Today was a serious crash day: I was out for nine hours straight from 9 am to 6 pm with only a short wakey period for lunch. I'm told that cats were sitting on me, apple sauce was being made, etc., all while I was snoozing away.
So let's see if I can squeeze in a quick post before midnight. How about something related to inflation? You all know how much I love inflation. Here it is:
This chart shows three measures of inflation that are designed to avoid the variability of headline CPI and give a better look at the real level of inflationary pressures:
- Core CPI omits food and fuel because they jump around a lot and don't really tell us anything about underlying inflation.
- Trimmed mean CPI cuts off the biggest gainers and losers so that a few outliers don't affect the bulk of the items in the CPI basket.
- Median CPI looks at the inflation rate of the median item in the CPI basket.
For those of you who want evidence that inflation isn't all that bad right now, median CPI is your hot ticket. Not only is it relatively low at 3.5%, but it's only 36% above its average 2016-2019 level. The others are 116% and 135% above their 2016-2019 averages.
And if you're an inflation hawk? Then ignore all of these and just use headline CPI, which is currently at 6.8%.
POSTSCRIPT: Here is the conclusion of a recent paper that studied the accuracy of various measures of core inflation:
The last two years have been highly informative about the behavior of alternative core measures. Headline inflation has fluctuated erratically....XFE [eXcluding and Food Energy] inflation has performed quite poorly....Fixed-exclusion measures of core that exclude a wider set of industries, such as the Atlanta Fed’s sticky-price inflation rate, have performed better, but the most successful measures have been weighted medians and trimmed means.
What about the trailing average of headline CPI? After all, people do eat and use fuel so it seems unfair to exclude them.
OTOH, if they make the chart too noisy, that's a problem so why not use trailing averages? Say, 12 months?
Ought to be fairly close to median but let's find out... 🙂
The usual year/year inflation numbers are in fact an average over the last 12 months. For some purposes it is better to use the month/month numbers since they more closely reflect the latest trend. I pointed this out some time ago and Kevin has used the month/month numbers at times, but today he seems to be using year/year (which is what is reported in the media). According to that the latest total or headline rate is 6.88%, but the month/month rate is 9.35%.
But again, you can't predict future inflation by fooling around with these different numbers. You must look at the actual factors which are causing price increases. The obvious things are a) supply-chain and other covid-related things and b) stimulus payments. These are probably transitory - there are no huge direct payments planned - so inflation is likely to come down eventually.
The idea that inflation becomes "baked in" actually is not consistent with the evidence. It is something that economists invented to explain how the Fed failed to control inflation in the 70's by raising interest rates.
The massive increases in the cost of housing in urban areas never seemed to cause reports of inflation in the media. Why is that?
Despite huge increases in home prices in many areas, the CPI for housing has not increased at a huge rate:
https://fred.stlouisfed.org/graph/?g=KmBI
The latest rate, 4.8%, is high, but not big enough to be driving total inflation, which has been higher (6.9%) and looks scarier for the purposes of the media. Only a small fraction of people are buying houses at any time, whereas everybody is buying food, etc. This is one factor which seems likely to be pushing inflation up in the future.
So really… all these inflation numbers are irrelevant. If the price of a chicken breast goes up, I might switch to something cheaper. Or maybe I have plenty of disposable income and simply don’t care. And since lots of people continue to buy pick up trucks and massive SUV, I don’t think they are sensitive to changes in gas prices. Bu the media would have us believe that 10s of millions are going to go hungry because of the dreaded inflation.
The government in power must be punished. So be it.
This is ridiculous. Gas prices collapsed in 2020. Calling a return to “normal” inflation is irresponsible.
“The cost of gasoline is up 58 percent in the last year, while health insurance prices have fallen almost 4 percent. Meat prices are up 13 percent, dairy 1.6 percent. Boys’ apparel is up 8.4 percent, while girls’ apparel is down 0.4 percent.
Differences like that mean that the inflation rate a person faces depends on what that person buys and where he or she lives and shops. People who live in more rural states, for example, most likely drive significantly more miles per year — so fuel inflation would matter a great deal to them.”
If this is best our economists can do, then they will ensure Biden’s defeat.
https://www.nytimes.com/2021/12/30/opinion/inflation-economy-biden-inequality.html
There are a few decent points about how inflation doesn’t really matter to lots of upper middle class people, but otherwise it’s a terrible article.
https://www.usinflationcalculator.com/gasoline-prices-adjusted-for-inflation/
In 2020 dollars - since any time series of money is supposed to be inflation-adjusted...
Year USD
2016 2.280
2017 2.263
2018 2.254
2019 2.258
2020 2.242
Their data goes back to 1978 and includes un-adjusted figures. The collapse you assert was ~16.9% in unadjusted prices between 2019 (2.698) and 2020 (2.242). You might try to argue then that ~20 percentage points of that 58% increase was "just getting back to where we were" but that still leaves a very large increase in the price of gasoline.
Gas price at my local station is $2.90 today. At their lowest in 2020, it was $1.90. A few weeks ago it was $3.25. Sounds like I’m experiencing deflation now. Oh well, fuel prices are volatile. In 2008, it was nearly $4.00.
If trying to understand inflation, why would you adjust prices and report they haven’t changed?
The perils of trying to extrapolate from a single datapoint.
I guess my real point is that if you want to talk about inflation, you should never talk about gas prices. In that piece, the writer highlighted gas prices as an issue. That was silly.
