Mo’ Money, Mo’ Macro Problems?

Money and Inflation

It’s one thing to count our money, as Macrosight discussed in last week’s post. It’s another thing to understand how money matters for the macroeconomy. While on the one hand it is pretty obvious—we use money to buy things—on the other the hand there is more than meets the eye when it comes to understanding how money does, and does not matter. Let’s start with the latter.

Money Doesn’t Matter

Money doesn’t matter? Isn’t that something rich people say to sound “deep”? Maybe. But when it comes to the macroeconomy, in the long run, it is true, money does not matter. That idea is known as the “long run neutrality of money.”

Long-run “neutrality” means that the amount of money in a macroeconomy does not matter for long-run economic growth. M2 could be $25 trillion or $50 trillion. The stock of money does not cause economic growth. If anything, the stock of money is the result of economic growth. This idea comes from an idea known as the “Quantity Theory of Money,” or QTM, for short.

The QTM tells us that the only thing money causes in the long run is inflation. As the stock of money grows, so does the price level (say, as measured by the CPI or the PCEPI). The connection between the two—the increase in the money supply and the increase in the price level—is not always obvious—but in the long-long run, the two variables appear to share a common trend. With that in mind, Figure 1 displays the Consumer Price Index and the stock of M2 since 1959 (at the monthly frequency).

Over time both series go up and up—hence a common upward trend. However, the upward path of each is not completely synchronous. Does that mean the QTM is wrong? Not necessarily.

Money Growth maybe matters

A key implication of the QTM is the rate of change of the money supply will equal (in the long run) the rate of change in the price level. Figure 2, for example, displays the rate of inflation—measured as a twelve-month rolling average—and the rate of M2 money growth, defined in the same way, from January 2000 through September 2024.¹ (Shaded regions indicate recessions.)

From eye-balling the figure, it is difficult to say with certainty if there is a common pattern between two (and even if we did, that would not be enough to say that a change in one variable caused the other variable to change). It would be easier to draw such a conclusion if we focused our eyeballs on the latter part of the graph, where we see the large spike in M2 starting in 2020. In that case, you might then think that the big M2 spike preceded the spike in the rate of inflation that started in 2022.

But, if you looked at other periods on the graph, that narrative is less-compelling. At least there does not appear to be a clear pattern of money-going-up followed by prices-going-up. For example, during the Great Recession, the growth rate of M2 increases, yet, the rate of inflation declines. As M2 declines, the rate of inflation increases. But perhaps this 24-year snap-shot is not long enough to see what the QTM says we should see.

Correlation (especially low-correlation) is not causation

As another check on the relationship between money growth and the rate of inflation, I calculated the correlation between the series from 1959 to September 2024, and for sub-periods in between. Then I did so to see if there was correlation between rates of money growth from the previous few months on the current value of monthly inflation. In all cases, the estimated correlations were very low.² In other words, I did not find much statistical evidence that M2 growth helped predict or “leads” the rate of inflation.

Why does this matter? If the QTM-predicted link between money growth and inflation were true, that would make monetary policy relatively easy to implement. You could figure out a way to set the growth rate of our money supply to, say, 2 percent, and the rate of inflation would consistently equal 2 percent.

Alas, however, the relationship is not reliable. As such, money growth is not a reliable guide to setting monetary policy. Instead, we as the macroeconomic public rely on the Fed. And they, for better or worse, rely on their judgement.

1. I focus on a shorter period here than in Figure 1 since the growth rate of M2 at the monthly frequency, in particular, is pretty volatile, rendering any eyeball examination of the series from 1959 to today more challenging. For that reason, too, I plot the rolling averages—to better see any possible patterns that emerge.

2. For those econometrically-inclined, I also ran various regressions models and vector autoregressions. In all cases, there was very little evidence of the growth rate of money leading the rate of inflation, either in terms of the magnitudes on the estimated coefficients or in terms of statistical significance of those coefficients.