As I was writing my recent post on the discussion of the importance of expectations in the lead-lag structure in the monetary transmission mechanism I came think that is really somewhat odd how little role the discussion of expectations have had in the history of the theory of transmission mechanism .
Yes, we can find discussions of expectations in the works of for example Ludwig von Mises, John Maynard Keynes and Frank Knight. However, these discussions are not directly linked to the monetary transmission mechanism and it was not really before the development of rational expectations models in the 1970s that expectations started to entering into monetary theory. Today of course New Keynesians, New Classical economists and of course most notably Market Monetarists acknowledge the central role of expectations. While most monetary policy makers still seem rather ignorant about the connection between the monetary transmission mechanism and expectations. And even fewer acknowledge that monetary policy basically becomes endogenous in a world of a perfectly credible nominal target.
A good example of this disconnect between the view of expectations and the view of the monetary transmission mechanism is of course the works of Milton Friedman. Friedman more less prior to the Muth’s famous paper on rational expectation came to the conclusion that you can’t fool everybody all of the time and as consequence monetary policy can not permanently be use to exploit a trade-off between unemployment and inflation. This is of course was one of things that got him his Nobel Prize. However, Friedman to his death continued to talk about monetary policy as working with long and variable lags. However, why would there be long and variable lags if monetary policy was perfectly credible and the economic agents have rational expectations? One answer is – as I earlier suggested – that monetary policy in no way was credible when Friedman did his research on monetary theory and policy. One can say Friedman helped develop rational expectation theory, but never grasped that this would be quite important for how we understand the monetary transmission mechanism.
Friedman, however, was not along. Basically nobody (please correct me if I am wrong!!) prior to the development of New Keynesian theory talked seriously about the importance of expectations in the monetary transmission mechanism. The issue, however, was not ignored. Hence, at the centre of the debate about the gold standard in the 1930s was of course the discussion of the need to tight the hands of policy makers. And Kydland and Prescott did not invent Rules vs Discretion. Henry Simons of course in his famous paper Rules versus Authorities in Monetary Policy from 1936 discussed the issue at length. So in some way economists have always known the importance of expectations in monetary theory. However, they have said, very little about the importance of expectation in the monetary transmission mechanism.
Therefore in many ways the key contribution of Market Monetarism to the development of monetary theory might be that we fully acknowledge the importance of expectations in the transmission mechanism. Yes, New Keynesian like Mike Woodford and Gauti Eggertsson also understand the importance of expectations in the transmission mechanism, but their view of the transmission mechanism seems uniformly focused in the expectations of the future path of real interest rates rather than on a much broader set of asset prices.
However, I might be missing something here so I am very interested in hearing what my readers have to say about this issue. Can we find any pre-rational expectations economists that had expectations at the core of there understand of the monetary transmission mechanism? Cassel? Hawtrey? Wicksell? I am not sure…
PS Don’t say Hayek he missed up badly with expectations in Prices and Production…
PPS I will be in London in the coming days on business so I am not sure I will have much time for blogging, but I will make sure to speak a lot about monetary policy…
Nick Rowe
/ March 13, 2012Lars: “Basically nobody (please correct me if I am wrong!!) prior to the development of New Keynesian theory talked seriously about the importance of expectations in the monetary transmission mechanism.”
IIRC, David Laidler did, in 1978, in his paper “Money and Money income:an essay on the ‘transmission mechanism'”.
My copy is in office, and I can’t download this from home:
http://ideas.repec.org/a/eee/moneco/v4y1978i2p151-191.html
Now, whether he said *enough* about expectations, i can’t clearly recall.
Lars Christensen
/ March 13, 2012Nick, thanks a lot. I will certainly have a look at the paper. You know better than anybody how great an economist David Laidler is.
By the way I would still love to hear what Laidler thinks of the market monetarist explanation of the Great Recession.
Lee Kelly
/ March 13, 2012I just posted this over on Scott’s blog. Am I wrong about this? Here is the comment in full:
There seem to be two versions of the expectations argument.
Suppose the Fed announces an NGDP level target (with some added catch up in the first couple of years)
In the Keynesian version, a sceptical market wins. The Fed cannot actually increase NGDP by brute force alone, because increasing the money supply is ineffective at the zero nominal bound. All it can do it promise to increase NGDP or inflation (or lower interest rates) more once it is no longer constrained by the zero nominal bound. If the market doesn’t believe the announcement, then nothing happens and the Fed remains stuck at the zero nominal bound.
In the market monetarist version, the Fed always wins. The Fed can increase NGDP despite a sceptical market by brute force alone, because money demand is never insatiable. Whether or not the market was initially sceptical doesn’t matter, because it’ll soon acknowledge rising NGDP and come to believe in the new target. Most of the heavy lifting, so to speak, will be done by the changing expectations, but the Fed doesn’t rely on expectations entirely.
Max
/ March 13, 2012Lee, if the central bank (or Treasury) threatened to buy private financial assets (not government bonds), then a level target would be highly credible.
