Where have I been all day?
Well, I woke up a little late and found nothing interesting in the news. The queen is still lying in state. The Russian army is still inexplicably incompetent. The FBI is still issuing subpoenas to the 1/6 rioters. Railroad owners are still threatening to shut down the entire country's rail system. And LA's sheriff is still a huge douchebag.
So I ate breakfast. Then I decided to let Charlie into the yard for the first time in several months to see how he did. Answer: he roamed farther than I'd like, but still only a couple of houses away and he came back of his own accord.
Then I ended up spending a bunch of time on some charts, but I got hungry before I was finished so I went to lunch and then bought myself a pair of astronomy pants.¹ I figured I would post the charts when I got home, but a friend sent me a question that I figured I could answer right away and . . . it turned out I couldn't because the stupid Census Bureau doesn't provide easy access to old income figures.²
Anyway, here are the charts. They aren't designed to make a point, just to satisfy some curiosity on my part. They show which categories of the economy are contributing the most to inflation. I left out gasoline and housing, since they present some difficulties I didn't feel like taking the time to solve, but everything else is there. Note that this time I simplified the charts by calculating the trendlines and then erasing the actual data.
¹What are "astronomy pants," you ask? Well, it's pretty warm out here, and years of chemo has made me cold blooded. So I normally wear shorts when I go out. But twice I've tripped and skinned my knee, so I figure it's time to wear long pants even if I don't like them. So I got some.
²In case you're wondering, he wanted to know how much annual earnings had changed since 1979 for (a) men, (b) 25 years and older, (c) with only a high school education, and (d) who are working full-time. The answer, adjusted for inflation, is $1,214 in 1979 and $952 today. That's according to the Census Bureau, anyway. The BLS has a different estimate. Note that virtually all of this decrease took place in the supposed boom times of the 1980s.
In case you missed this:
https://www.cnet.com/culture/fashion/these-cool-clothes-could-help-fight-global-warming/
That coupled with this is very interesting:
https://nordot.app/942816952346492928?c=592622757532812385
A car that CONSUMES CO2? It's not a lot but it's a start.
Made from recyclable plastic.
And stylish too
Any sign of that gray hairstreak butterfly when Charlie returned home?
Thus getting rather divorced from reality. Unless, as others have mentioned, you have some special insight into how economies work to know which sort of trendline to use…
A lot of things follow a zigzag or sawtooth path, including especially GDP growth. Inflation is even less regular, although it also is up and down. There is practically no such thing as a straight line or consistently curved path in economics, over the long term.
Thing do often follow straight paths for a limited time, like the up path in the GDP=growth sawtooth. But if you can't predict the inevitable inflation points (and nobody can), you are not really predicting the economy
inevitable inflection points (which obviously occur in the inflation path also)
‘Zigzag path’, indistinguishable from ‘noise’.
No, there are real expansion and recession paths in the GDP (for example), which form an overall sawtooth pattern. The pattern is irregular, so you can't predict when recessions will occur. These ups and downs are not measurement noise, although the month/month differences in some things (not GDP) that the media play up are sometimes noise.
The difference between 2% and 8% inflation is also not noise, when it keeps going for many months. The US economy is just not prone to continuous high inflation - inflation goes up only when special factors are involved. We can often see what is causing inflation, but at present several factors are involved, and nobody can accurately predict when all of these will return to normal. But when they do return to normal, inflation will subside. Wage increases are not a major factor currently.
Not divorced from reality. The advantage of fitting a line to the data is that it uses information from multiple data points, which should attenuate noise relative to signal. The disadvantage is, as you say, implying that we know more about the direction of the economy than we actually do. A further disadvantage of ordinary least-squares (which I presume is what Kevin is using) is that it weights all data the same, older data equally as newer. A better analysis would use LOESS, which is agnostic as to underlying functions, and weights nearby data more than farther data, for each point on the smoothed line. Thus the shape of the rightmost segment of the smoothed line represents recent data more than older.
OK, give some references to accurate predictions that have been made by fitting curves, however sophisticated. Has Kevin done this? He should post the results of his past predictions on inflation and everything else in the economy. Actually so should any economist who claims to have a prediction, but they will never do it.
The economy, and aspects of it such as inflation, often follow simple linear or curved paths for limited times, but those paths inevitably change.
I guess I wasn’t clear. LOESS is a data-smoothing algorithm; it doesn’t extract a trend, so it doesn’t support prediction. It’s a better look at what has been happening, not what will happen.
The acronym is for LOcally Estimated Scatterplot Smoothing.
That is good to know. I've never been motivated enough to look that up.
I can make a pretty confident hypothetical prediction that inflation will come down before too long (within a year or two?) if the Fed does not jack up interest rates. I can do this because this is what happened in the many years between the 20's and the late 60's when the Fed did not jack up interest rates and episodes of inflations, caused by massive government spending in war time, nevertheless came to an end.
The prediction is hypothetical because the Fed obviously intends to jack up interest rates again. When inflation does come down the Fed will take credit regardless of what the real causes were. Or I should say that economists will give the Fed credit because they have quasi-religious faith in the dogma that the Fed control the economy with interest rates.
The problem for them is that Russia will never accept any peace that isn't their victory, they really hold a grudge and will keep at trying to destroy you or ruin your lives any trivial or greater way they can for centuries, because that is who they are.
You may win against them today but they'll keep at undermining Ukraine or Georgia and everywhere else on their border until they think they can try again and you'll have to fight them all over again in a few decades.
Russia requires not just political revolution but a massive and extreme societal upheaval that will, at the least spin chaos off in every direction for most of the rest of the century.
As Kipling said back in 1898: Make ye no peace with Adam-zad, the Bear that looks like a man!
Moldovans and Georgians ought to volunteer for a ‘foreign legion’ in Ukraine to learn how to beat the Russians …
Chisinau, brah.
Moldova can just call in Romanian Dracula battalion.
i'm vaguely disappointed in your explanation of "astronomy" pants.
WE demand a picture of "astronomy pants". (accent like we demand a shrubbery)
where's our swdish troll ...
https://www.nytimes.com/2022/09/14/world/europe/sweden-election-result-right.html
Perhaps he was called up for duty in Donbas…
Havn't seen silver OR sturestuhl in ages.
Prolly busy at Internet Research Service Stockholm Desk agitating for reprisals against the Aland Islanders in the event they push for reunion with Finland.