January 21, 2022
The news media have been constantly hyping inflation in recent months. While everyone has been seeing the huge rise in gas prices over the last year (that’s what happens when the world reopens after a pandemic), used car prices have risen almost as rapidly. From December 2020 to December 2021 they rose 37.3 percent. This accounted for 1.03 percentage points of the 7.0 percent overall inflation in the last year.
We know the story of these price increases. A fire in a semiconductor plant in Japan has created a worldwide shortage of semiconductors, which has slowed car production. With people unable to get new cars, they are bidding up the price of used cars.
But beyond the specifics, there is an interesting accounting issue (oxymoron?) here. The Consumer Price Index (CPI), which is our most used measure of inflation, uses a different methodology for used cars than the Personal Consumption Expenditure (PCE) deflator calculated by the Commerce Department.
The CPI counts the full value that consumers pay for a car in determining its weight in the index. In December, this weight was 3.42 percent. By contrast, the PCE uses a net measure that subtracts out what consumers are paid for the used cars they sell. In the PCE, used cars had a weight of 1.65 percent for November, the most recent month available. This means that the weight of used cars is approximately 1.8 percentage points higher in the CPI than in the PCE.
Typically, this 1.8 percentage point difference would not matter much, but when the price of used cars and trucks is going up 37.3 percent in a year, it matters. In contrast to the 1.03 percentage points that used vehicles contributed to the CPI over the last year, they contributed just 0.62 percentage points to the inflation rate in the PCE. This means that if we were measuring inflation in the CPI using the net methodology of the PCE, it would be roughly 0.4 percentage points lower over the last year. (There are other differences in methodology that make this calculation a bit more complicated.)
Of course, even subtracting 0.4 percentage points still leaves us with a 6.6 percent year-over-year inflation rate, which is high by anyone’s standard, but this gap does make a difference in how we see the world. For example, the average hourly wage for production and nonsupervisory workers rose 5.8 percent over the last year. Measured against the 7.0 percent inflation rate, this implies a 1.2 percentage point decline in real wages. (They rose 4.0 percent in the prior year, this is a pandemic-composition story.) However, measured against a CPI that uses the PCE deflator net measure, the decline was just 0.8 percentage points.
Declining real wages are still bad news, and would be especially bad if we expected these declines to persist for any period of time, but using an alternative and reasonable measure for used vehicle prices eliminates one-third of the drop over the last year. That seems worth noting.
 Prices for used vehicles in the PCE deflator actually rose slightly faster over the last year, going up 42.5 percent.