MMs are correct to focus on expectations and level targeting, but they go wrong by obsessing over the “quantity of money”.
George Farnon
/ March 13, 2012I’d give a cheeky push for Thornton to be the earliest? Explaining how a central bank that believed the real bills doctrine could possibly generate unlimited inflation if the expected rate of profits for merchants exceeded the lending rate..
Ratex was a structural break because any of the previous expectation models (And there were plenty – Fisher’s distributed lag, adaptive, extrapolative, error-learning, regressive, etc.) could be compatible with it under specific circumstances. Expectations are always highly subjective and the failure of Friedman was in trying to generalize a specific model. So I would put Friedman in the earlier tradition rather than as a direct precursor to ratex. I just so happened that in his state of the world where ‘monetary policy in no way was credible’ ratex and adaptive might have well been the same thing.
The later tradition, Ratex was pretty much an import from optimal control/stochastic process theory (which only developed in the 1950’s itself). Muth 61 was a response to Grunberg/Modigliani 1954 “the predictability of social events” and Simon 1954 “possiblity of election predictions” – Cowles began to portray agents as mini-econometricians after their attempts to empirically verify earlier models failed.
And this was the walrasians’ response to the recasting of the economic agent as an information processor… So (albeit indirectly) this second tradition was Hayek’s doing all along! 😉
David Eagle
/ March 13, 2012Lars notes that (with some minor grammar corrections in brackets) “Friedman more [or] less prior to the Muth’s famous paper on rational expectation came to the conclusion that you can’t fool everybody all of the time and as consequence monetary policy [cannot] permanently be [used] to exploit a trade-off between unemployment and inflation. This is of course was one of things that got him his Nobel Prize.” I wish to remind everyone that Sargent and Wallace (JPE, 1975) used rational expectations to argue their monetary policy irrelevance proposition, which is part of the reason Sargent was co-winner of this year’s Nobel Prize. That monetary policy irrelevance proposition is what Lars described as “the conclusion that you can’t fool everybody all the time and as a consequence monetary policy [cannot] permanently be [used] to exploit a trade-off between unemployment and inflation.” Thus, my understanding is that conventional economic thought now gives credit to Sargent and Wallace for this conclusion, not to Friedman, although that conclusion was consistent with Friedman’s monetarist movement. Of course, some quotes from Friedman could convince me otherwise.
Scott Sumner has pointed me to a very good speech by Ben Bernanke in 2003 (except that it ends with an inflation-targeting flavor instead of supporting nominal GDP targeting)(http://federalreserve.gov/boarddocs/speeches/2003/20031024/default.htm). In this speech, Bernanke summarized Friedman’s views how monetary policy affects the economy. Bernanke lists as Friedman’s seventh proposition:
“7. Although money growth can affect output in the short run, in the long run output is determined strictly by real factors, such as enterprise and thrift.”
This statement is not as strong as Sargent and Wallace’s statement, so until I see quotes from Friedman that indicate otherwise, I am going to side with the conventional economic thought of giving Sargent and Wallace credit for the monetary policy irrelevance principle rather than giving that credit to Friedman.
George Farnon
/ March 13, 2012David… Friedman in his own words of ‘what monetary policy can do’ http://www.aeaweb.org/aer/top20/58.1.1-17.pdf
Steve
/ March 13, 2012Lars,
Good post. I should read your blog more often.
I think everyone who practices finance intuitively understands that expectations are everything, but expectations isn’t an idea that translates cleanly into an academically analytical framework.
Your point on Friedman is spot on. In my effort to creating catchy slogans for Market Monetarism, I would rephrase Friedman: “Credibility is earned with long and variable lags.”
Martin
/ March 13, 2012@David,
““7. Although money growth can affect output in the short run, in the long run output is determined strictly by real factors, such as enterprise and thrift.””
This seems to me a proposition with which the classical economists would agree with as well (ever since Hume?): long run neutrality of money. In fact, I believe this is what Hayek used in his theory of the cycle. Crudely stated: because money cannot affect output in the long run, a money induced growth in output has to be self-reversing.
Mikael Olai Milhøj
/ March 13, 2012“However, Friedman to his death continued to talk about monetary policy as working with long and variable lags. However, why would there be long and variable lags if monetary policy was perfectly credible and the economic agents have rational expectations?”
Part of the answer is that in-real-life agents are not rational, but are forward-looking and learning. Rational expectations are a better approximation than different more naive expectations, but it is not a perfect expectation-theory at all. I’m must more into theories like “adaptive learning” which are better at explaining real behaviour.
Jonathan M.F. Catalán
/ March 14, 2012I would look into Hawtrey. If Hicks is right, then Hawtrey had a lot to say about expectations, monetary policy, and the “transmission mechanism.”
Sophia Borree
/ November 20, 2015Hey thanks for the informative information. I will keep it in my mind. Automatic transmission