“It’s Not Straightforward”: An Atlanta Fed Economist Discusses Measuring Inflation

February 23, 2021

southeastern United States stylized as a colonial-era watercolor map

Few economic concepts create more confusion than inflation. Even within the economics field, definitions of inflation have evolved over the decades.

Nonetheless, inflation matters a great deal to the Federal Reserve. As part of its dual mandate, the nation’s central bank is charged with maintaining low and stable prices. Decoding exactly where inflation stands has proven uncommonly difficult during the pandemic, however, as dramatic swings in consumer behavior have scrambled prices.

To discuss this phenomenon and inflation generally, Economy Matters staff writer Charles Davidson spoke with Atlanta Fed economist and resident inflation expert Brent Meyer.

Charles Davidson: Atlanta Fed president Raphael Bostic often emphasizes that what’s really important is the trajectory of inflation, as opposed to the current reading or month-to-month changes. Why is the trajectory critical?

Brent Meyer: Focusing on trajectory is his way of clueing in the public about when he would be uncomfortable with rising inflation. In the long run, we’re trying to hit 2 percent on average. But if we’re running at 2.2 percent, let’s say, and we’ve gotten there gradually, he’s not going to see that as an imminent danger to price stability. But if we get to 2.2 percent very rapidly, and it looks like it’s going to continue to rise, then he’s going to be uncomfortable.

CD: And is trajectory synonymous with the notion of “underlying inflation,” which is what the Fed aims to control?

BM: They are not exactly synonymous. Underlying inflation measures—there’s a basket of them that we track—attempt to disentangle relative price changes [of individual products or services] from an inflation signal that we care about (see the chart). That signal is telling us whether monetary policy is well positioned to the current economic fundamentals.

Underlying Inflation Dashboard

12-Month Growth Rate (Percent)

Measure of Underlying Inflation

January 2020

January 2021

Core Consumer Price Index (CPI)



Cleveland Fed Median CPI



Cleveland Fed 16% Trimmed-Mean CPI



Atlanta Fed Sticky CPI



Core Personal Consumption Expenditures (PCE) Index 



Market-Based Core PCE



Dallas Fed Trimmed-Mean PCE



San Francisco Fed Cyclical Core PCE Inflation



Cyclically Sensitive Inflation (Stock and Watson [2019])



Sources: U.S. Bureau of Labor Statistics; U.S. Bureau of Economic Analysis; Federal Reserve Banks of Atlanta, Cleveland, Dallas, and San Francisco; Stock and Watson (2019); staff calculations 


CD: The layperson often thinks of inflation as measuring the cost of living, but it doesn’t. Why is underlying inflation, as opposed to the cost of living, so critical to what we want to understand?

BM: There are two things going on. What we’re trying to measure with underlying inflation is not necessarily related to the cost of living. They tend over long periods to be related, sure. But what we’re trying to do when we’re measuring inflation is to understand how the stance of monetary policy is relative to what it needs to be. If monetary policy is too loose, there’s too much money floating around in the economy, then prices over a long period of time will increase. That’s not a great situation for firms and households wanting to make long-range plans.

Changes in the cost of living are very much a real phenomenon. If it goes up, and my wages don’t, then I’m going to feel poorer. So, there’s not only that disconnect, but the other disconnect is this: the consumer price index and the personal consumption expenditures index are trying to gather the price changes for a representative household, an archetypal household, and this archetypal household does a lot of weird things.

It smokes, but just a little. It buys a small amount of infant and toddler apparel and equipment whether or not it has kids. It’s constantly paying legal fees. The point is that the experience of that “typical household” may not mirror your spending patterns at any point in time over the year. But on average, these things do add up.

photo of Melinda Pitts
The Atlanta Fed’s Brent Meyer. Photo by David Fine

CD: Someone is always paying legal fees, though you may not be. Lots of people are always buying infant stuff, though you may not be, right?

BM: These indexes are not really trying to mirror your cost of living changes directly. The indexes are trying to get at basically the average for the overall economy.

CD: You’ve written, and Raphael has said, that inflation is getting harder to measure; that signal is getting harder to find. Why is that, and can we devise better tools to try to gauge it?

