Month: March 2023

Here's the latest from the IPCC on our progress toward reining in climate change:

Bottom line: Current policies "lead to warming of 3.2°C." That's about 6°F for real Americans who don't use the metric system.

I sometimes think that this whole business of using temperature rises to describe climate change is misguided. In the US, of course, you have the problem that most people see 3° and just vaguely think it's 3°F, which doesn't seem like much. So summers used to 78° and now they're going to be 81°. So what?

Of course, I'm not sure what would be better as a single, all-around metric. Grain production? Acres of land flooded? Expected heat deaths?

We lost over a million people in the COVID-19 pandemic and the biggest effect seems to have been a huge national temper tantrum over mask wearing. Do you think a country like that is going to be even slightly moved by the prospect of 80,000 extra deaths several decades from now?

I'm stumped. If we could figure out a way to correlate temperature with gun availability or number of abortions, maybe we could get somewhere.

The Fed is fighting back. It's been criticized for not supervising Silicon Valley Bank properly, so now it's leaking the news that it did indeed supervise them:

In 2021, a Fed review of the growing bank found serious weaknesses in how it was handling key risks....But the bank did not fix its vulnerabilities. By July 2022, Silicon Valley Bank was in a full supervisory review — getting a more careful look — and was ultimately rated deficient for governance and controls. It was placed under a set of restrictions that prevented it from growing through acquisitions. Last autumn, staff members from the San Francisco Fed met with senior leaders at the firm to talk about their ability to gain access to enough cash in a crisis and possible exposure to losses as interest rates rose.

It became clear to the Fed that the firm was using bad models to determine how its business would fare as the central bank raised rates: Its leaders were assuming that higher interest revenue would substantially help their financial situation as rates went up, but that was out of step with reality.

This raises as many questions as it answers. Bloomberg, which reported this story first, says the Fed initially called out "operations and technology" problems. It didn't point out problems with interest rate risks until late last year.

So I remain puzzled. If the Fed really was up in SVB's grill, how is it that SVB was able to ignore them? And given SVB's capital and leverage positions, what exactly was the Fed worried about? In the end, even with deposits flowing out at a fairly heavy rate, SVB was able to raise $20 billion—which should have been plenty. It's true they had to take a substantial loss on the sale, but not one of life threatening size.

And beyond that, although the Fed is saying that it warned SVB of problems, there's no hint in either of these stories to suggest what the Fed wanted SVB to do about it. Nothing very dramatic, apparently. Just how concerned was the Fed about these problems, really?

More to come, I imagine.

Brad DeLong comments this weekend on a couple of essays that lament the difficulty of living in an era where technological progress is improving exponentially:

I read these, and I think: this is how it has been since the 1870s. Perhaps this is how it has been since the late 1830s.

Well, yes. But there's a difference. Back in 2013, in Part 1 of my robot trilogy,¹ I illustrated exponential growth using the analogy of filling up Lake Michigan. In the first year, you add one drop of water. Then two. Then four.

In the second to last year, you add a quintillion gallons of water. Then two quintillion. Then four. For five or six decades you can barely see anything more than a bit of sogginess. Then, in the final decade, suddenly the job is done and the lake is filled.

Both ends of the chart are exponential growth. But to human beings, the right edge is very, very different from the left. Exponential growth is much more impressive when it produces an elephant than when it produces a flea, even if, technically speaking, both are equally miraculous.

So go ahead and feel more disoriented than your great-great grandfather! After all, here's what Brad has to say about the wonders of ChatGPT:

The “entity” we are conversing with is very sub-Turing indeed. It is page-level autocomplete.

....I do anticipate the use of this page-level autocomplete will spread around the world, considerably faster than use of personal computers or use of the Internet. We are primed to have conversations....A kind of animal that attributes Turing-class intelligence to the lightning—which we, at least some of us, personified as a huge guy with a red beard and severe anger management problems—is primed to make much use of page-level autocomplete.

....Will we use it for good or ill? That is in our hands.

Quite so.

¹Part 2 is here. Part 3 is still in the works. Maybe next year?

