@GatorGuy wrote:
What you are describing is chained CPI. The vast majority of time when inflation and CPI is mentioned it is not chained CPI. Chained CPI is also not used for i-bonds. There is nothing being kept "artificially low". It is just one more measure of inflation.
You're correct in that I drifted a bit, and should have been clearer. That's on me.
But CPI is being kept artificially low. CPI is directly tied to major government expenditures, because it's used to calculate social security checks, government pensions, food stamps, and much more. Since the federal government isn't exactly known for staying out of debt, they have a a strong incentive to favor lower over higher numbers. And they're acting in suspicious ways. They don't release the raw data they use, which means they don't have to worry about independent audits. They've also tweaked the methodology they used to calculate the CPI at least 20 times, over the last few decades. So they have a strong incentive, and the means and opportunity. And at least one of those changes involved substitution -- while substitution between categories (chained CPI) is not used for CPI-U (iBonds), one of the tweaks the Bureau of Labor Statistics (BLS) made in 1999 incorporated substitution within categories. All these changes mean CPI numbers from the 1970s are not directly comparable to CPI in the 2020s. They're measuring different things. If you'd like to see the article the BLS published where they describe incorporating substitution within categories, it's here:
https://www.bls.gov/mlr/1998/10/art1full.pdf
While the BLS is not transparent when it comes to how the CPI is calculated, they've publicly talked about how they use census data and consumer spending patterns. So even without considering the methodological tweaks, the CPI measures cost of living, rather than inflation (the rise in prices).
While these do explain why iBonds won't keep up with inflation, it is a bit tangential to the thread's topic, so I'm going to drop the subject.