BM: Recently, the primary reason is because of the preponderance of these relative, or individual, price changes due to the onset of the pandemic. Airfares were plunging early in the pandemic because no one was flying, for instance. Some areas are experiencing dramatic price increases due to COVID and social distancing and such: food at home, used cars, home goods, appliances, furnishings. They’re just huge. The volatility of the underlying market basket [from which inflation is calculated] has just exploded.

CD: Prices of certain individual goods and services have been jumping all over the place?

BM: Imagine you measure all the prices and sort them highest to lowest, and there’s a distribution of price changes. Let’s say the normal distribution of price changes has a variance of 1—volatile but not super-volatile. Now that’s a made-up number. At the onset of the coronavirus hitting the price system, April and May, that variance went from 1 to, like, 5. It was the highest variance that we have on record.

CD: Is this mainly a result of these extraordinary changes in consumer behavior that happened virtually overnight?

BM: Yes. Again, you’re trying to separate the fundamental inflation signal from the relative price changes. This makes it really challenging. In fact, something that’s going to happen in April is because we saw this sharp decline in April and May of 2020, we’re going to see the year-over-year growth rate in inflation spike above 2 percent. So, the PCE [personal consumption expenditures] core figure is likely going to be at 2.3, 2.4 percent. It doesn’t mean underlying inflation’s running that high, though. It’s another distortion that comes from the size of the shock we had.

CD: Some of the changes in behavior could last. We may do more online shopping. We may entertain more at home and eat out less. Is there a way to know how much of this is going to persist beyond the pandemic? And if so, then, how do we get our arms around this as far as picking up that real inflation signal?

BM: It’s not straightforward to figure these things out. We need to infer them from current research, our survey work, REIN [the Atlanta Fed’s Regional Economic Information Network] talking to firms on how they’re responding to conditions, talking to households.

Firms in our Survey of Business Uncertainty said they were slashing business travel markedly. That’s going to fundamentally change the makeup of the economy—this reallocation away from travel and industries that support business travel and toward industries that support housing, retrofitting housing for offices, and such.

CD: I wanted to ask a little about how the inflation data are gathered, the processes. Can you talk about how the Bureau of Labor Statistics [BLS] has changed the data gathering?

BM: The BLS suspended in-person price checking, basically, in March and resorted to collecting prices online and through phone surveys. It’s not clear that the quality is different, but overall collection rates are down a bit.

CD: They normally send people out to stores and walk up and down the aisles checking prices?

BM: Oh, yeah. I chatted with one of the agents collecting prices once. They go to random places. They went to a manufacturing plant to just make sure the prices in the vending machine were the same. And they’ll say, “From this grocery store we need the prices of vinegar, eggs, and peanut butter.” And it’s not just saying, “The price of Jif was $2.99 last month and it’s still $2.99.” They make sure it’s the same price, it’s not on sale, the packaging isn’t bigger.

Three quarters of these price checks were personal visits in 2019, and now virtually none of it is personal visits. That’s been replaced by internet price scraping and phone surveys. But we have some sense that the prices the BLS is getting are OK. So that doesn’t appear to be a huge issue. The bigger issue is the imputations. For instance, almost nobody has leased a car in the United States in months. The amount has fallen so low that the BLS is unable to have any confidence in that price index, so it hasn’t published it. We don’t know what’s happening. The other is admission to sporting events. We have no idea what’s happening to those prices because no one’s going to sporting events. Again, that’s mitigated by the fact no one’s going, but it does impart noise into the inflation measurement.

CD: Is it possible to put this period in inflation into a historic context?

BM: Here in the U.S., I’m not sure I can relate what we’re seeing now to anything in history given the disparate and dispersed nature of the shock. It’s not a great time to be digging into price data. It’s just not very useful at the moment. It’s going to be July, I think, before I have any confidence in what the underlying inflation measurements are telling us.

photo of Charles Davidson
Charles Davidson

Staff writer for Economy Matters

The Atlanta Fed moderates all comments posted by visitors to this website. We reserve the right to delete content that is abusive, harassing, or threatening; or obscene, vulgar, or profane. In addition, no off-topic remarks or spam is permitted. Learn more at our Disclaimer & Terms of Use and Online Privacy Policy.