This was from his 1933 Nobel acceptance speech (!):
I should like to suggest to you that the
cause of all the economic troubles is that we have an economic
system which tries to maintain an equality of value between two
things, which it would be better to recognise from the beginning
as of unequal value. These two things are the receipt of a
certain single payment (say 100 crowns) and the receipt of a
regular income (say 3 crowns a year) through all eternity. The
course of events is continually showing that the second of these
is more highly valued than the first. The shortage of buyers,
which the world is suffering from, is readily understood, not as
due to people not wishing to obtain possession of goods, but as
people being unwilling to part with something which might earn a
regular income in exchange for those goods. May I ask you to
trace out for yourselves how all the obscurities become clear, if
one assumes from the beginning that a regular income is worth
incomparably more, in fact infinitely more, in the mathematical
sense, than any single payment? In doing so I think you would
then get a better insight into the way in which a physical theory
is fitted in with the facts than you could get from studying
popular books on physics.
I thank Eric R. Weinstein (Twitter!) for the pointer.
Of course you can read this paragraph as offering differing (better?) microfoundations for a liquidity trap argument. The problem is not expectations of unfavorable interest rate changes (no one buys bonds with cash), but rather no one parts with bonds to receive goods and services. Is it a behavioral argument, namely that no one wants to give up the feeling of "forever" for the feeling of getting something which is only "temporary." Or is it a maximizing argument, namely that a kind of zero discount rate hyper-rationality takes over the people who won't spend? Does the argument require that consols are the dominant form of bonds?