4 ms·

The base effect is relevant. The consumer price index is 267 right now, 256 a year ago and 255.5 two years ago. So the average inflation over two years is 4.5%

The base effect is relevant. The consumer price index is 267 right now, 256 a year ago and 255.5 two years ago. So the average inflation over two years is 4.5% or 2.2% annualised -- a quite normal pace of price inflation. The main reason why inflation was near-zero in 2020-04 and too high in 2021-04 is oil prices dropping from $70 to $0, and then recovering back to $70, with the corresponding effects on gasoline prices. Had oil prices staid stable at $70, year-over-year CPI figures would have fluctuated much less and would now be 1-2 percentage points lower. That is the base effect.

jerf 5y ago
When a company in 2021 reports revenue growth of 30%, it's not really 30% growth, with all the conclusions about the company you'd normally come to based on that data, and calling "base effect!" on it is reasonable. The number has no real use. This inflation is real. The base effect doesn't make it less real. Prices have really gone up, in the relevant time period. Your choice to average it out across a particular two year window isn't invalid on its own terms, but it starts raising a lot of other questions about which years you pick and how much data massaging you can do by picking your choice of start and end and all that other sort of stuff, and it produces numbers that may also be real, but are different numbers. By contrast, if the company reports absolute revenue in 2021, the base effect does nothing to that number. The revenue is the revenue. Likewise, the inflation measurement is perfectly valid. Inflation, as defined by the measure (since by no means would I claim that it's somehow the only option or the objectively best option) really is what it is. What conclusions you come to based on that, well, we'd be arguing about that regardless of whether there's a "base effect" or not anyhow. But I'm generally underwhelmed by argumentation that appears to be trying to explain away, rather than explain. Start with the brute facts and move out from there, rather than immediately cutting to trying to argue away the brute facts. (To be clear, I'm not making any particular implicit argument. I don't currently have a solid opinion on this matter; it is still a fairly large range of possibilities to me. In general, usually some space and time is needed to even begin to "explain" any economic fact.)
- matthewdgreen 5y ago
  The question is not what you call it, but what you do about it. If we had a temporary pandemic-induced price crash and then everything returned to pre-pandemic normal, then the answer is probably "not much". But some people are concerned that this is the start of a crisis and want the Fed to raise interest rates. Deciding between those two very different policy outcomes is the important thing here, not quibbling about semantics.

jerf 5y ago

When a company in 2021 reports revenue growth of 30%, it's not really 30% growth, with all the conclusions about the company you'd normally come to based on that data, and calling "base effect!" on it is reasonable. The number has no real use. This inflation is real. The base effect doesn't make it less real. Prices have really gone up, in the relevant time period. Your choice to average it out across a particular two year window isn't invalid on its own terms, but it starts raising a lot of other questions about which years you pick and how much data massaging you can do by picking your choice of start and end and all that other sort of stuff, and it produces numbers that may also be real, but are different numbers. By contrast, if the company reports absolute revenue in 2021, the base effect does nothing to that number. The revenue is the revenue. Likewise, the inflation measurement is perfectly valid. Inflation, as defined by the measure (since by no means would I claim that it's somehow the only option or the objectively best option) really is what it is. What conclusions you come to based on that, well, we'd be arguing about that regardless of whether there's a "base effect" or not anyhow. But I'm generally underwhelmed by argumentation that appears to be trying to explain away, rather than explain. Start with the brute facts and move out from there, rather than immediately cutting to trying to argue away the brute facts. (To be clear, I'm not making any particular implicit argument. I don't currently have a solid opinion on this matter; it is still a fairly large range of possibilities to me. In general, usually some space and time is needed to even begin to "explain" any economic fact.)

matthewdgreen 5y ago
The question is not what you call it, but what you do about it. If we had a temporary pandemic-induced price crash and then everything returned to pre-pandemic normal, then the answer is probably "not much". But some people are concerned that this is the start of a crisis and want the Fed to raise interest rates. Deciding between those two very different policy outcomes is the important thing here, not quibbling about semantics.