Over on Twitter (oops! X) @DisruptorStocks noticed something interesting. His observation triggered a bit of controversy about the October CPI.
From October 2023 to October 2024 the health insurance component of the CPI fell by 34 percent. Here’s the data and a graph. The Excel workbook is here.
For those who want to investigate further, the Series ID is CUUR0000SEME. Trust me, that will be very helpful. I e-mailed the BLS asking for help. A very patient economist explained it to me. Rather than quoting his e-mail, I’ll just point to this special note in the press release.
If you’re really brave, you can read all about this change by clicking here.
So What’s New?
BLS is making two changes. First, they’re switching to a moving average of the data. During the transition (from today to the March 2024 CPI) the moving average period will be two years. After that, the data will be smoothed at six-month intervals. This is the second change: data being collected and reported at six-month intervals rather than annually. The moving average period will remain at two years.
A 34% decrease in a CPI component triggered my paranoia. Is it possible that BLS was manipulating the data to lower the measured inflation rate? Why not restate the historical data using the new methodology.
There are two answers to this. First, the new process uses new data. There is no historical data to make the calculations. Second, more practically, this is a negligible part of the CPI. In the December, 2012 weighting, health insurance contributed 0.7 percent of the overall index. (I’ll save you some trouble. Click here to download the Excel workbook with weights every other year starting in 2011.)
There is a third piece to the answer. The CPI is updated frequently. For example, the series CUUR0000SEME was added to the index in 2005. This is necessary and important. Changes in consumer spending patterns (especially new products) must be reflected in the CPI. For example, what was the price of an 18-terabyte computer hard drive in 1980? Nobody knows because there was no such thing in 1980. But I know they exist today. There are eight sharing the office with me.
Sometimes paranoia takes me to a dead-end. This is one of those cases. If nothing else, I got to re-learn some basic facts about the CPI. Thanks to the energetic BLS economist who helped me with this.