Thanks for the link to the Austan Goolsbee article. To be fair, I think Goolsbee mentioned gasoline prices only to kick off a quick illustration of how variable the effects of inflation are. As in, "Sure, gas prices are up a lot, but other things are hardly up at all, or even went down."
Inflation and deflation are defined as decreases and increases, respectively, in the value of a currency. As such, the effect on prices would be a fixed multiplier, in the absence of other factors. Variations in the percent change from commodity to commodity are entirely due to other, particular factors.
Had Trump won reelection, likely I would select the highest of these figures and used it as evidence of policy failure.
If I am being fair-minded, I think the comparison of US inflation to Eurozone inflation is reasonable: the US is about (yes it depends on what index you use) 2% higher, in terms of inflation, than Europe.
In other "inflationary" news, on Jan 1 the minimum wage in New Mexico goes up to $11.50. Cue the Republican howls of outrage! The squeals of business owners who insist they'll go bankrupt!
And the Republicans in CA will howl about the minimum in CA going up to $15 or $14, depending on size of business.
"Businesses are fleeing CA!"
Expect a lot of new fast food joints in NM since they have fled CA,
It would seem the Median CPI serves the purposes of Inflation Doves only recently. The rest of the time covered by the chart makes it look like it would be to them a slimy cherry if picked.
Only if you think inflation at 2.5% is dreadful. I don't.
It’s odd that a median lies above a kind of average (trimmed mean) for an extended period. I checked, and the trimmed mean discards the most extreme 8% from each tail (highest and lowest) of the price-change distribution. The median discards the highest 50%-1 and/or lowest 50%-1 values. Presumably the central 84% is skewed toward larger increases.
Thinking about the factors underlying price changes, inflation/deflation = change in the value of a currency, which would be a constant additive factor in all prices. Prices are obviously also affected by idiosyncratic factors. All aggregate measures of inflation based on a market basket assume that these idiosyncratic factors average to zero. That seems unlikely; there are obviously systemic factors that affect large groups of prices (energy, labor, shipping). It’s also quite possible that prices are not symmetrically ’sticky’. Bottom line is that mean or medians of market-baskets are likely ‘contaminated’ by factors other than currency value.
Sorry, multiplicative factor in prices, not additive.
Reporting the percent change of a percent change is extremely misleading, especially when the underlying percent change (change in prices) is itself small. What if there were a recession and inflation was zero or negative? OMG the percent change of the inflation rate is infinite!
You know, in an earlier post, Kevin highlighted all the really good metrics that the US economy is enjoying. I think a better rhetorical stance is to enumerate these - really strong GDP growth - after accounting for inflation. There is a strong job market. And so on. And some inflation is the price that the policies we enacted cost.
We are facing a very serious issue with the pandemic and countermeasures. We're doing really well, but it doesn't come for free.
I remember when the Domino's Pizza guy was testifying to the costs of insurance requirements if the rules went thru--it was like a nickel or maybe a quarter per pizza. And people sneered at him. He also said the company would have to sell a million more pizzas (if I could call it that) a year--during the advertising campaign saying they were giving away a million pizzas.
Inflation is relative. And people notice pain, i.e. increased prices, more than cheaper products. But mainly if the price increases are on a more "essential" item. Casual purchases, not so much--unless it crosses a threshold and you are paying in cash. If $20 covers everything, including tip, plus you'd get change--you remember when it no longer covers the tip or the purchase itself.
Just a covid mention---DC is ca. 300 new cases/day/100K; NY an NJ ca. 200 (and rising), ten others at ca. 100 to 200, with a few more ready to join the club.
IF new daily infections are reported to be 0.2% of the population, that's actually about 1 to 2% of the population being infected, and if you use the 5 day infectious period, 5 to 10% of the population has active infectious cases. That also means 5 to 10% of the population is getting infected every week. Almost all the vulnerable population will be hit within, what, two or three weeks once cases take off. Cases will plummet after the first or second week in Jan in many places.
Hospitalizations are up. but not skyrocketing--at least not yet. The Merck and Pfizer pills may not be in widespread use until after this peak. Right now we have to hope that omicron is indeed less nasty than other variants and/or residual "natural: immunity is good enough. Yes, even with breakthrough cases, it's mainly the unvaccinated who are hospitalized.
I thought the likes of DC, NY, and NJ were virtually fully immunized.
Percent of population fully vaccinated (official sources)
DC: 67
NY State: 69.7
NJ: 70.42
And fully vaccinated doesn't mean boosted.
Golack,
I would like to have confidence oj your figures. If. In reality , there are 10 to 20 times more actual omicron infections than confirmed cases, then omicron is far far less deadly than delta and really at cold level below flu. And the build up of natural omicron immunity will end this quite quickly.
Unfortunately I just feel 10 to 20 times is too high, even for omicron. But still possible and I have some hope you are right. The data from London where cases do seem to be peaking already and from South Africa do indicate that the actual infections are a much higher multiple vs. delta, because cases cannot peak until natural immunity reaches the herd immunity level.But I am now somewhat discounting that because I really think that micro infections play a role in boosting immunity below the threshold we call an infection.
So maybe your high " infection " numbers are correct, but only if you include micro infections that did nothing except boost immunity . We would not pick up most micro infections even of everyone was tested every other day.
But end result is same as what you say so still good. Only difference is in whether there are a lot of actual infections which we are not picking up because not tested. Or some are micro infections which testing mostly is not going to pick up.
And maybe partially because it appears that you do have a significant number of false negatives with pcr test with omicron .
A week ago I was quite worried as to how bad the worst case might be. But, with another week of data , that worst case possibility seems to be unlikely.
Good to know, thanks for posting Kevin.