In my previous post, I looked at the development of modern macroeconomics. Several commenters responded by discussing what they thought was wrong with macro. Here I’ll put in my own two cents, and then explain how my views relate to those of my commenters.
In my view, the biggest problem with modern macro is the treatment of monetary policy. Simply put, modern macroeconomists don’t know what monetary policy is.
Of course I’m slightly exaggerating. Economists have a general idea as to what the term monetary policy means, but the concept remains frustratingly vague. Exactly what is monetary policy? I feel so strongly about this issue that I recently published an entire book on the subject.
Here are a couple excellent comments from the previous post:
Thomas Hutcheson:
For me the biggest failing of macro modeling is leaving out (leaving implicit) the Fed’s reaction function. It the modeler guesses right, then the model can be useful in examining other variables, fiscal “policy” for example or shocks to the price of petroleum.
Michael Sandifer:
The most disappointing thing about the development of macroeconomics to me, as a non-economist, is the lack of progress at understanding the intersection of macroeconomics and finance, particularly with regard to liquid asset markets.
What is the fiscal multiplier? Without knowing the Fed’s reaction function, that’s not even a coherent question. Suppose the Fed were an NGDP targeter. If fiscal policy changed, then the Fed would adjust its interest rates target so that NGDP growth would be expected to remain on target. In that case, in what sense is there any “fiscal multiplier”? And even if interest rates are stuck at zero (as in late 2012) the Fed adjusted other tools like forward guidance and QE in order to offset the impact of fiscal austerity.
If you argue that the fiscal multiplier is the effect of fiscal policy on GDP holding monetary policy constant, then you must define what you mean by “monetary policy”. The money supply? The nominal interest rate? The real interest rate? The gap between the policy rate and the natural interest rate? There are lots of possibilities.
In my view, there is only one useful definition of the stance of monetary policy—the expected future level of the nominal aggregate being targeted. If NGDP is the target (which is my preference), then the future expected level of NGDP measures the current stance of monetary policy.
Sandifer is correct that macroeconomists pay too little attention to financial asset prices. Any efficient estimate of future expected NGDP would involve at least some financial asset prices. In a perfect world, we’d have a highly liquid NGDP futures market. The price of NGDP futures would represent the current stance of monetary—which would move around in real time. Even without this market, we can look at a wide range of other asset prices (including things like TIPS spreads) and infer a rough estimate of the market forecast for NGDP growth.
Macroeconomics will never become a mature field until we get serious about defining the stance of monetary policy. Step one is to abandon interest rates as an indicator of policy and move on to more promising alternatives—especially market forecasts of future NGDP.
READER COMMENTS
Kurt Schuler
Sep 29 2023 at 9:54pm
“In my view, there is only one useful definition of the stance of monetary policy—the expected future level of the nominal aggregate being targeted.”
I can imagine a policy that targets the real exchange rate. Whether you think that’s a good idea or a bad one, isn’t it then the stance of the expected future level of a real magnitude that is important?
By the way, the post that preceded this one was, I thought, unusually good even for you.
Scott Sumner
Sep 30 2023 at 12:42pm
Kurt, I suppose this is a question of usefulness. Yes, you could define monetary policy in terms of a real exchange rate target. However, I don’t believe it’s possible to target real variables in the long run, so I don’t see that as a particularly useful definition.
At the same time, I realize that my preferred definition is not objectively true, rather it’s an approach that I happen to find useful.
Thanks for the comment on the previous post.
Thomas Hutcheson
Sep 30 2023 at 8:22am
“Simply put, modern macroeconomists don’t know what monetary policy is.” I’m reminded of the joke from the time of Bretton Woods that only two people really understood the system. One was in the Bank of England and the other was in the Bank of France and that they did not agree.
“In my view, there is only one useful definition of the stance of monetary policy—the expected future level of the nominal aggregate being targeted.”That seems close to right to me. The Fed should have in mind a trajectory of settings of its instruments that, conditioned on its expectations of changes in exogenous variables, will result in the future levels macroeconomic outcomes it is targeting. Unless it can explain very well what those expectations of exogenous variables are (which I doubt) it should make no speculation about its future actions beyond that it will do at each point in time whatever it thinks best at that point to achieve its target(s). I do not see the value of “stance” or “policy indicator.”Whatever we are targeting (I think in practice a FAIT and NGDPLT can be equivalent) a liquid NGDP futures market would be a big help to the Fed. But would not a Trillionth security be the same thing and easier todo at the stroke of a pen? Could the Fed itself create and trade in such a security? And shouldn’t we have more intermediate tenor TIPS, too?
spencer
Sep 30 2023 at 1:15pm
re: “Step one is to abandon interest rates as an indicator of policy”
Even Bloomberg gets this wrong. Savings flowing through the nonbanks never leaves the payment’s system. But there is an increase in the supply of loan-funds (affecting interest rates), which lets the MMMFs purchase additional securities (where S = I).
How do you think the Treasury funded the 3rd quarter trillion-dollar debt issuance?
Michael Sandifer
Oct 2 2023 at 10:34am
Thanks for the compliment. Good post.
Off-topic, I’ve been using ChatGPT-4’s new data analysis tool, and discovered it’s far more capable econometrically than I expected. It can quickly do stationarity tests, Granger causation tests, VARs, etc. Quite interesting and useful, though it can be frustrating until you learn how to use it.
spencer
Oct 2 2023 at 11:00am
The earliest construction of Fisher’s “equation of exchange” represented money flows and the value of all transactions. It degraded into the relationship between money flows and GDP. Transaction’s velocity can, and has, moved in the opposite direction as income velocity.
rick shapiro
Oct 5 2023 at 10:13am
Fiscal (not monetary) policy is the one worthy of attention. But paralysis of fiscal policy abdicates macroeconomics to monetary policy, which basically pushes on a string. Consequently, the fed pushes so hard that they break things, keeping rates so low that they provoked asset inflation, and now continuing to raise rates before allowing raises to work through the economy, thereby assuring the next recession.Monetary policy that pushes so hard that it actually has an effect, suffers from enormous instability resulting just from delay in effect, no matter which form of targeting is chosen. Viewing monetary policy as the complex transfer function in a negative-feedback control system, these delays inevitably cause transfer function zeros to appear in the positive half-plane. At best this results in instability (in economic terms, extreme business cycles), at worst in pegging to a voltage rail (in Keynesian terms, switching to an inferior equilibrium).
Comments are closed.