The Theory of Monetary Policy, Then and Now

Has Confusion Increased or Decreased?

In 1975, Benjamin Friedman (no relation to Milton) published Targets, Instruments and Indicators of Monetary Policy. It helped to clarify issues that monetary economists considered important at the time.

Friedman’s idea was to separate the three concepts in the title. The Fed’s target variable could be, for example, the rate of inflation. Its instruments might be open market operations, changing the discount rate, or changing reserve requirements (for what those mean, consult a source in freshman economics). Two indicators that Friedman contrasted were the short-term interest rate and the rate of growth of the money supply.

In fact, what the Fed tried to do (and still tries to do, as far as I can tell), is assemble a forecast for its target variable(s) and adjust policy when the forecast deviates from the target number. For example, if the Fed’s target for inflation is 2 percent, and the current forecast is that it will be 3 percent, then the Fed knows that it has to engage in contractionary policy until the forecast drops to 2 percent. If the current forecast instead is 1 percent, then the Fed has to engage in expansionary policy until the forecast rises to 2 percent.

Milton Friedman claimed that this approach of targeting the forecast was likely to fail, with the Fed alternating between undershooting and overshooting. He thought that the “long and variable lags” between changing an instrument and seeing the results of that change would prove destabilizing. The Fed might see 3 percent inflation coming, tighten policy, and produce 1 percent inflation. Then it would loosen, producing 5 percent inflation. Then it would over-tighten, and so on. “Like an amateur shower tuner, getting alternately scalded and chilled,” in James Tobin’s pithy metaphor.

Milton Friedman said that the Fed could avoid overshooting by keeping the growth of the money supply stable. In Benjamin Friedman’s terminology, Milton thought that the money supply growth rate was a more reliable indicator than the Fed’s current assessment of economic conditions.

This controversy never really got settled. Some monetarists insist that Milton Friedman’s claim that inflation is determined by the growth in the money supply, with “long and variable lags,” is still correct. Many economists, myself included, believe that with the modern financial system that includes many more instruments than the cash, checking accounts, and savings accounts that were predominant in Milton Friedman’s theory, the growth of the money supply has become less reliable as an indicator.

But starting about the time that Benjamin Friedman’s paper appeared, the “rational expectations revolution” started to sweep through academic macroeconomics. This put the spotlight on what markets are expecting about the future.

One way to take expectations into account is for the Fed to use market forecasts rather than internal forecasts. For example, the market might forecast 3 percent inflation, and the Fed could use that as a basis for tightening policy. That approach was dubbed Market Monetarism and is associated with Scott Sumner, among others.

Market Monetarism is revolutionary only to the extent that market forecasts differ from the Fed’s internal forecasts. Sumner and others place great faith in market forecasts (that is consistent with the “rational expectations revolution.”) But given the way that Fed staff and Wall Street economists draw from the same pool of information and analytical methods, I do not think one is likely to find a significant difference between market forecasts and internal forecasts. Given that market forecasts and internal Fed forecasts tend to be closely aligned, the shiny new Market Monetarism is just the same-old way the Fed used to do business. With market forecasts no better than internal forecasts, the Fed could still act like an amateur shower tuner.

Another way to take expectations into account is for the Fed to offer guidance concerning its future actions. For example, several months ago, the Fed announced that it was planning a series of increases in its target for the interest rate on Fed Funds.

“Forward guidance” gives the Fed a new policy instrument. When the market believes that the Fed plans a series of Fed Funds rate increases, this should lead investors to raise interest rates on longer-term bonds more than they would if they thought that the Fed were going to keep the Fed Funds rate the same going forward.

Forward guidance is an attempt to shape market expectations. But if investors were to have a very different forecast from the Fed, they could always second-guess the Fed. For example, if they think that the Fed’s guidance for future interest rate increases would cause inflation to undershoot its target, investors could build in expected future interest rates that are lower than the Fed’s guidance.

Here is a summary of where I think we are:

1. The Fed tries, and has always tried, to use its instruments to hit a target, using a forecast.

2. The forecast that the Fed uses is very close to the forecast that money market participants use, because the Fed and Wall Street investors think alike and communicate with one another.

3. It is possible that Milton Friedman’s warning is valid, and the Fed will alternate between over-loosening and over-tightening.

4. Milton Friedman’s solution, of using the growth rate of money as an indicator, is less likely to work today, because the connection between his measures of the money supply and economic activity has gotten looser.

5. The Fed has a new instrument, “forward guidance.” But the market still has the option of second-guessing the Fed.

All of those five points stand on their own. Your belief in those points should not change when you read the rest of this essay.

A Heterodox View

I cannot discuss the theory of monetary policy without mentioning a heterodox view that I hold. What I will suggest is that the Fed has little influence over the course of the economy these days. Why does almost everyone disagree with me? I think that we are very biased to imagine that someone is in charge of the economy, and the Fed has emerged as the Wizard of Oz in which everybody believes. (I am not trying to refer to the claim that L. Frank Baum was writing a parable of monetary theory. I’m just using the Wizard of Oz as a metaphor for somebody who is believed to have magical powers but does not.)

Instead, I think that the Fed’s main role is to help manage government debt. As John Cochrane put it the other day,

For broad brush macroeconomics, the Fed and Treasury are left and right pockets of the federal government.

In recent years, the debt managers at the U.S. Treasury (the right pocket) tried to reduce taxpayer risk by issuing some long-term debt, in case interest rates might rise subsequently. But the Fed (the left pocket), using “quantitative easing,”bought the long-term debt and issued short-term liabilities, such as “interest on reserves” and “reverse repo.” When inflation and interest rates rose, the Treasury’s debt managers reaped a windfall, as the buyers of long-term bonds were stuck with low yields. Unfortunately for taxpayers, those buyers were mostly the Fed. So the gain to the right pocket (Treasury) was offset by a gigantic loss in the left pocket (the Fed).

The preceding paragraph is not all that heterodox. It is just a clear-eyed view of the meaning of “quantitative easing” from 2008 to present.

Now for the heterodox part. I think of the Fed as just another bank, like Citicorp. We don’t think that Citicorp controls inflation, unemployment, or nominal GDP. Neither does the Fed, in my heterodox view.

Yes, unlike Citicorp, some of the Fed’s liabilities get counted as “high-powered money” by economists. And some of its liabilities pay no interest. But most of the tremendous growth in the Fed’s balance sheet since 2008 is financed by interest-bearing liabilities. This is not your 1970s textbook Fed any more.

What I think matters for inflation these days is the total debt issued by the Treasury. The modern financial markets turn government debt into the fuel for inflation that we used to associate with money.

But inflation is not a simple mechanical process. It does not follow a simple ratio or other formula.

I think that inflation has two regimes. In one regime inflation is low and stable. In another regime, inflation is high and variable. In the 1970s, eliminating the dollar’s gold peg resulted in a transition to the high and variable regime. Getting out of that regime took a steep recession and high interest rates at a time when most consumer bank deposits had regulatory caps on interest. Back then, financial markets were not nearly as well integrated as they are now, and the Fed probably mattered more as a result.

More recently, the COVID policy to expand demand (deficits) while restricting supply (lockdowns) resulted in another transition to the high and variable inflation regime. Maybe this time the transition back to low and stable inflation will be less painful. At this point, it is too early to tell. But I think that with financial markets highly integrated, the Fed cannot by itself drive out inflation.