Book club: Scott Sumner’s 'Alternative Approaches to Monetary Policy'
Chapter 1: What is monetary policy?
Scott Sumner has published a new book, Alternative Approaches to Monetary Policy. The book is freely available online, making it a perfect candidate for a book club. Scott’s intention is for the book to be a living document that is improved over time, so let’s help Scott do that. I will discuss each of the chapters here in turn over coming weeks. Feel free to add to the discussion in comments. If you have a newsletter and want to take-up the discussion there, let me know and I will link back.
‘Alternative Approaches to Monetary Policy’ in context
We should first put the book in the context of Scott’s work and monetary economics more generally. The book follows his 2021 The Money Illusion (TMI), which aimed to do for the financial crisis of 2008-09 what Friedman and Schwartz’s A Monetary History of the United States did for the Great Depression, namely, explain the critical role of monetary policy in driving that episode. TMI also serves as a very good primer on monetary economics.
I was pleased to have published one of Scott’s first long-form treatments of the thesis of TMI 10 years ago when I was editor of the now defunct Policy. As I noted back in 2021, TMI appeared on the 10th anniversary of Lars Christensen coining the label ‘market monetarism’ to describe Sumner’s distinctive approach, heralding the second monetarist counter-revolution (the first counter-revolution was Friedman’s post-war revival of the quantity theory). Over the last decade, market monetarism has established a broad following in academia, think tanks and elsewhere.
Scott is also the author of The Midas Paradox: Financial Markets, Government Policy Shocks, and the Great Depression (TMP) (2015). In that book, Scott re-interprets the Great Depression in terms of central bank gold reserve ratio and real wage shocks, in contrast to Friedman and Schwart’s focus on money aggregates, providing what is perhaps the only comprehensive and fully integrated economic narrative of that episode.
Sumner’s work follows in the tradition of Friedman and Schwartz in relying on narrative methods to identify the effects of monetary policy. This makes his work very accessible to a lay reader, in that it does not lean too heavily on formal economic modelling or econometrics, even though his work is very well informed by economic theory. There is some limited econometrics in TMP, but nothing a lay reader could not follow. TMP is also a great work of historical recovery in its use of primary sources. Narrative identification is a powerful tool because it relies on the historical and institutional information that is frequently ignored in formal economic modelling.
TMI and Scott’s latest book also serve as treatises on monetary economics, which I would distinguish from textbooks. A treatise was a standard mode of exposition in economics historically, but is relatively little used today. Keynes’s A Treatise on Money and Don Patinkin’s Money, Interest, and Prices, which sought to develop microfoundations for Keynesian macroeconomics, fall into this category. In the New Keynesian tradition, we also have Michael Woodford’s Interest and Prices: Foundations of a Theory of Monetary Policy.
Friedman never produced a systematic exposition of his own approach and it remains contradictory in parts. Ed Nelson restates Friedman’s work in contemporary terms in his monumental but unfinished intellectual biography of Friedman. Arthur Marget’s A Theory of Prices from the late 1930s and early 1940s is one dated exposition in the monetarist tradition, dismissed by Kaldor as ‘mid-Victorian.’ Textbooks in monetary economics typically cover a wide-range of theory, institutions and empirics but typically do not land on a coherent approach because they are trying to convey the diversity of monetary economics.
What sets Scott’s work apart is its accessibility. His success as a blogger speaks to his ability to write for a general audience. TMI serves as a very good primer on monetary economics and Scott’s distinctive approach. However, there is clearly a need for a back-to-basics exposition that is not tied to a specific macroeconomic episode, which is what Scott has attempted with AAMP.
Back to basics: What is monetary policy?
Scott starts with the most fundamental problem in monetary policy, which is the identification of the stance of policy. He highlights three broad approaches:
The quantity of money approach favoured by monetarists.
The interest rate approach favoured in the Keynesian tradition.
The price of money approach favoured by market monetarists.
The first two approaches correspond to some of the policy instruments that are typically used by central banks, namely the money supply and the official interest rate. But as Scott argues, we cannot reliably read the stance of monetary policy from these operating instruments.
The quantity of money approach, whether in terms of the money base or broader monetary aggregates, is complicated by changes in the demand for money. Monetary aggregates can sometimes provide a useful guide to the stance of monetary policy, but growth in monetary aggregates by itself cannot tell us whether the stance of monetary policy is easy or tight (or becoming easier/tighter). I discuss some of these issues in this post.
