July 24, 2002
Vulgar Monetarism

Vulgar Monetarism:

Do you want to show that Alan Greenspan is an idiot inflationist, who caused the bubble of the late 1990s by pumping up the money supply beyond all reason? If you're a vulgar monetarist, it's easy: you just trot out a chart of the annual rate of growth of the money stock, M3,* and your case is proved. Look at the extraordinary acceleration in the rate of growth of M3 after 1994! Look with what ludicrous and unprofessional enthusiasm the Federal Reserve pumped up the liquidity of the economy!

Do you want to show that Alan Greenspan is an idiot deflationist, who took grave risks with the economy of the late 1990s by following root-canal economics that starved the economy of liquidity, and slowed growth by artificially making money scarce? That we barely escaped disaster from this "Great Deflation?" If you're a vulgar monetarist, it's easy: you just trot out a chart of the annual rate of growth of the money stock, M1**, and your case is proved. Look at the extraordinary deceleration in the rate of growth of M1 after 1993! Look with what ludicrous and unprofessional enthusiasm the Federal Reserve tried to squeeze the life out of the economy by starving it of liquidity!

Now in the real world in which we live in--as opposed to, say, the Gamma Quadrant--Alan Greenspan is not an idiot, and anyone whose argument needs that assumption should be viewed with profound suspicion. The problem with vulgar monetarism--in all of its different and contradictory branches--is that it never acknowledges that different measures of the money stock frequently say different things. The M1 and M3 measures of the money stock were giving wildly different assessments of the state of monetary policy in the late 1990s. And if you want to blame somebody, there will almost always be some measure of the money stock you can look at in isolation and claim demonstrates that monetary policy was awry.

There are no strong theoretical reasons to prefer either one of the measures to the other (there are weak theoretical reasons to prefer M1, but they are weak indeed). There are no strong empirical reasons--no consistently tighter correlations with spending--to prefer one to the other. Thus anyone interested in helping you figure out what is going on would show you a money supply growth with more lines on it, showing just what the different concepts of the money stock are saying, and would admit that from a monetarist perspective the situation is very confusing:

Was monetary policy overly tight in 1991-92 (as M3 says), or extremely loose? Did the Federal Reserve move to tighten monetary conditions between 1993 and 1995, or did it move to loosen them? You pick your definition of the money stock--M1, M2, or M3--and you take your choice. Why do these different measures of the money stock tell different things? Because there is a lot of structural change going on in our financial markets. People are hunting for different ways in which to hold that portion of their wealth they want to keep liquid. And, perhaps most important, people's demands to hold liquid assets are not an especially stable or simple function of their spending and of interest rates. And people's demands for different kinds of relatively liquid assets are not stable. To use changes in the outstanding amount of liquid assets to figure out whether money is "cheap" or "dear" requires figuring out not just what is happening to the relative supply of but what is happening to the relative demand for that definition of the money stock. That is an extremely complicated--I would say hopeless--task.

So which measure of the money stock is the most reliable measure of monetary conditions? None of them. M3 gives the false signal that Alan Greenspan's mid-1990s monetary tightening was, instead, a monetary loosening. M1 gives the false signal that 1986 was a year of extraorinary monetary stimulus. M1 also gives the false signal that the late 1990s were a time of the most extraordinarily tight monetary policy since the end of WWII.

How, then, should we figure out what is going on in monetary policy? Given the instability of monetary demand, all kinds of monetary demand, and the weak links between changes in money stock measures and changes in spending, it is futile these days to look at money stock changes. The money stock tells you how much the economy is supplying to itself in the way of liquid--easily spendable--ways to hold wealth, but supply is uninformative without a measure that also takes account of shifts in money demand. The most obvious such measure is the short-term real interest rate: it tells how expensive it is (in terms of how much you must pay back when your borrowed loan matures) to get hold of additional cash now.

