Skip to content

These are backlit aerial photographs of Los Angeles adorning the entrance to the Metro Center station at the union of the blue, red, purple, and yellow lines. I suppose they're there to remind passengers of what they're avoiding by riding the subway.

February 21, 2022 — Los Angeles, California

Jason Furman reminds me this morning that GDI numbers were released a week ago. Why does it take a month longer to release GDI than GDP? Does anyone know?

Anyway, GDI is Gross Domestic Income, and it's an alternate way of calculating economic output. In theory, GDI and GDP should be identical, but thanks to measurement issues they're usually a little different.¹ Lately, though, they've been way different, and they were way different once again in Q3:

In Q3, GDP came in at 5.2% while GDI was 1.5%. That's a huge difference.

So which one is more accurate? As Furman notes, there are good reasons to think that an average of the two is the most plausible number. In Q3 that was 3.3%, which certainly seems more likely on a gut level. I mean, did summer feel like a season of massive growth? Or a fairly ordinary one? Based on everything else going on (employment, wages, etc.) it seemed fairly ordinary.

¹This is officially known as the "Statistical Discrepancy" because the people who name these things have no imagination.

The stock of news stories asking why Americans are so sour on the economy continues to soar. If we could just figure out a way to give people a nickel for every one of these pieces our problems would be over.

But can I please interrupt all these thousands of words? The answer is simple:

  • Republicans are unhappy about the economy, not "Americans."
  • Sentiment doesn't turn on a dime. Inflation is our #1 economic concern by a mile, and increases in inflation are way more noticeable than slowdowns. Concerns about inflation will eventually abate, but it will probably take until next summer, after rates have been low for about a year.

That's it. That's all there is. Now can we all shut up about this?

An op-ed in today's New York Times discusses an evergreen topic: prescription drug shortages. But how many drug shortages are there? Emily Tucker cites three sources which provide answers of 160, 49, or 37 for the year 2022. Here's the FDA's number:

If we use this number—which is about the same as that reported in Europe—it represents a little less than 0.3% of the 17,000 pharmaceuticals currently cleared by the FDA. I'm reluctant to downplay the severity of shortages for patients who need these drugs, but I wonder realistically if this number can get an awful lot lower?

One thing Tucker mentions is that large hospital chains often get wind of shortages before others. When that happens, they order huge quantities, which just makes the shortage worse for everyone else. Oddly, she doesn't mention this as something that could be addressed, even though it seems quite likely that the relevant regulatory body could issue rules about hoarding fairly easily.

Instead, she wants legislation to enforce higher quality factories; more adaptable factories that can increase production easily; and buffer stocks of essential medications, presumably held by the federal government. I guess that sounds OK, though I wonder how effective it would really be, especially since (a) the FDA already mandates pretty high quality standards and (b) factories that can increase production more easily is just another way of asking for bigger factories that normally run at, say, half their capacity. As for buffer stocks, that sounds doable, though it depends on just how many drugs we're talking about and how stable they are over time.

All this said, it doesn't appear to be the case that the drug shortage problem is getting worse. Nor is it the case that it's been around for two decades with nothing being done about it. Drug shortages have been cut by about two-thirds since 2010.

Rep. Elise Stefanik had three university presidents on the hook today and decided to goad them by asking if calling for the genocide of Jews violated their university's code of conduct or anti-harassment policies. "Yes or no?" she demanded repeatedly, but all three presidents demurred. Instead they replied with various versions of "it's context dependent"—which outraged Stefanik and a bunch of other people too.

There wasn't much of anyone left by the time Stefanik unleashed her outrage, but it was still an opportunity for some good video clips. Campaign season is coming up, after all.

The problem is that Stefanik very deliberately didn't ask the presidents how they personally felt. I think we can all guess that. She asked about their universities' written policies. At Harvard, for example, racial harassment is defined as actions

that demean or abuse another individual or group because of racial or ethnic background. Such actions may include, but are not restricted to, using racial epithets, making racially derogatory remarks, and using racial stereotypes.

