Month: July 2023

I don't have any special reason for writing this post. There's no hook. It's just something that bears knowing about.

The failure of Democrats to maintain support among the white working class is a frequent topic of concern. But the truth is that it's not really a thing. Not nationally, anyway. Democrats have lost a huge amount of support among white voters in the South:

Starting in the mid-90s, Democrats began shedding support in the South to Republicans in massive numbers. In 1992 there was little difference between the parties. Today, white voters in the South prefer Republicans by 27 points on a "warmth" scale of 1-100.

But that's the South. Here's what it looks like everywhere else:

White voter preferences have tracked almost identically between Democrats and Republicans all the way to the present day. Outside the South, Democrats are mostly appealing to white voters just fine.

Now, it's true that these charts are averages for voters of all education levels, and education is a big factor among voters these days. But this doesn't change the picture much. White high school and college grads have almost identical party sentiments in the South; it's only elsewhere they diverge. It's not a non-factor outside the South, but it's not an overwhelming issue either.

The moral of this story is that in American politics you always need to break apart the South and non-South. It is the great dividing line. Looking at national averages will almost always mislead you.

How tight is the labor market? I've long been mildly skeptical that it's as tight as people seem to think, and today Jason Furman tweets the following chart:

If this is correct, the labor market was equivocal around 2021, with two measures suggesting looseness and three measures suggesting tightness. Since then, the measures showing tightness have come down and are now only slightly above normal.

So where does that leave us? My take is straightforward: after you've charted and analyzed every possible measure of tightness in the world, you're left with one thing: wages. If the labor market is tight, wages go up. If wages aren't going up, then the labor market isn't tight. This is an iron law. So how have we been doing on this score?

Since 2019 real wages have barely budged, and since 2021 every single wage measure is down. I don't understand it, but this just isn't consistent with a tight labor market. In the early days of inflation you could explain it by saying employers were simply behind the curve and didn't realize they needed to increase pay substantially. But it's been over two years now, and there's plenty of evidence that employers have raised prices well beyond inflation levels and reaped higher profits as a result. They know exactly what's going on and they can afford to pay more if they want to. But they still haven't raised wages to attract supposedly scarce workers.

So tell me again how this can possibly represent a tight labor market?

Here's a pigeon hanging out on a traditional French chimney. But where? I can't quite remember. It's just across one of the bridges on the right bank of the Seine, but I can't put my finger on exactly where I was.

Here's a little something not to take too seriously. Today the Institute for Supply Management released its latest indexes, one of which is a price index for services. This is a number worth following since services continue to be the component of inflation that's staying stubbornly high.

However, ISM doesn't report a specific price increase or decrease. It surveys its members to find out how many say prices are going up and how many say prices are going down. Any index above 50 means prices, on average, are going up.

But how much? This is the sketchy part. ISM doesn't say, but you can make a reasonable assumption that the more people who are reporting price rises, the higher those rises are. Then you can do a bit of curve fitting to extract something that looks reasonable. Here it is:

This is dodgy. And it's probably more comparable to PPI than CPI, since it's based on a survey of corporate purchasing managers. However, one advantage of this is that it doesn't include housing prices, so that's not skewing their results. In the end, it suggests that services inflation is coming down faster than CPI indicates, which jibes with other market signals.

By itself this doesn't mean much. But combined with other trends it's yet another weakish indicator that inflation is already coming down significantly.

The latest JOLTS data is out. Here's a nickel summary of changes from the previous month:

• Hires are up 1.8%.
• Job openings are down 4.8%.
• Quits are up 6.6% (shows confidence about getting new job).
• Layoffs are down 2.2%.

Aside from the drop in job openings, these are decent numbers. However, the longer term trends are all negative. Over the past year, job openings are down by 2 million. Hires have dropped 0.6 million. Quits have declined by 0.5 million. And layoffs have increased by 0.2 million.

None of this means a recession is in the offing. The trends aren't that bad. But it probably does mean that the economy is visibly softening.

Over at the New York Times, Jessica Grose urges us to be optimistic about vaccines even with nutballs like RFK Jr. running around badmouthing them. And she's right. Here's the latest CDC data for childhood vaccines:

Vaccine takeup has been generally flat or up for the past decade and showed no sign of slowing down during the COVID pandemic. The takeup rate for the combined 7-vaccine series increased from 69.8% to 70.1% between 2019 and 2021.

Takeup of the COVID vaccine has been (slowly but) steadily improving too:

Things are pretty flat, but in the US the vaccine rate went up half a point in early 2023 and is now safely over 80%. That's pretty solid.

On the bad news front, only about 40% of Americans have received a booster COVID dose, the lowest of any major country. On the childhood front, vaccine takeup is strongly related to insurance. Here are the rates for the 7-dose series:

• Private: 78%
• Medicaid: 64%
• Other: 67%
• Uninsured: 45%

In a country as rich as the US, there's really no excuse for this. Every kid should have easy access to every vaccine.

Judge Terry Doughty's judicial handiwork is having an immediate impact. He must be feeling so proud:

One day after a Louisiana federal judge set limits on the Biden administration's communications with tech firms, the State Department canceled its regular meeting Wednesday with Facebook officials to discuss 2024 election preparations and hacking threats.

....The canceled meetings show that the injunction is affecting government efforts to protect elections....“There is so much wrong with this decision — not least of all that it will make us less secure going into the 2024 elections,” wrote Yoel Roth, the former head of Trust and Safety at Twitter, in a social media post. Roth said the most glaring problem with the decision is that it asserts the companies were “coerced” to remove posts simply because they met with government officials. “That’s just … not how any of this works,” he wrote.

So no more meetings with Facebook and no more warnings of foreign election interference. We are shooting ourselves in our collective butts if we keep this up.

I know this is obvious, but . . .

Everybody seems to have forgotten what Elon Musk's real problem with Twitter is. The wellspring of everything that's happened since last year was a single moment of stupidity in which he agreed to massively overpay for it. And it's not like he doesn't know this. He tried desperately to get out of the deal, and took over the company only after a court forced him to.

At that point he was stuck with a massively unprofitable business, so he laid off half the staff and then embarked on a series of hare-brained schemes to raise revenue. Some of these schemes have been dumber than others, and none of them have even remotely worked, but that's because nothing will work. I'm unable to conceive of any plan that would raise enough money to make Twitter a break-even proposition, let alone profitable. And the only alternative is to pump endless billions of personal dollars into it, not exactly an appealing proposition.

So, sure, Musk is destroying Twitter. But who wouldn't? I don't think there's a human being on earth who could fix Twitter. At best they could be a little less flaky about sending it down the drain, but that's probably all.

The state of Wisconsin allows the governor to eliminate parts of legislation he doesn't like using a line item veto. Today, Gov. Tony Evers vetoed a couple of digits and a hyphen from a two-year increase in the school budget. The original said this:

for the 2023-24 school year and the 2024-25 school year, add $325

After the vetoed digits it said this:

for 2023-2425, add $325

So it's now a 400-year increase. But this can't possibly be legal, can it? Could Evers carefully excise individual letters from a sentence to change, say, this:

no amount over $5,000 may be allocated above the governor's recommendation as fully authorized by appropriation

to this:

move $5,000 to the governor daily

That's ridiculous, right? You have to veto a provision, or at least a clause, not a bunch of individual letters. What's going on here?

UPDATE: It turns out this is not only legal but common. It makes no sense, but the story is here.

Who has the most new solar capacity? Obviously big states do. But what if you adjust for population and look at solar module shipments per capita?

It turns out it's mostly still the big states that are solar leaders: California, Texas, Georgia, and Illinois. But Nevada has them all beat by a mile.