4 ms·

> Food, energy, and housing are what impact consumers the most. And both food and energy are very chaotic (especially due to geopolitical events: thanks Russia

> Food, energy, and housing are what impact consumers the most. And both food and energy are very chaotic (especially due to geopolitical events: thanks Russia!). A couple of years ago there was much rending of garments over gas prices in the US: now, not so much. So that is why inflation is often also looked at without those two components because they can vary so widely, even month-to-month (anyone else remember when oil went negative in 2020?).

nightski 3y ago
I still hear a lot about gas prices today to be honest. They are still quite high relatively. But I think that is the largest failure of Core CPI to the point it makes it almost worthless. Is it more stable? Sure, but stability isn't great if it doesn't measure the right thing.
- throw0101b 3y ago
  > But I think that is the largest failure of Core CPI to the point it makes it almost worthless. It's not: if policy was determined on non-Core, then one month the Fed would have to raise/lower rates, and then the very next month it would have to do the opposite. The point of Core is to removed the gyrations of commodity (traders). See Chart 1 of: * https://www.frbsf.org/education/publications/doctor-econ/2004/october/core-inflation-headline/ https://www.frbsf.org/education/publications/doctor-econ/200... The red line is whipsawing all over the place. You want to set policy on that? Paper from the Bank of Canada on the topic: * https://www.bis.org/publ/bisp05d.pdf https://www.bis.org/publ/bisp05d.pdf
  - dragonwriter 3y ago
    > if policy was determined on non-Core, then one month the Fed would have to raise/lower rates, and then the very next month it would have to do the opposite. Well, except Fed policy doesn’t really use CPI (Core or otherwise) at all. The Fed's evaluation of prices that informs monetary policy centers on the PCE, not CPI. And while that's their key indicator, their actual analyiss underlying policy decisions isn't just applying a fixed function of current top-level indicators to get a monetary policy as output; there is considerable analysis of trends and what is going on below the top-level indicators, etc. You don't need to board consisting of multiple Senate-confirmed officers with large supporting expert staff to apply a simple function of headline indicators from other agencies to produce an output, so the BLS not separately calculating and publishing core CPI wouldn't effect the Fed doesn't use CPI, and does a level of analysis with the indicators it does use that if it thought something like the particular aggregate used for “Core” was necessary and not in the publication, they’d just recreate it from the lower-level components.
- dragonwriter 3y ago
  Core CPI isn’t headline CPI. Its an additional measure that is published alongside the headline CPI; the removal of components with a high non-seasonal volatility [0] is thought to potentially make it more useful in providing early indication of changes in longer term trends; literally no one presents it as a better measure of current inflation. [0] the components involved also have a high seasonality, but that can be addressed via seasonal adjustment, but the high non-seasonal volatility means that fluctuations in those components can drown out early trend indications in ways that seasonality adjustments can’t compensate for.

nightski 3y ago

I still hear a lot about gas prices today to be honest. They are still quite high relatively. But I think that is the largest failure of Core CPI to the point it makes it almost worthless. Is it more stable? Sure, but stability isn't great if it doesn't measure the right thing.

throw0101b 3y ago
> But I think that is the largest failure of Core CPI to the point it makes it almost worthless. It's not: if policy was determined on non-Core, then one month the Fed would have to raise/lower rates, and then the very next month it would have to do the opposite. The point of Core is to removed the gyrations of commodity (traders). See Chart 1 of: * https://www.frbsf.org/education/publications/doctor-econ/2004/october/core-inflation-headline/ https://www.frbsf.org/education/publications/doctor-econ/200... The red line is whipsawing all over the place. You want to set policy on that? Paper from the Bank of Canada on the topic: * https://www.bis.org/publ/bisp05d.pdf https://www.bis.org/publ/bisp05d.pdf
- dragonwriter 3y ago
  > if policy was determined on non-Core, then one month the Fed would have to raise/lower rates, and then the very next month it would have to do the opposite. Well, except Fed policy doesn’t really use CPI (Core or otherwise) at all. The Fed's evaluation of prices that informs monetary policy centers on the PCE, not CPI. And while that's their key indicator, their actual analyiss underlying policy decisions isn't just applying a fixed function of current top-level indicators to get a monetary policy as output; there is considerable analysis of trends and what is going on below the top-level indicators, etc. You don't need to board consisting of multiple Senate-confirmed officers with large supporting expert staff to apply a simple function of headline indicators from other agencies to produce an output, so the BLS not separately calculating and publishing core CPI wouldn't effect the Fed doesn't use CPI, and does a level of analysis with the indicators it does use that if it thought something like the particular aggregate used for “Core” was necessary and not in the publication, they’d just recreate it from the lower-level components.
dragonwriter 3y ago
Core CPI isn’t headline CPI. Its an additional measure that is published alongside the headline CPI; the removal of components with a high non-seasonal volatility [0] is thought to potentially make it more useful in providing early indication of changes in longer term trends; literally no one presents it as a better measure of current inflation. [0] the components involved also have a high seasonality, but that can be addressed via seasonal adjustment, but the high non-seasonal volatility means that fluctuations in those components can drown out early trend indications in ways that seasonality adjustments can’t compensate for.