We also cannot infer the stance of monetary policy from the level of interest rates. As Friedman famously argued, but the economics profession never properly internalised, high interest rates can be indicative of high rates of inflation, pointing to monetary policy being too loose. By the same token, low interest rates can be an indication that monetary policy is too tight. Because we do not know the (time-varying) real equilibrium interest rate, we cannot reliably read the stance of monetary policy from the level of interest rates.
Sumner argues instead for the price of money approach, which is perhaps one of his most important contributions. The price of money does not refer to interest rates, a common misconception. Interest rates are the price of credit. Instead, the price of money is the reciprocal of the general price level. However, the price of money approach can be generalised to other prices use to denominated trade in goods and services and other transactions, such as the exchange rate or even commodity prices.
As Sumner puts it:
The ultimate goal is to find a policy indicator that is as clearly controllable as the monetary base or short-term interest rates, and as closely linked to our policy goals as inflation or [nominal gross domestic product] NDGP growth (p. 17).
Sumner argues that:
The optimal instrument or indicator is some sort of financial asset price that is closely linked to the macroeconomic goal variables.
Long-time readers of Sumner’s work will not be surprised to learn that he favours the price of nominal GDP futures contracts as both the optimal monetary policy instrument and indicator.
The monetary policy transmission mechanism
An important advantage of this approach is that it abstracts from the monetary policy transmission mechanism, that is, how changes in monetary policy instruments work their way through various channels such as investment spending and mortgage interest rates to influence the macroeconomic goal variables, such as inflation and employment. The price of money approach allows us to be agnostic about the precise transmission mechanism, because we want to assess the stance of monetary policy based on macroeconomic outcomes, not the stance of policy instruments or intermediate variables.
Sumner argues that ‘no one knows how monetary policy works’ (p. 22). While I think this is an over-statement, Sumner is right to argue we spend too much time trying to identify and quantify the various transmission mechanisms. Sumner is a pragmatist and maintains that debates about the monetary transmission mechanism have ‘no important theoretical implications.’ Again, I think this is over-stating things, but a preoccupation with the transmission mechanism can distract from the outcomes policymakers should care about.
As a thought experiment, Sumner suggests a monetary policy regime that targets five-year Treasury Inflation Protected Securities (TIPS) spreads at a rate of 2% (a regime like this has been a long-standing policy proposal in the monetarist tradition). Such a regime would not imply that TIPS spreads caused monetary policy to affect the economy. Rather, the TIPS spread is just a financial market price, but one that reflects both the stance of monetary policy and its macroeconomic effects.
Summary
Chapter One is a succinct re-statement of some of the core propositions that have informed Sumner’s distinctive approach to monetary policy and monetary economics. As Sumner notes, there is nothing particularly radical or unorthodox about these propositions. He quotes Ben Bernanke making essentially the same case in his famous 2003 speech on the legacy of Milton and Rose Friedman. Sumner’s approach is in many ways just a more rigorous and comprehensive application of the discipline’s own orthodoxy. Much of Scott’s work and commentary is built around the intellectual slippage that occurs between monetary theory and monetary policy practice, which leads to bad, and sometimes disastrous, policy outcomes.
Many of the monetary policy misconceptions addressed by Sumner’s work are apparent in financial market discussions of monetary policy. That might be thought to argue against conditioning monetary policy on financial market prices. But those prices coordinate tacit knowledge in ways that do not necessarily depend on the beliefs of individual market participants.
In the next installment of our book club, we will consider Chapter Two: The Strange World of Interwar Monetary Policy.
ICYMI
The New Libertarian Elitists. What’s behind the dangerous new notion that democracy should be left to the well-educated.
Memes and themes:
I wish I had seen this in real time.
Perhaps now is too late, but I would question the need for the concept of "stance" and consequently how to identify it. I think it is sufficient to opine that the central bank should use one or more of it policy instruments to increase or decrease one or more of its objectives which in the US have been defined for it in legislation: stable prices and maximum employment. Granted that the legislation (fortunately) does not define either goal, leaving the Fed free to define a flexible average inflation target (FAIT) as "stable." And presumably the value of the target and the degree of flexibility take account of maximizing employment of resources (not just labor)
Steven, coincidentally, on the same day Scott´s book came on line, I posted "When Monetary Policy tightening is not what you think", a "narrative approach" covering the last 30 years.
https://marcusnunes.substack.com/p/when-monetary-policy-tightening-is