The Great Inflation Mystery

The people who set interest rates don’t know what causes inflation, how to measure it, or how to move it up and down.

Illustration: Tomi Um for Bloomberg Businessweek

The first principle of war is “know thine enemy.” But as Federal Reserve Chairman Jerome Powell and his colleagues raise interest rates to keep the U.S. economy from overheating, there’s a lot they don’t know about the foe they’re trying to contain: inflation.

Here are some of the things about inflation the Fed and other central banks are uncertain of: what causes it; what effects it has; what to count in measuring it (stock prices?); how low, or high, it should be; and how to move it up and down.

The Fed, in other words, is driving blind. Daniel Tarullo, in a tell-all address at the Brookings Institution in October after his resignation from the Fed’s Board of Governors, said, “We do not, at present, have a theory of inflation dynamics that works sufficiently well to be of use for the business of real-time monetary policymaking.”

On March 21, the Federal Open Market Committee raised its target range for the federal funds rate to 1.5 percent to 1.75 percent. In his first press conference as chairman, Powell said, “There’s no sense in the data that we’re on the cusp of an acceleration of inflation.”

Like a cook lowering the flame under a pot, the Fed is trying to reduce the pace of growth from a boil to a simmer to extend the life of the almost nine-year expansion. History shows that once the central bank begins to raise interest rates, it often goes too far. Five of the seven credit-tightening cycles since 1970 have choked growth and ended in a recession. But without understanding more about inflation, it’s hard to know if the Fed is tightening too quickly, not quickly enough, or at about the right pace.

The chart above illustrates one problem facing the central bank: Prices don’t rise or fall in unison. What the Fed calls “inflation” is just a weighted average of the ups and downs of the prices of all goods and services in a basket that reflects the spending of all American consumers. And those prices change for a variety of reasons, including technology, consumer preferences, and the cost of imports.

How people experience inflation varies. It was higher for the old, for the poor, and for large families than for the rest of the population from 2004 to 2013, according to research into half a billion transactions by economists Greg Kaplan of the University of Chicago and Sam Schulhofer-Wohl of the Federal Reserve Bank of Chicago.

The sales channel matters, too. Inflation of goods sold online ran 1.3 percentage points lower than the Consumer Price Index for the same product categories from 2014 to 2017, according to research by Austan Goolsbee of the University of Chicago Booth School of Business and Pete Klenow of Stanford, who analyzed data from the Adobe Experience Cloud, which includes 80 percent of the transactions of the top 100 U.S. retailers.

Another big divergence is between goods and services. Since the end of the last recession, the Fed’s favorite measure of inflation—the change in the price index for personal consumption expenditures—has averaged 1.5 percent a year. But when you break the index into its two main components, services inflation has averaged 2.2 percent annually while goods inflation has come in at just 0.3 percent. The problem for the Fed is that monetary policy is a blunt instrument that can’t treat the services side and the goods side of the economy differently. “The average of landing at two airports is called a crash,” says Brian Barnier, an expert in operations research who’s head of analytics at ValueBridge Advisors LLC in New York.

The problem with inflation theories is that they tend to rest on “unobservables,” such as the concept of the natural rate of unemployment. “Attempts to estimate them have strongly suggested that they aren’t constants,” writes Edward Yardeni, the Wall Street economist, in a new book, Predicting the Markets: A Professional Autobiography.

The Fed’s challenge with inflation isn’t just a lack of knowledge; it’s also a lack of power. The economist Milton Friedman asserted in 1963 that “inflation is always and everywhere a monetary phenomenon,” implying that the Fed could exert near-total control over it by adjusting the supply of base money. He theorized that prices couldn’t possibly rise unless the Fed pumped more money into the economy. But he was wrong. Even if the Fed does nothing, the opportunity for inflation can increase if banks make more loans and if money circulates through the economy faster.

There’s even an argument that raising interest rates now, far from damping price pressures, could stoke them. The idea goes back to economist Irving Fisher in the 1930s, based on the simple equation that the nominal rate of interest equals the real rate plus inflation. For example, if the nominal rate is 2 percent and inflation is zero, the real interest rate is also 2 percent. If the Fed raises the nominal rate to 4 percent but the real interest rate stays at 2 percent, then inflation will have to rise to 2 percent to make up the difference: 2 + 2 = 4. Federal Reserve Bank of St. Louis President James Bullard said in a presentation in Germany in 2016 that “Neo-Fisherian ideas may have an important impact on our thinking about monetary policy in the future.”

Many of the people asking the sharpest questions about the Fed’s thinking on inflation aren’t anointed doctors of economics. An extreme example of the self-taught econ skeptic is Matt Busigin, who dropped out of high school at 15 but has managed to have a dual career as a software engineer in Buffalo and a portfolio manager for Los Angeles-based New River Investments Inc. Running tests on 20 years’ worth of data, he found a negative correlation between increases in the Fed’s base money and future inflation: When one rises, the other falls, and vice versa. Theory predicts a positive correlation.

Closer to the center of power is Patrick Harker, president of the Federal Reserve Bank of Philadelphia, an engineer by training with expertise in operations research, a scientific method for management of systems ranging from transportation networks to factory floors. Operations researchers tend to value data collection over theory-spinning. Thinking out of the box, Harker’s bank is even toying with using the emerging science of machine learning to make macroeconomic predictions and policy. It sponsored a conference on the topic last year and hired an expert in the field to start work in August.

What can the Fed and other central banks do, given the uncertainty? A good start is to admit the problem exists. Powell’s predecessor, Janet Yellen, kept the Fed “data-dependent”—dragging her heels on raising rates until she saw evidence of inflation, not just theories predicting it. Said Yellen last year in New York: “You have to keep an open mind and not assume you have a monopoly of truth.”

    BOTTOM LINE - The Fed has a preferred way of measuring inflation, but it’s an open question whether the institution has the right tools to act on prices—and hence growth.
    Before it's here, it's on the Bloomberg Terminal.