The site I posted is using the Bureau of Labor Statistics methodology. All price comparisons of different eras normalize the price to a specific era. But the price is not being 'normalized' how you are inferring (forced to a specific value). I reproduced a generic version of the formula below.brkriete wrote: ↑Fri Sep 09, 2022 6:32 amI read more carefully and that site is using the "CPI for gas" to normalize gas prices. It's essentially backing into a number that makes gas prices a constant. At least that's how I'm reading the methodology. So...not surprising that when you normalize the cost for gasoline to be constant it stays pretty constant.
X-year consumer good price for Z x (Y-year CPI for Z / X-year CPI for Z) = Adjusted Gas Price for Z in Y-year Dollars
We want to know the relative value or purchasing power of money for consumer goods at a previous era. Using CPI allows us to compare the price of gasoline to all other consumer goods. It would not be appropriate to use CPI for non-consumer goods. Just like it is not appropriate to use other inflation measures (other price indexes or GDP inflation) and apply them to consumer goods. We like to discuss inflation as one monolithic variable. It is not.
Common critics of CPI are that it can overestimate the rate of inflation* and it does not account for consumers substituting goods of lesser or higher quality. CPI also attaches 'weights' to various goods and services. There is disagreement with the BLS on how they weight things. Like all models, CPI is wrong. But has proven useful.
*I would argue it has underestimated the rate of inflation since the Great Recession as CPI does not include the price of housing.