First, I raise some issues about a current class of models of monetary transmission in
which a short-term interest rate represents the transmission process. The class of models is so
widely accepted that the conclusions I challenge have become part of the canon. A different
class of models -- a more useful one, I believe -- does more than give different answers. Some
issues do not arise; they are no longer relevant. And some issues remain relevant but receive a
different answer. The role of money is one such issue.
Second, I discuss some of the evidence I have gathered from my study -- A History of the
Federal Reserve -- the work that has been my main occupation for the past four years. The two
pieces are joined, as I hope to show. The evidence from history shows that the transmission
process cannot be summarized by a single interest rate. In the final section, I present some
econometric evidence to supplement the historical data.