The student handbook adds this:

It is important to note here that speech not specifically directed against individuals in a harassing way may be protected by traditional safeguards of free speech, even though the comments may cause considerable discomfort or concern to others in the community.

Calling for genocide certainly falls within the "racially derogatory" bucket, but like it or not, it's factually correct to testify that it's protected as free speech unless it's directed at an individual in a harassing way. In other words, it's context dependent.

The presidents were not waffling, nor being morally bankrupt. They surely all think that calling for Jewish genocide is ugly and antisemitic. But it's also protected speech in some circumstances, so that's what they said. It was the correct answer.

Bad news today:

Californians passed a bond measure for HSR between Los Angeles and San Francisco in 2008. It is now 15 years later and we aren't even close to finishing the first little segment. In fact, let's review a few facts and figures from the latest CAHSR business plan:

  • The initial "train to nowhere," a 171 mile segment from Merced to Bakersfield, is still seven years away even if the latest schedule is met. It will supposedly be completed by 2030 after more than a decade of work and attract 12 million riders per year. Between Merced and Bakersfield!
  • The other 350 miles linking San Francisco to Los Angeles will take only three more years. You betcha.
  • By 2035 the authority will be running 120 trains daily between the Bay Area and Greater LA. Assuming an 18-hour day, that's one train every nine minutes. You betcha.
  • Ridership will clock in at 32 million per year compared to about 12 million who fly between Greater LA and the Bay Area today. That's 800 passengers per train and 90,000 passengers per day. You betcha.
  • The authority continues to claim that the full LA-SF trip will take 2 hours and 40 minutes, even though this would require continuous operation at average speeds higher than any HSR in the world. You betcha.

Oh, and it still lacks about $80 billion in needed financing even after today's $3 billion infusion. Hell, even the train to nowhere is still $10 billion short.

What's remarkable is that all of these assumptions are actually more realistic than they used to be. They've gone from pure fantasy to applied fantasy, but they're still fantasy. And now we're sinking another $3 billion into it. Sigh.

Our friends at National Review point us today to Matthew Dickerson's recent review of the federal government's improper payments rate:

The Biden Administration has reported a total of $764 billion in improper payments in just three years The improper payment rate has averaged 5.9 percent during President Biden’s administration. The pre-pandemic historical average was a 4.23 percent improper payment rate. If the Biden Administration had simply kept improper payments at the historical norm, improper payments would have been $210 billion lower over the last three years.

This finger-pointing is misguided. The spike in improper payments was largely due to pandemic-era programs that were hastily and sloppily put together by the Trump administration. What's more, as Dickerson himself points out, improper payments from the unemployment expansion were accounted for in 2023 even though they were actually paid out in 2021 and 2022. He also doesn't account for recoveries of improper payments. Here's an apples-to-apples chart:

Following the pandemic spike, the improper payment rate dropped to 3.8%. Over the last 15 years the trend has been almost perfectly flat.

POSTSCRIPT: I have to say that I'm impressed with the speediness of this report. The 2023 fiscal year only ended two months ago and we already have improper payments all tallied up.

But that also prompts a question: If agencies can figure out this quickly that improper payments have been made, why do they make them in the first place?

This is not news, but here's a recent poll finding about the 2020 election. It's for Republicans only:

Among Republican voters who support Trump in the primaries, 86% think he was the rightful winner of the 2020 election and 61% think there was outright fraud from Democrats.

Those are basically Trump cultists. But even among the non-Trump voters, 41% think Trump rightfully won the election. That's a majority of those who expressed an opinion.

For what it's worth, this is one of the answers to the question of why even moderate Republicans still plan to vote for Trump rather than Biden. In their view, it's Democrats who are a threat to democracy, and all Trump is doing is fighting back.

In other news, it looks like Republicans have pretty much given up on Ukraine:

Republicans only like wars that Republicans start. They get tired of wars supported by Democratic presidents pretty quickly, even if, as in this case, it means giving Vladimir Putin whatever he wants.