No, it really isn’t. From your link:

“If someone were to rent you home today, how much do you think it would rent for monthly[...]”

That’s a very poor proxy for housing cost.

First of all it’s a guess. A guess from whom? What do these oracles know about how much their home can rent for if they haven’t put it up?

Second, the homes available in the rental market are vastly inferior from (for example) my home. I cannot find a home remotely as nice as my home (bought in 2017 for 320k, currently valued for 400k - pretty average middle class home in my city) but I can’t rent mine for what I’d want to rent it for [1] - so the supply and demand price don’t overlap for my type of home in my city. Without overlap of price, the question, as it pertains to my home is meaningless.

On the other hand, there is an easy way to include housing price for owners. Add price of homes to the CPI. In fact, investments generally should be included, since they are goods whose price raise with the increase in monetary mass

[1] is the accelerated deterioration of my home, and the headache of being a landlord worth the $1400/month renting it to a family? No. Can I rent it for $1800 by splitting it into rooms, say, for students? $1800/month is not worth having destructive students live here.

It's certainly true that calculating imputed rent can be tricky and is probably subject to some error. However that does not make is meaningless.

Ok, let’s say not meaningless.

Let’s say it has (my guess of my particular case) a 30% error. What happens to CPI if the weighted average of housing is off by 30%? Assuming housing is a third of CPI, that puts CPI up to 10%.

Look, I’m actually looking to buy a home out in the country and remote work from there. I’m actually living the problem of “should I sell my home or rent it out?” It’s not a survey question, that’s for sure!

It's less of a problem than you think because the error can't compound. An error of 30% compounded over 10 years can't grow to a 1300% because if you compare the two years directly you'll see there's a huge mismatch.

“If someone were to rent you home today, how much do you think it would rent for monthly[...]'

That’s a very poor proxy for housing cost."

Agreed. Rental data is publicly available. Why bother asking only to get a worse answer?

The question is asked to owners. Renters are asked how much they actually paid in rent. The main reason why owners are asked for their estimate is because owner-occupied houses can be systematically different (probably more expensive/higher class) than renter-occpied houses, so estimating owner-occupoed rents with actual rents can be biased.

The GP is mistaken in assuming that rent is excluded from the CPI, however mortgages and other costs of home ownership are not considered consumption and are therefore excluded. This is the first item of the FAQ in the document that you linked.

As a matter of fact, home ownership has become a highly speculative play for the ordinary wage earner and this is largely due to asset price inflation as a result of low-interest rate policies (worldwide, not just the US).

It's important to understand that while mortgages are not part of CPI, imputed rent is. So the consumption costs of living in a home one owns are taken into account.

Imputed rent is meaningless since there’s no market discovery of the price.

It’s like asking what the position and momentum of particle is simultaneously. You can’t know the price of something you don’t intend to rent unless you rent it, but you’re not renting it because you have no intention of renting.

Even if you “pretend” to rent by putting out a fake add and then not signing the rental agreement, that doesn’t count. That only gives you the demand side of the price. Since you know that you don’t intend to actually rent your home, you will not give serious consideration the opportunity cost of renting out your home.

>Even if you “pretend” to rent by putting out a fake add and then not signing the rental agreement, that doesn’t count. That only gives you the demand side of the price. Since you know that you don’t intend to actually rent your home, you will not give serious consideration the opportunity cost of renting out your home.

The point of the fake ad is to gauge what the current market clearing price is, which presumably already takes into account the current supply/demand. It's not any different than using the current stock price to judge what an entire company is worth (market cap), even though only a fraction of the shares are actually on the order book.

Two problems with that:

1. Clearing price assumes efficient markets. My house in my neighborhood is one of maybe three (probably much less judging by the rental units friends of similar means live in) like it that will be available throughout the year.

I don’t agree a fake add will get the clearing price in these conditions.

Even if I put up an ad to rent a room in a student house -one of tens of thousands in a typical college town- the hypothesis fails since students are extremely sensitive to distance to their main building, nearby amenities, availability of parking, etc.

2. Even in the stock market the clearing price is dubious; the market cap of a company is a useful bit of fantasy. By multiplying the clearing price by the outstanding shares there is an implicit and incredible assumption that all the shareholders value the shares equally - which is obviously not true, some share-holders believe the stock has more promise, others are sentimental (the founder’s family members).

And that ‘price*shares == company worth’ is not true is evident in mergers and acquisitions which can be very acrimonious events. If the clearing price times the outstanding shares were a fair measure of the market cap, why would rational investors care about a merger except for the costs associated w/ closing?

But bringing this all back to the OP. Housing cost for the majority of the middle class is not meaningfully included in the CPI, except with a flawed metric.

>1. Clearing price assumes efficient markets. My house in my neighborhood is one of maybe three (probably much less judging by the rental units friends of similar means live in) like it that will be available throughout the year.

Yes, low liquidity in the rental market means that there's a spread in the estimates, but I don't see how it follows that "imputed rent is meaningless". The low liquidity can cause the estimate to be both above or below reality. Do you have evidence that the inflation numbers for the housing component of the CPI is totally detached from actual market prices?

>2. Even in the stock market the clearing price is dubious; the market cap of a company is a useful bit of fantasy. By multiplying the clearing price by the outstanding shares there is an implicit and incredible assumption that all the shareholders value the shares equally - which is obviously not true, some share-holders believe the stock has more promise, others are sentimental (the founder’s family members).

Again, this seems to be more of a complaint that the market cap figure isn't 100% accurate, rather than that the number is "meaningless" as claimed in your original comment. It seems like the entirety of your argument (for both imputed rent and market cap) is "well there are these factors that causes the number to be inaccurate, therefore the number is totally meaningless and should be ignored".

> It's not any different than using the current stock price to judge what an entire company is worth (market cap)

It's interesting to note, the stock price is a small part of a company's valuation. There are many different structures for stock issuance.

It's not meaningless. It's an estimate based on market data.

Every component of CPI is an estimate, that doesn't mean it's meaningless.

As others have pointed out, this is handled differently in the UK versus the US. You are correct for the US standard definition of CPI, but the author of the article is a professor in the UK, writing on a UK domain for a (primarily) UK audience. Here's a link that another poster provided that summarizes the UK definitions: https://www.thetimes.co.uk/money-mentor/article/rpi-versus-c....

My biggest beef with housing in the CPI is the fact that housing costs are underrepresented in general, and drastically underrepresented for some segments of the economy. The weight of housing in the index is something like 24%, but the average housing costs for the top 20% of incomes is closer to 29% of income, and it gets above 40% of income by the time you get to the bottom 20% of incomes.

So by underrepresenting housing's weight by that much, combined with the fact that housing prices are growing much faster than everything else in the index, combined with the drastic disparity in housings effectively "felt" weight on inflation for people of different incomes, and you get this problem: things may seem fine to a banker on the Board of Governors of the Federal Reserve, while the bottom 20% rightly thinks our nation is going up in flames. There is a massive disconnect between what the metrics say and what people are actually experiencing, and nothing is happening about it because the metric still feels right to the wealthy people that control it.

Shelter accounts for 33% of the index, not 24%: https://www.bls.gov/news.release/archives/cpi_02102021.htm - I'm not sure where your figure originates, but it's wrong.