3 ms·

The big numbers ARE driven by only one or two inputs though. Without used cars and housing, the supply of which took a huge hit for Covid related reasons, the p

The big numbers ARE driven by only one or two inputs though. Without used cars and housing, the supply of which took a huge hit for Covid related reasons, the print would be much lower.

lumost 5y ago
Perhaps we should also swap rent increases for living with additional roommates, clearly consumers are choosing more communal living situations and we can thus discount rent price increases as well. Just like we took home prices out of CPI /s If we get to choose which categories get excluded from CPI after every bad report it ceases to be a good metric.
- sokoloff 5y ago
  CPI is a survey-based methodology and captures how people actually spend their money. If something becomes rare and expensive and people curtail their consumption of it, that is itself a useful signal. If something becomes common and inexpensive and people buy a lot more of it as a result, also a useful signal. Home purchase price was taken out of CPI-U in 1983 (and CPI-W in 1985) and replaced with an owner-equivalent rent measure that gets at the cost of consumption of shelter (the service that housing provides). Tl;dr: Housing expenses were not removed; house purchases were (and replaced with a rent-equivalent measure).
  - harambae 5y ago
    >owner-equivalent rent >a rent-equivalent measure The accusation is that OER is a terrible way to measure rent increases. OER is determined by surveys asking homeowners "If someone were to rent your home today, how much do you think it would rent for monthly, unfurnished and without utilities?" It's a crazy way to measuring housing inflation for most people. Especially since there are statistics that directly just measure rent increases and the BLS could use those.
nickff 5y ago
Well there are only a few main groups, and if we generally expect that some will respond faster than others, I'd say two or three going up would be quite indicative of general inflation. It seems like food, energy, and commodities (less food and energy) have gone up at different rates. The only major group that's been relatively stable (still about 4%) is services, and it seems likely that it's because many of them have been made illegal/unattractive for many recent months. https://www.bls.gov/news.release/pdf/cpi.pdf https://www.bls.gov/news.release/pdf/cpi.pdf

The big numbers ARE driven by only one or two inputs though. Without used cars and housing, the supply of which took a huge hit for Covid related reasons, the print would be much lower.

lumost 5y ago

Perhaps we should also swap rent increases for living with additional roommates, clearly consumers are choosing more communal living situations and we can thus discount rent price increases as well. Just like we took home prices out of CPI /s If we get to choose which categories get excluded from CPI after every bad report it ceases to be a good metric.

sokoloff 5y ago
CPI is a survey-based methodology and captures how people actually spend their money. If something becomes rare and expensive and people curtail their consumption of it, that is itself a useful signal. If something becomes common and inexpensive and people buy a lot more of it as a result, also a useful signal. Home purchase price was taken out of CPI-U in 1983 (and CPI-W in 1985) and replaced with an owner-equivalent rent measure that gets at the cost of consumption of shelter (the service that housing provides). Tl;dr: Housing expenses were not removed; house purchases were (and replaced with a rent-equivalent measure).
- harambae 5y ago
  >owner-equivalent rent >a rent-equivalent measure The accusation is that OER is a terrible way to measure rent increases. OER is determined by surveys asking homeowners "If someone were to rent your home today, how much do you think it would rent for monthly, unfurnished and without utilities?" It's a crazy way to measuring housing inflation for most people. Especially since there are statistics that directly just measure rent increases and the BLS could use those.

nickff 5y ago

Well there are only a few main groups, and if we generally expect that some will respond faster than others, I'd say two or three going up would be quite indicative of general inflation. It seems like food, energy, and commodities (less food and energy) have gone up at different rates. The only major group that's been relatively stable (still about 4%) is services, and it seems likely that it's because many of them have been made illegal/unattractive for many recent months. https://www.bls.gov/news.release/pdf/cpi.pdf https://www.bls.gov/news.release/pdf/cpi.pdf