July 09, 2021
Jeanna Smialek and Ben Casselman had a good piece in the NYT on the inflation of the 1970s and the differences with the current situation. However, it left out one important part of the story.
In the 1970s, actually in the late 1960s also, the Consumer Price Index (CPI) had an error in its construction that led it to overstate the rate of inflation relative to the current measure. In some years, especially in the late 1970s, the error was especially large, peaking at 2.6 percentage points in 1979. Here’s the picture for the official CPI used at the time, compared with the Bureau of Labor Statistics CPI-U-RS, which calculates the inflation rate using the current methodology.
1978 9.0% 7.8%
1979 13.3% 10.7%
1980 12.5% 10.7%
This mattered a lot in the 1970s because, as Smialek and Casselman point out, many wage contracts were directly indexed to the CPI. This means that an overstatement in the measured rate of inflation would show up directly in higher wages. Many rental contracts were also indexed to the CPI. This measurement error undoubtedly contributed to the wage-price spiral of the 1970s. Presumably we don’t have to worry about the same sort of measurement error today, and even if there were such an error, many fewer contracts are now indexed to the CPI. ( I wrote about this issue many years ago.)