FDR's immaculate inflation

David Glasner has an excellent post on the relationship between inflation and unemployment. At one point he links to Paul Krugman, who makes the following comment:

Even if you think that inflation is fundamentally a monetary phenomenon (which you shouldn’t, as I’ll explain in a minute), wage- and price-setters don’t care about money demand; they care about their own ability or lack thereof to charge more, which has to – has to – involve the amount of slack in the economy. As Karl Smith pointed out a decade ago, the doctrine of immaculate inflation, in which money translates directly into inflation – a doctrine that was invoked to predict inflationary consequences from Fed easing despite a depressed economy – makes no sense.

David has lots of interesting things to say about Krugman’s argument, but I’d like to make a few additional observations.

1. Inflation does not require an elimination of “slack”. Here Krugman is confusing microeconomics with macroeconomics. Microeconomics is about relative prices, which are determined by conditions in individual markets. Macro is about nominal prices, which are determined by the supply and demand for base money. Let’s take FDR’s policy of devaluing the dollar, which occurred during 1933-34, a time of 25% unemployment. I’d call that “slack”, wouldn’t you? As you can see, the policy led to a high rate of inflation:

This policy increased the expected future rate of inflation, which reduced the current demand for base money. The very high unemployment did not prevent inflation from rising sharply, because inflation is a monetary phenomenon.

Back to Krugman:

And to get back to my broader point: economics is about what people do, and stories about macrobehavior should always include an explanation of the micromotives that make people change what they do. This isn’t the same thing as saying that we must have “microfoundations” in the sense that everyone is maximizing; often people don’t, and a lot of sensible economics involves just accepting some limits to maximization. But incentives and motives are still key.

You don’t need microfoundations such as overheating product/labor markets to get inflation; you simply need to reduce the value of the medium of account. Suppose Mexico replaces 1000 old pesos with one new peso, in a currency reform. What motivates Mexican businesses to cut all prices by 99.9%? Is it “slack” in the economy? Obviously not. It’s enough to assume that Mexicans are rational, and know that the currency reform will reduce prices by 99.9% in the long run. Similarly, back in 1933, Americans understood the implications of dollar devaluation.

At the same time, Krugman is right that inflation and unemployment are often negatively correlated in the short run. David points out, however, that Krugman is confusing a correlation with a causal relationship. Consider the following hypothesis:

1. Inflation is always and everywhere caused by monetary (supply and demand) factors.

2. Changes in inflation are often negatively correlated with unemployment, due to sticky wages.

These two hypotheses are consistent with the Phillips Curve holding during some periods (such as 1932, 1969, 1982 and 2009, and not holding during other periods (1933-34, 1974, 1980-81, 1998-2000, etc.) But these two hypotheses do not imply that unemployment has any causal impact on inflation.

If we assume that monetary factors are the underlying cause of inflation, and the Phillips Curve relationship is contingent on a set of special factors, then we can explain the history of American inflation. If we assume that unemployment (i.e. “slack”) is the causal factor that determines the rate of inflation, then we are left with all sorts of puzzles.

Back to Krugman:

Consider, for example, the case of Spain. Inflation in Spain is definitely not driven by monetary factors, since Spain hasn’t even had its own money since it joined the euro. Nonetheless, there have been big moves in both Spanish inflation and Spanish unemployment:

The fact that Spain lacks its own money has no bearing on whether Spanish inflation is caused by monetary factors. The typical state in American (let’s say Indiana), did not have its own money in 1966-81, but nonetheless suffered from high inflation due mostly to “monetary factors” (i.e., excessively expansionary Fed policy.) That’s not to say that real factors don’t also matter—presumably they explain why Spain’s inflation rate might differ from a neighboring Eurozone member—but that doesn’t mean that both Spanish and Eurozone inflation are not largely monetary.

I would never deny that slack might have played some role in the inflation process of 1933-34. Thus the same policy of dollar depreciation might have produced even higher rates of inflation had the unemployment rate been 3%, rather than 25%. But you can certainly get plenty of inflation in a depressed economy, as we saw in 1933-34.

PS. We lack good data on inflation from the 1930s, which is why I use the WPI. Broader indices have been estimated, and they show lower rates of inflation. But even so, the inflation rate by any measure was far higher than predicted by Phillips Curve models.