So do what Alan Greenspan does: look at interest rates, short-term real interest rates, the difference between the (nominal) interest rate on Treasury bills and the current rate of inflation. Greenspan looks at monetary policy by looking at the gap between the current level of the interest rate and the "natural real interest rate." The "natural real interest rate" is what the short-term real interest rate would be if there were no business cycles, generated by a combination of people's impatience and the productivity of capital. The interest rate set by the Federal Reserve should oscillate around this "natural real interest rate"--the actual interest rate should be higher when inflation is gathering force and the economy is overheated, and should be lower when aggregate demand is deficient and unemployment is wastefully high.

If you look at the real interest rate on short-term Treasury bills in this framework, monetary policy over the past two decades is much more intelligible than if you have to explain the false signals provided by money stock estimates as if they were deliberate shifts in policy by the Federal Reserve. From the interest rate perspective, the most important monetary policy move is the Volcker disinflation: Federal Reserve Chair Paul Volcker's decision around 1980 that interest rates needed to rise and needed to rise a lot, because the Federal Reserve had lost and badly needed to recover its credibility as an inflation-contgrolling institution. In the late 1980s, confident that inflation was under control, the Fed reduced interest rates--especially after the stock market crash of 1987. At the end of the 1980s it raised interest rates to fight inflation, and then cut them to fight the 1990-1992 recession. In the early 1990s it kept interest rates low to encourage deficit reduction, in the belief that a loose and stimulative monetary policy would damp fears that reducing the deficit would send the economy into a recession in the short run.

In the mid-1990s, the Federal Reserve raised interest rates to what it thought was a good estimate of the "natural real rate." The idea was to get back to a "neutral" monetary policy and then react to whatever was the next unfavorable shock to the economy. And for the rest of the decade, the Federal Reserve... waited. It shifted its balance from one foot to another as things like the Asian financial crisis or the collapse of LTCM caused it to assess the balance of risks differently. But real interest raets from 1995 to 2000 were remarkably stable, as the Federal Reserve watched unemployment fall, productivity growth rise, the stock market boom, the computer revolution take place, and inflation remain quiescent. There were no large moves in inflation-adjusted real interest rates from 1995 to 2000. There were fierce debates within the Federal Reserve. But their outcome was always, "let's wait a little longer and see."

Then came the recession of 2001, and the Federal Reserve did respond quickly and massively to the next large shock to hit the economy.

That's the way to tell this story of monetary policy. It makes sense. It allows you to understand what is going on, and to think about what policy should be.

So what should be done with the vulgar monetarists? What about these money-stock people who show up waving graphs with their single favorite money supply line alone? And the cacophany of their (M1) cries of a "great deflation" in the late 1990s, or of wildly overinflationary (M3) bubble-feeding in the late 1990s, or of a Fed that didn't increase the money supply enough (M3) during the 1990-1992 recession, or of a Fed that risked reigniting inflation by boosting the money stock too much (M1) in a vain attempt to reelect George Bush in 1992? Ignore them. The instability of money demand, and the inconsistency of the signals sent by different measures of the money stock, have turned vulgar monetarism from a branch of economics into a branch of palm-reading.

*The sum of currency, demand (checking account) deposis, travelers checks, other checkable deposits, retail money market mutual fund balances, savings deposits, time deposits, RPs, eurodollars, and institutional money market mutual fund balances.

**The sum of currency, demand (checking account) deposis, travelers checks, and other checkab le deposits.

Posted by DeLong at July 24, 2002 09:50 AM | Trackback

Email this entry
Email a link to this entry to:

Your email address:

Message (optional):


This is the most readable, rational and realistic tract I've seen on monetarism in a long time. Kudos.

Now, a question:

The "natural real interest rate" is what the short-term real interest rate would be if there were no business cycles, generated by a combination of people's impatience and the productivity of capital.

How is this value divined?

Posted by: George Zachar on July 24, 2002 11:43 AM

It's nice to see an explanation of that Great Deflation line in that kooky NRO article. Thanks!

Posted by: Jason McCullough on July 24, 2002 01:53 PM

I've generally heard that neutral monetary policy is a 2.5-3% real or inflation-adjusted interest rate.

