Chairman Smith, second try

I got distracted yesterday and ended up at my favorite topic -- misunderstanding the economy. I'm back now, trying to focus on Chairman Jason Smith's statement on the Biden inflation. I have to begin by restating something I said, wow, almost a month ago now.

On the graph below, blue shows "Headline" CPI inflation and red shows "Core" CPI inflation. The two follow roughly the same path, but the blue line shows a lot more up-and-down motion than the red line does. The blue line, in other words, is more volatile.

Headline and Core inflation have their differences. But often when Headline inflation is rising, Core is rising also. And often when Headline is falling, so is Core. Sometimes the two follow the same path, but blue is much more jiggy than red. The jigginess is volatility. And then sometimes, Headline inflation shows a substantial drop or substantial increase, or both in turn, while Core makes nothing but small steps. Again, there is less volatility in Core than in Headline.

The red line is less "jiggy" because Core PCE is less "volatile". Core PCE is less volatile because they leave out the most volatile data -- food and energy. The stats people want to take some of the noise out of the numbers -- or at least that's what I would want to do -- so they omit that volatile data.

Graph #1:  Headline (blue) and Core (red) CPI Inflation since 1980

Before I made the graph and actually looked at it, I thought that Core PCE understated inflation. It does not. It understates changes in inflation. Usually, with Core PCE, the highs are less high and the lows are less low. Usually, but not infallibly, Core PCE smooths out the data. It's all there on the graph. All I had to do was look.

When inflation is in the news, it is because inflation is increasing. When inflation goes high, if we compare Core inflation to Headline on a graph, Core will often appear to understate the inflation, because the highs of Core inflation are less high than those of Headline inflation.

When inflation goes low, Headline inflation (the more volatile one) goes down rapidly. Core inflation (the less volatile one) goes down more slowly. At such times, Core inflation appears to overstate the inflation, because it understates the decline. 

As a rule, Core inflation understates the changes. The highs are less high, and the lows are less low.

//

Chairman Smith's statement ends with half a dozen bullet points under the heading "Key Background". The last of these bullet points is

• Inflation has become so deeply ingrained in the economy that when removing food and energy this month’s core inflation number of 5.3 percent is even higher than the topline figure.

I want to restate that, for clarity:

This month’s core inflation number of 5.3 percent is even higher than the headline number.

I am left with several considerations:

1. Smith mentions "removing food and energy" from the inflation data. This removal is done by the people who provide the stats. Food and energy prices are seen as more "volatile" that other prices. Removing these more volatile prices from the calculation leaves the less volatile "Core" data that the Chairman mentions. So the jiggies of Core inflation are smaller than the jiggies of Headline inflation.

2. I figure what Smith calls a "topline" number is the one that I've seen called the "headline" number. "Topline" sounds like a decent synonym. To avoid confusion, I prefer not to use synonyms. Except, of course, "jiggies" for "volatile".

3. I don't know what the Chairman means by "deeply ingrained" inflation. I suppose he means that inflation expectations are no longer anchored at 2 percent (as Jerome Powell might put it) and that people expect higher inflation. I don't know if the statement is true or not (and I won't guess), but it is how I interpret Chairman Smith's words.

4. Smith's statement doesn't say whether he's looking at PCE inflation or CPI. But the statement is dated 15 June 2023. By that date the May data was probably available for the Chairman's use. A FRED search for headline inflation turns up 6 datasets, including Core PCE and Core CPI:

• Core PCE (FRED PCEPILFE) for May 2023 is given as 4.68783 percent. That rounds to 4.7 percent, which does not match Smith's 5.3 percent.
• Core CPI (FRED CPILFESL) for May 2023 is given as 5.33225 percent, or 5.3 percent when rounded. This is the data Chairman Smith was looking at.

5. Smith says Core CPI was "even higher than" Headline CPI for May, 2023. Headline CPI (FRED CPIAUCSL) for that month was 4.12069 percent, which rounds to 4.1 percent. Core CPI for that month, at 5.3 percent, was certainly higher than Headline CPI. Chairman Smith is right about this. And I get a Milk-Bone for my painstaking effort here. 

You can have half, for reading these tedious notes.

//

Now that I know what Chairman Smith is saying, and the data he used, I can evaluate what he said. As I read him, he says

Inflation has become so deeply ingrained in the economy that the core inflation number is even higher than the headline figure.

His statement strongly suggests that in his view Core inflation is almost never higher than Headline inflation. That's what I used to think. As I said above: Before I made the graph and actually looked at it, I thought Core PCE understated inflation. If Core understated inflation, it would be unusual for Core to be higher than Headline. But Core doesn't understate inflation. It understates changes in inflation. 

It is not unusual for Core inflation to go below Headline. Core is below Headline when it understates rising inflation. 

It is not unusual for Core inflation to go above Headline. Core is above Headline when it understates falling inflation. It's all right there on the graph.

It is likely that we are most concerned about inflation when inflation is rising. So it is probably true that usually, when we look at inflation on a graph, we see Core understating the increase. But if we should look at a spot on the graph where inflation is falling, or falling rapidly, we would very likely see Core inflation understating the decline. We would very likely see Core inflation higher than Headline inflation because Core understates the decline.

I got the Excel data for Graph #1 above, and subtracted the Headline values from the Core values. I put the results on a bar graph. When Core is greater than Headline, the result is above zero. When Core is less than Headline, the result is below zero:

Graph #2: Core CPI is Greater than Headline CPI when the bars are Above Zero

This graph shows 534 months of data, from January 1980 to June 2024. 295 of those months are above zero (Core is greater than Headline). 239 of those months are below zero (Core is less than Headline). For this sample, Core inflation is more often above Headline inflation than below it. It happens during disinflation, when the rate of inflation is falling. Core goes above Headline when it understates falling inflation. It happens often.

I only figured this out a month ago. But I'm a hobbyist, studying the economy for my own satisfaction. It troubles me that Chairman Jason Smith of the House Ways and Means Committee doesn't know it.

On Graph #2, most of the 1980s show the blue bars above zero, meaning Core inflation was higher than Headline inflation. This was during the Volcker disinflation. Headline inflation came down rapidly. Core understated that decline every month from July 1981 to July 1987.

And again, every month from March 2023 to end-of-data show the blue bars above zero, meaning Core inflation was understating the decline of Headline CPI. I guess we'll have to call this the Biden disinflation.

Core inflation understates changes in Headline inflation. It's a pretty good rule. I just wish the Chairman of the House Ways and Means Committee understood it.

//

The Core inflation Smith mentions for May 2023 was higher than the Headline number. Chairman Smith seemed to think this most unusual. But it is often true that Core inflation is higher than Headline inflation.

Looking at the May 2023 data, Chairman Smith said Core inflation was higher than Headline inflation because inflation "has become so deeply ingrained" in us. Maybe he meant we expect inflation to be higher than 2 percent. Or maybe he just meant we expect inflation, always.

We do expect inflation, always, because we always have inflation. Granted, we don't always have 9 percent inflation. Usually it is much lower. But almost always, we have inflation.

Core inflation was higher than Headline in May 2023 -- in fact, higher in every month from March 2023 to the most recent data, June 2024 -- but not because inflation is "deeply ingrained". Core inflation has been running higher than Headline inflation because Headline inflation has been coming down. Headline inflation has been coming down, and Core has been understating that change. So Core inflation  has been higher than Headline inflation. It is not unusual, and it is not because inflation is "so deeply ingrained."

This is the Biden disinflation, remember. Inflation is coming down. and Core understates that change. So Core inflation has been running higher than Headline inflation. Just like the Volcker disinflation.