I get that this may annoy a few Over My Shoulder readers. I've discussed inflation many times in my letters, pointing out that the CPI is a general model and not an exact measure. I am constantly asked what I think about John Williams and ShadowStats. The basic answer is that if you believe that ShadowStats' inflation numbers are correct, you have to somehow think that bureaucrats in 1990 knew more than bureaucrats in 2000 or 2014. The BLS is a group of sincere economists trying to accurately model what is going on in the real economy. As this note points out, CPI closely reflects the results of the MIT study called the Billion Prices Project.

I also get that your experience of inflation may be significantly different from the inflation model of CPI. But the CPI does not measure your inflation, it is a model of the general inflation in a country of 330 million people.

I think what Williams does is a little bit more sophisticated than this essay suggests, but the correlation with Billion Prices is clearly there. Food for thought.