Very well written article! My only concern is that it gives he general impression that the money supply isn't important, while actually it's what everything revolves around. The monetary aggregates you refer to are obviously erratic, and therein lies the monetarist dilemma.

I'm surpised you don't address any of the newer measures of the money supply, such as MZM or L. There's a good discussion of monetary aggregates at http://www.mises.org/journals/scholar/Define.pdf .

Posted by: Matthew Mullenweg on July 24, 2002 03:32 PM

It would be so cool to read this real interest graph in paralell with a decomposition of investment into its three sources of funding:
- domestic private saving
- domestic public saving and
- (inbound) (net) foreign direct investment
Expressing this as % of GDP would probably make it easiest to interpret.

With two graphs we would know the key aspects of:
- the Fed's attemps to smooth cycles with monetary policy
- the stance of fiscal policy, its impact on investment, and hence most likely on productivity, given domestic residents propensity to save and foreigners' willingness to plug for the difference.

My take is that things currently look pretty ugly from this second perspective. It may be the rignt short-term stimulatory policy but it will, I think, come at the cost of longer-run growth.

To what extent expectations of slower growth can feed back into the current conjoncture is something I have been scratching my head about... Are equity holders that forward looking? It doesn't seem so, but every now and then I hear economists say that the stock market is a good indicator of expected future inflation... Mmm... Or is it the above decomposition and its implications that are not widely understood?

Posted by: Jean-Philippe Stijns on July 24, 2002 04:45 PM

Divisia money isn't a great deal of help either ...

I think that the "neutral/natural real interest rate" is a concept which probably deserves to go the way of the Nairu. The "productivity of capital", as Joan Robinson will tell you, is a concept which is in very dubious epistemological standing indeed, for reasons which are intimately bound up with the rate of interest. (In other words, the point I'm trying to make is roughly that capital investments have varying payback periods, and thus the productivity of capital depends on the discount rate (rate of interest) which you use to put different types of capital with different payback periods on the same valuation basis).

Furthermore, "people's impatience" isn't a stationary number either. People become more or less impatient depending on their position in the life cycle; as death nears, we are less prepared to wait. And this means that the impatience of the marginal investor is a demographic fact.

All of which suggests to me that there might be mileage in characterising the bubble as a massive example of "reswitching". The timeline would be:

1) Baby boomer retirement funds pour into the market. Impatience is low, because the boomers have ages to go before they anticipate needing to draw down their savings.

2) Because impatience is low, technologies with long payoff periods become the most desirable. Let's lay global cables, build huge telecoms infrastructure, capitalise marketing costs, etc ,etc. Measured productivity is massive, because there is huge capital investment and rentiers are not demanding immediate returns.

3) Boomers getting a bit older and more impatient. At higher "natural rates", these long-payback technologies are absurd, ridiculous, unprofitable. Crunch.

I don't think this is perfect, but it captures a certain truth; the investment boom in my view was a consequence, not a cause, of the stock market rally (I say this purely to wind up Brad :)

One thing I'd suggest would be that I would look at the yield on 90-day single A commercial paper as being a truer measure of available liquidity than T-bills. There are liquidity shocks which are endogneous to the financial system, and they have real effects. 90-day CP tells a materially different story about conditions right now ...

Posted by: Daniel Davies on July 25, 2002 01:02 AM

Another example of this:

'We dodged a bullet yesterday. A threatened run on the banks on top of the plummeting stock market was halted.'

'The case is not yet open and shut. In quantity terms, the best money measure is the St. Louis Fed's adjusted monetary base, which year-to-date has grown at about a 10% annual rate. Last year that base grew 8.7%, following a deflationary drop of 2.5% in 2000. Meanwhile, the central bank's unadjusted balance sheet, referred to as Federal Reserve credit or reserve-bank credit, is currently rising at a near 11% pace, more than twice the rate registered earlier this year.'

Posted by: Jason McCullough on July 26, 2002 04:57 PM
Post a comment

Email Address:



Remember info?