gagnon recommends a three-pronged attack by the Fed:

First, the Fed should lower the interest rate it pays on bank reserves to zero….Three-month Treasury bills currently yield 0.15 percent, and that rate, too, should be brought down to zero. Second, the Fed should bring down the rates on longer-term Treasury securities by targeting the interest rate on 3-year Treasury notes at 0.25 percent and aggressively purchasing such securities whenever their yield exceeds the target. That is a 65-basis point reduction from the current rate of 0.90 percent…. Finally, the Fed could bolster the stimulative effects of these actions by establishing a full-allotment lending facility to enable banks to borrow (with high-quality collateral) at terms of up to 24 months at a fixed interest rate of 0.25 percent.

In response to this suggestion, Krugman writes:

We don’t know how well the Gagnon plan would actually work — but there’s no harm in trying, and large potential benefits. The only possible reason for the Fed not to be more aggressive now is fear of embarrassment, of not getting big results.

And now we really see Krugman’s cockiness, with the parts I’ve put in bold. I can understand if you’re using a Keynesian model, and you think that all things considered, in the real world it would be a good idea for the Fed to get more aggressive.

But to say that Gagnon’s quite specific recommendations have no possible downside?! Krugman really can’t see even a theoretical drawback to the Fed setting up a new lending facility?

In case you think I’m being a hypocrite, consider something on my side of the fence. In Carlos and my new book (which will be available for purchase and free download very soon–we’re just working out the kinks), we have a section talking about how the Fed and Treasury could link the dollar back to gold. Now I will be the first to admit that the specific plan we outline, could blow up in our faces. We think it would be better than the status quo, but it’s possible if we implemented the plan, that something could go wrong. There’s always problems like that, when you try to tinker with the regulatory apparatus.