[jnordwick]

You're making a couple mistakes on this.

The first is you're using the Nationwide averages as opposed to the regional numbers for New York where a lot of these increases are much greater than on the nation.

The second thing is the way it includes housing is by using a thing called the owners imputed rent. And what that does is it tries to back out the rental from a housing unit. The problem is in New York City rent has been rising way faster than that.

30 is the cpi's consistently underestimated a number of its own provisions because of the way it does hedonics and substitution. It basically says that while meat might have risen 50% people switch to fish now and it uses in lower value for inflation.

The CPI over the last 30 years have been so massively game it's almost useless anymore

[gruez]

>The first is you're using the Nationwide averages as opposed to the regional numbers for New York where a lot of these increases are much greater than on the nation.

Another commenter has pointed out new york house prices actually rose slower compared to the rest of the country.

https://news.ycombinator.com/item?id=42043147

>The second thing is the way it includes housing is by using a thing called the owners imputed rent. And what that does is it tries to back out the rental from a housing unit. The problem is in New York City rent has been rising way faster than that.

Most Americans own their home. OER might not be perfect, but pretending that they pay market rent doesn't make much sense either. Even for people who don't own their home, new york has rent control, which provides similar inflation protections compared to owning a home.

>30 is the cpi's consistently underestimated a number of its own provisions because of the way it does hedonics and substitution. It basically says that while meat might have risen 50% people switch to fish now and it uses in lower value for inflation.

The part about hedonic adjustment is misleading. While it's true that such adjustments are used. It's only used for small minority of categories (basically clothes and technology), and doesn't include stuff like food (like in your example).

Meanwhile the part about substitution is straight up false:

https://www.bls.gov/cpi/factsheets/common-misconceptions-abo...

[jnordwick]

Read that link. The BLS says "yes we try to adjust for the substitution, but really we are trying to calculate a basic level of satisfaction:

> In January 1999, the BLS began using a geometric mean formula in the CPI that reflects the fact that consumers shift their purchases toward products that have fallen in relative price. [It continues by saying it allows substitution withing categories and not between categories.]

This is exactly what I was saying. It has become a cost of living index, not a measure of inflation anymore. The hedonic adjustments can be laughably hand-wavy (it is a very difficult problem on how you measure inflation when using CPUs -- the real answer is you shouldn't and CPI's basked is a bad bad way to do it). That FAQ is always pretty funny and defensive (a lot of "give us a break -- it doesn't really affect anything much").

> Even for people who don't own their home, new york has rent control, which provides similar inflation protections compared to owning a home.

This is one of the huge issues. (1) Not sure if you live near NYC, but ask any new yorker about how much rent control has helped -- the answer is usually pretty low. don't have a heart attack looking at this rental rise in the last 5 years https://comptroller.nyc.gov/reports/spotlight-new-york-citys...

However that isn't the important part - rent control shouldn't effect measures of inflation -- this is precisely why CPI is bad. You can't legislate away inflation, you mere push the money around and prices rise asymmetrically but you still have the fundamental issue that CPI no longer even tries to measure -- too many dollas chasing too few goods. So you stop some of using those dollars on housing -- they just bid of something else instead.

CPI is a terrible horrible very bad measure of inflation. Its been that way since the 70s at least when the concept was turned on its head.