Cats, charts, and politics
For some reason Hopper has been hogging up my camera time lately. Sometimes Hilbert hasn't been around when I have my camera out, other times my pictures of him are blurred or out of focus. But obviously he deserves his moment in the sun, so I made sure to grab a quick pic of him this week. In this one he's lounging in a patch of sunshine and keeping a close eye on his sister.
I don't remember what inspired me to look at this, but it's been a while since I've plotted pandemic job losses separately for men and women. Here's a version of the chart I posted earlier this morning split up by sex:
As you can see, the lines cross each other several times, which means your conclusions depend a lot on your ending month. If I wrote about this in November, I'd say men had lost more jobs than women. In December I'd say it was about equal. This month it looks like women have lost more jobs than men.
Which they have. As of January, men have lost 4.1 million jobs while women have lost 4.5 million. What's more, over the past three months men have been trending (slightly) upward while women have been trending (slightly) downward. And the labor participation rate has dropped more for women than for men.
This is still too small and variable a change to call the pandemic a "women's recession" or anything like that, but we might be heading in that direction. It all depends on the nature of the recovery in the second half of the year.
This has gotten a lot of attention over the past couple of days:
Kia has approached potential partners about a plan to assemble Apple Inc.’s long-awaited electric car in Georgia, according to people familiar with the matter....The likelihood of a final agreement was thrown into question when Kia’s parent company, Hyundai Motor Group, said last month, then sought to play down, that it was in negotiations with Apple to cooperate on an electric driverless car.
....Hyundai has talked to Apple about investing more than $3 billion in a deal that would see its subsidiary Kia begin building cars under the tech company’s brand as soon as 2024, a person familiar with the matter said. Under such an agreement, up to 100,000 vehicles could be assembled in the first year in Georgia, where Kia has a factory, the person said.
For some reason, all the articles about this have focused on the production of physical cars, which is frankly not very interesting. There are already lots of cars on the market, and plenty of electric cars as well.
The only interesting thing about all this is that, apparently, Apple plans to bring a driverless car to market "as soon as 2024." That would be big news if it were true, but I've read nothing recently suggesting that Apple is anywhere near this far advanced in driverless technology. I'll grant that Apple is the master of secrecy, but there's a limit to how secret you can get with this stuff. You have to have trials, you have to get permits, and you need to collect mountains of data. A few test cars won't do the job; you need whole fleets of them—and needless to say, those fleets all have to be very public.
But then again, Apple has been working on this for years, and I suppose it's possible they're farther along than anyone thinks. What's more, a huge production deal with Kia certainly suggests they feel pretty good about their chances of success.
If Apple is serious about the 2024 launch date, they'll be joining a crowd of car manufacturers who are talking pretty publicly about having driverless cars in the 2023-2026 timeframe. Even assuming that everyone is being optimistic, it does suggest that by 2025 or so driverless cars at reasonable prices are finally going to hit the market. Like so many things related to AI, it will have taken way longer than anyone predicted back in the aughts, but that doesn't mean it was all claptrap. It just means it was harder than anyone thought.
The American economy gained 49,000 jobs last month and the unemployment rate fell to 6.3 percent. Those numbers are fairly meaningless, however, so instead I'll show you this chart of job losses over the past year:
The feeble employment gains in January are being taken as bad news, but I don't think that's the right interpretation. Roughly speaking, what happened was a big disruption last March, followed by a stabilization at -10 million jobs. That's the basic job loss from the pandemic, and there's no special reason to think it's going to change until businesses fully open up following the widespread adoption of vaccines.
In other words, we're likely to stay where we are until the middle of the year, at which point we should start to see some progress as restaurants and other public-facing businesses fully reopen. By the end of the year we'll know whether some of the job losses are permanent, but right now it's just too early to tell.
The number of vaccine doses administered has been accelerating and we are now inoculating more than 1 million people per day. The chart below is a guess at how this is going to play out:
The line shows the average number of doses per day, with the acceleration slowing down over time and topping out at about 2.5 million per day. The numbers are the cumulative number of doses administered.
If we end up somewhere in this ballpark, we will have administered 400 million doses by July 1. That's 200 million people, or about 80% of the adult population of the country—assuming, of course, that 80% of the country wants the vaccine. That's well over the number needed to get the virus well under control and get the case count plummeting.
Don't take this as any kind of rigorous prediction. It's just to give you a rough idea of how the cumulative numbers add up given acceleration in the number of daily doses administered. The actual numbers will depend on how much vaccine is available and how much better we get at distributing it. Nonetheless, I think this is a fairly conservative guesstimate. By May we should start seeing real improvement, and by the end of June we should finally be in pretty good shape.
Assuming, of course, that our vaccines remain effective against whatever new variants show up . . .
Yikes! The US reported nearly 5,000 deaths on Thursday.
Here’s the officially reported coronavirus death toll through February 4. The raw data from Johns Hopkins is here.
Here's my latest M-protein test result:
The last few months have been almost eerily stable. That's good news, of course, since as long as my M-protein level is stable it means my multiple myeloma is under control
Bret Stephens has a column in the New York Times this week taking on the now-tedious subject of what a hellhole California is. This is not a topic I really feel like wasting time on anymore, other than to say (a) every state has problems, and (b) California is a high-tax-high-service state. That's not for everyone, I suppose, but it's a pretty good model for a lot of people.
However, Ramesh Ponnuru points out that one of California's genuinely worst problems is its poverty rate, due largely to our high housing costs. This got me curious about the relationship between poverty rates and urbanization. Is California's poverty rate high largely because it's heavily urban? Here's a look:
You can see two things here. First, poverty clearly increases significantly in heavily urban states. Second, California is nonetheless well above the trendline, which means our poverty rate is higher than you'd expect even given our high rate of urbanization.
There's a lot of noise here, and urbanization doesn't explain a lot about poverty, but it does explain some. In California's case, I'd guess that our heavily Hispanic population also has a lot to do with it.
Smartmatic, the voting software company that Donald Trump’s lawyers falsely accused of manipulating vote counts in the 2020 presidential election, has filed a $2.7-billion defamation lawsuit against Fox News and three of its on-air hosts — Maria Bartiromo, Lou Dobbs and Jeanine Pirro — who presented the disinformation on their programs. The suit filed Thursday in New York State Supreme Court also names Trump’s lawyers Sidney Powell and Rudolph Giuliani, who were frequent guests on Fox News programs in the weeks after the November election.
....“The company’s reputation for providing transparent, auditable, and secure election technology and software was irreparably harmed,” the suit said. “Overnight, Smartmatic went from an under-the-radar election technology and software company with a track record of success to the villain in [the] Defendants’ disinformation campaign.”
I have no idea if Smartmatic has any chance of winning this suit, or even if it has much of a case, legally speaking. And God knows, Fox has settled plenty of suits in the past for large sums, none of which seem to have slowed them down.
Still, it's worth putting them on notice that there's a price to be paid for going off the deep end on dangerous and absurd conspiracy theorizing. Maybe losing (or, more likely, settling) a few more suits like this will finally teach them a lesson or two.
This is a hamadryas baboon at the San Diego Zoo. Isn't he handsome? In the bottom picture he's playing with a baby baboon, just as any good daddy should.
