Recall what Keynes wrote (p. 207):
"(2) There is the possibility, for the reasons discussed above, that, after the rate of interest has fallen to a certain level, liquidity-preference may become virtually absolute in the sense that almost everyone prefers cash to holding a debt which yields so low a rate of interest. In this event the monetary authority would have lost effective control over the rate of interest. But whilst this limiting case might become practically important in future, I know of no example of it hitherto. Indeed, owing to the unwillingness of most monetary authorities to deal boldly in debts of long term, there has not been much opportunity for a test. Moreover, if such a situation were to arise, it would mean that the public authority itself could borrow through the banking system on an unlimited scale at a nominal rate of interest."
So Scott is right - this really does seem like a problem for this particular passage, at least. One interesting thing to note is that the last sentence Keynes has here seems right whether the first part of the paragraph holds true or not (in other words - the case for fiscal policy still stands just fine - a point that Krugman makes, and argues is independent of all those "Keynesian things").
The question, of course, is why is he wrong? I don't feel like I have guidance on that from Scott. Why would investors prefer a negative yield to a zero yield assuming both assets have negligible risk? This isn't a Keynesian question that deserves an answer - it's a basic rationality question that deserves an anwer. My first thought roamed into international economics (which makes me nervous beause I don't feel like I have a good intuition for it). If foreigners are buying a lot of Treasuries and there is some exchange rate risk involved, the yield on cash may not be zero. But that doesn't seem to work because the exchange rate risk is going to impact demand for Treasuries (which are denominated in dollars) as well. Perhaps transaction costs associated with working with bonds is lower for foreigners than cash? I don't know.
The only other thing I could think of is that building a cash reserve must be limited somehow, so while a zero yield would be preferable its not a viable option. Is this true? That would seem to explain it. If the reason why interest rates can transcend the lower bound of the liquidity trap is that the supply of cash is constrained, you would want policy that creates more cash: like public works that create jobs or more bond purchases. If you'd like people to get jobs before banks get cash to sit on, you might prefer the former. If you'd like banks to get cash to sit on before the jobs start flowing, you might prefer the latter.
I also wanted to point out that Barkley Rosser is talking about liquidity traps this morning, and he is particularly taking economists to task for allegely rejeting the idea of negative prices. Again, like the Sumner post, the discussion is interesting but it doesn't get at the real question that's bugging me: fine, I accept prices can be negative (although can't that just be thought of as a flip-flop of the role of buyer and seller? No matter). The real confusing question for me is how this price could be negative when the price of an otherwise equal product is zero? Barkley's example (brides who alternatively trade for dowries or bride prices) isn't relevant here because the brides aren't the same quality product. The whole problem with a liquidity trap is that cash and government bonds presumably are. Barkley's first commenter shares my concern for the real question here: "So what's the mechanism by which negative interest rate bonds come about?". He comes to the same conclusion about the limited supply of cash, and he is more amenable to the foreign flight to quality issue (I'm still not so sure... flying to quality is sensible, but why fly to American bonds rather than American cash??).
I feel too dumb to answer this, but I feel smart enough to confidently say that Scott and Barkley and no one else I've seen has answered it either.
There is another disconcerting implication of all of this. On page 202 of the General Theory, Keynes points out that "...if the general view as to what is a safe level of r is unchanged, every fall in r reduces the market rate relatively to the "safe" rate and therefore inreases the risk of illiquidity; and, in the second place, every fall in r reduces the current earnings from illiquidity, which are available as a sort of insurance premium to offset the risk of loss on capital account, by an amount equal to the difference between the squares of the old rate of interest and the new. For example, if the rate of interest on a long-term debt is 4 per cent., it is preferable to sacrifice liquidity unless on a balance of probabilities it is feared that the long-term rate of interest may rise faster than by 4 per cent. of itself per annum, i.e. by an amount greater than 0.16 per cent. per annum. If, however, the rate of interest is already as low as 2 per cent., the running yield will only offset a rise in it of as little as 0.04 per cent. per annum. This, indeed, is perhaps the chief obstacle to a fall in the rate of interest to a very low level. Unless reasons are believed to exist why future experience will be very different from past experience, a long-term rate of interest of (say) 2 per cent. leaves more to fear than to hope, and offers, at the same time, a running yield which is only sufficient to offset a very small measure of fear."
We usually hear the "more to fear than to hope" logic as we approach the zero lower bound, right? But these calculations squaring the interest rate are symmetric.
What does it say about expectations for the future path of interest rates if interest rates are currently negative?
This sort of thinking explodes quickly. Part of me wants to paraphrase Herbert Stein and say that path is not sustainable, so it won't be sustained. There must be something driving this other than expectations of future interest rate paths, because that answer is the explosive answer. Another part of me thinks that even if the explosive solution doesn't come about, the fact that investors aren't afraid of a negative yield says something real about their expectations of future yields.
I minor pet peeve on the workings of the internet - no need to comment on this in the comment section - why in the world are there more Austrian links in the reference list of the Wikipedia entry on liquidity traps than Keynesian links? It's really a shame that people who try to learn about economics from the internet get a such a distorted sense of the field, simply because old Mises Institute blog posts are so accessible relative to the actual economics literature.