You might have seen that the inflation rate rose from the 6.8 percent reported in early December to 7 percent reported today. It’s important to understand what exactly is being reported. The reporters don’t necessarily do a terrible job, but readers often misunderstand.
What is being measured in each case is the inflation rate over a 12-month period. Inflation between November 2020 and November 2021, which is what was reported in December, was 6.8 percent. Inflation between December 2020 and December 2021, which is what was reported this morning, was 7.0 percent.
So it looks as if inflation is rising. Is it? No. The reason is that to know if inflation is rising or falling, we need to look at the latest data. So we need to compare the inflation rate in December with the inflation rate in November. In November, the Consumer Price Index rose by 0.8 percent, which implies an annualized inflation rate of 9.6 percent. But in December the CPI rose by 0.5 percent, which implies an annualized inflation rate of 6.0 percent. So inflation actually fell.
One of the most careless statements I’ve seen on-line, and it’s in the headline of the news story referenced above, is that inflation rose by 7 percent. No. Prices over the last year rose by 7 percent. If you want to stick with annual, then to be correct, you would say that the inflation rate rose by 0.2 percentage points. If you want to report as accurately as possible what happened to inflation this last month, you would say that the monthly inflation rate fell by 0.3 percentage points and that the annualized inflation rate fell by 3.6 percentage points.
One of the most important things I learned from Ben Klein in his Ph.D. monetary theory course at UCLA was always to distinguish between rates and levels. Levels tell us that prices rose. Rates tell us that the inflation rate fell.
UPDATE: Another wrong headline.
READER COMMENTS
Mark Brady
Jan 12 2022 at 10:12pm
“In November, the Consumer Price Index rose by 0.8 percent, which implies an annualized inflation rate of 9.6 percent. But in December the CPI rose by 0.5 percent, which implies an annualized inflation rate of 6.0 percent. So inflation actually fell.”
Are the numbers that you cite for the rise in the Consumer Price Index from month to month (between October and November and between November and December) seasonally adjusted or not?
David Henderson
Jan 12 2022 at 10:34pm
Seasonally adjusted.
Mark Brady
Jan 13 2022 at 12:35am
Thank you, David.
And “if you want to report as accurately as possible what happened … this last month,” “prices” didn’t rise, a weighted average of the prices of a basket of goods rose.
I’m pretty sure Ben Stein would have made that point, and I’m pretty sure you would have said the same thing in a lecture to your students. 🙂
David Henderson
Jan 13 2022 at 9:51am
Correct about me. Why are you so sure about Ben Stein?
Mark Brady
Jan 13 2022 at 11:13am
Based on your enthusiasm for Ben Stein, I assumed that he would at least sometimes recognize that “prices” are a reference to a weighted average of the prices of a basket of goods. 🙂
David Henderson
Jan 13 2022 at 9:39pm
I’m enthusiastic about Ben Stein?
Chris Hutchison
Jan 13 2022 at 11:26pm
I think he meant Ben Klein…not Stein.
Mark Brady
Jan 23 2022 at 2:55pm
Yes, I meant Ben Klein. Thank you, Chris Hutchison.
Philip George
Jan 12 2022 at 11:15pm
Inflation rate rose by 0.2 percentage points.
Chris Hutchison
Jan 13 2022 at 12:36am
How does 0.2% monthly correspond to 7% annually?
Johnson85
Jan 13 2022 at 12:22pm
The annual inflation rate measured from December 2020 through November of 2021 was 6.8%. The annual inflation rate measured from January 2021 through December 2021 was 7.0%. So the annual inflation rate is .2% higher compared to the prior month.
The annual inflation rate went up even though the inflation from Dec. 1 2021 through Dec. 31, 2021 was lower than the inflation from November 1, 2021 through November 30, 2021. Presumably, the monthly inflation in December 2020 was lower than that in December of 2021, so the annual inflation rate rose even though it was dropping if you looked at it on a monthly basis. That nomenclature might not be exactly precise/correct, but that is the way I think of it.
I track my monthly spending so this is easy to see. Often times, my monthly spending will go up but my running 12 month average will go down, because my monthly spending, while higher than the prior month, was still lower than the same month the preceding year that is falling out of the rolling 12 month average.
Chris Hutchison
Jan 13 2022 at 11:58pm
Thanks for clearing that up! I got confused with the statement “if you want to stick with annual”…and that there are 2 ‘annuals’ we are dealing with here. 1. The average annual rate, and 2. The annualized rate.
Alan Goldhammer
Jan 13 2022 at 7:51am
Best post of the new year!!! Innumeracy is still a plague to be dealt with.
David Henderson
Jan 13 2022 at 9:51am
Thanks, Alan.
Daniel R. Grayson
Jan 13 2022 at 11:55am
From today’s New York Times: “Inflation climbed to its highest level in 40 years at the end of 2021.” (January 13)
Bob Murphy
Jan 13 2022 at 5:37pm
Daniel Grayson wrote: “From today’s New York Times: “Inflation climbed to its highest level in 40 years at the end of 2021.” (January 13)”
I presume, Daniel, that you’re showing David this, as another example of a false headline. But I don’t think so. They have in mind that they are measuring inflation as a % rise in prices over a 12-month period, and by that measure, their statement is correct.
I realize the point David was making in his post, but I think if we take that logic to its full conclusion, we would have to say that even with the monthly figures, we technically have no idea if inflation is rising or falling. After all, maybe the reason the December 2021 average is what it is, is that the first 3 weeks saw steady prices while the last week saw a swift uptick, such that the December average was 0.5% higher than the November average. In that case, the “most recent data” would show a very high annualized inflation rate.
In the extreme, we would have to report the instantaneous rate of inflation as the delta-t goes to zero. But David was comfortable using the monthly average as his unit, when he said “inflation actually fell.”
David Henderson
Jan 13 2022 at 9:41pm
Good points.
Chris Hutchison
Jan 13 2022 at 11:51pm
Yes, very true. In my opinion (I’m not an economist) the use of an annualized rate based on the current month’s data is a silly thing. It just exaggerates whatever is going on, but doesn’t really help. I also think it is useful to look back at what has happened in the past year (6.8% vs 7.0%). I think the best way to present the data is to say, “inflation rose by 0.2% to 7.0% over the past year, and rose 0.3% over the past month.” No need for annualized rates.
Chris Hutchison
Jan 14 2022 at 12:00am
…oops…meant *fell 0.3% over the past month
Phil
Jan 14 2022 at 6:29am
David:
If the CPI grows 0.8% per month for 12 months, isn’t the annual inflation rate 10.0%?
(1+0.8%)∧12 = 1.100
David Henderson
Jan 14 2022 at 9:43am
I think you meant (1 + .008)^ = 1.10.
And you’re right about 10%.
I just assume that the data are not so precise that we can assume that the BLS knows that prices rose by exactly 0.8 percent last month, as opposed to, say, 0.77 percent or 0.83 percent. If the former, then the annualized rate is 9.6 percent.
Comments are closed.