The Washington Post has a story today about the demise of ER physicians. It used to be a coveted position for residencies, but now senior doctors are warning against it:

They warn of burnout after covid and patients’ increasing suspicion of doctors. The pay is not as good, they say, especially as hospitals rely more on nurse practitioners and physician assistants to staff emergency departments. And job prospects may be grim, they caution, as emergency medicine residency programs aggressively expanded in recent years.

....Emergency departments are under strain as they become congested with patients waiting for beds, veteran providers quit and violence against the remaining staff grows. These factors are damaging the emergency room’s reputation as an ideal place to learn by caring for a steady stream of patients with a wide range of problems.

Every year, graduating students apply for residencies and are matched with programs that are interested in hiring them. Here's what that looks like for emergency medicine:

Emergency medicine was in the SOAP in 2023. That is, there weren't enough applicants for all the open positions, which means that some ER residency programs had to hire doctors from the Supplemental Offer and Acceptance Program, a sort of second-round draft for everyone who didn't get an offer from the first round of matching.

Of course, it's worth noting that, in the end, every ER position was filled. And every specialty has its ups and downs. Perhaps next year, with ER departments back to semi-normal, we'll see a rebound.

You're all probably tired of hearing me natter on about Silicon Valley Bank, but the conventional narrative continues to bug me.

The most persistent part of the narrative is that SVB had a bond portfolio exposed to lots of interest rate risk and they were morons for not hedging this risk. But normally, you hedge interest rate risk only if it's an extraneous factor to your main business and you don't feel like worrying about it. You don't hedge it if it's a deliberate investment strategy. What's the point?

So of course SVB didn't hedge their interest rate risk. If they wanted less exposure to interest rates, they would have put together a different portfolio in the first place.

What's more, they did hedge their short-term interest rate risk in the strongest possible way: by making sure that short-term losses wouldn't land on their books at all. Their $90 billion bond portfolio was deliberately marked as Hold to Maturity, which means that gains and losses only show up when the bonds are redeemed. That's several years away at a minimum.

Basically, SVB was (a) betting that interest rates would be low in five or ten years and (b) not betting at all on interest rates over the next two or three years. That may or may not be the smartest investment strategy, but it's perfectly reasonable.

Within the investor community, SVB's unrealized losses caused a bit of nervousness but not a lot. SVB collapsed not because of that, but because clients pulled out their money en masse. The proximate causes of this were (a) a Moody's downgrade, (b) a massive miscalculation by Goldman Sachs, and (c) a decision by Founders Fund to pull all their companies' money out of SVB. None of this was substantially related to SVB's unrealized losses.

Tom Nichols is unimpressed with today's story about John Connally's 1980 mission to let the Iranians know they should wait a while before releasing the embassy hostages:

Has it occurred to anyone that the Iranians hated Carter's guts because of the Shah and did what they did just to spite him? Of *course* they were trying to hurt Carter; they didn't need John Connally's help with that. https://t.co/X3Lv6cABfn

— Tom Nichols (@RadioFreeTom) March 18, 2023

This has occurred to everyone. In retrospect, we all know that Khomeini was vanishingly unlikely to ever release the hostages while Jimmy Carter was president.

But that doesn't matter, because we didn't know it back then. Nor does it matter if Connally was an idiot. Or badly informed. Or whether his message ever got through.¹

What matters is that he tried. He tried to make sure the hostages would stay in captivity as long as there was any chance their release might help Carter. This is appalling beyond belief.

But did Reagan know about any of this? We don't have any firm evidence. But we already know there was a parallel effort underway with the same goal. And we know the hostages were released within minutes after Reagan was sworn in. And we know that arms began flowing from Israel to Iran a few days later, something that could happen only with a US blessing.

This fact pattern already suggests that Reagan cut a deal with the Iranians. The Connally story adds yet more evidence. And of course, we also know that the Reagan campaign was terrified Carter might negotiate a hostage release before the election. And we also know that Reagan was perfectly happy to trade arms for hostages later in his presidency.

Given all this, you'd have to be willfully obtuse to continue thinking it unlikely that Reagan—or at least his senior staff—knew what was going on. It's true that we don't have silver bullet proof. But come on.

¹Well, it matters a little bit because it let the Iranians know that Reagan was anxious for a deal—which meant they could trick Reagan into offering more than he had to. In particular, they took advantage of this to get arms in return for releasing the hostages, something Carter was unwilling to consider.

As we all know, Jimmy Carter lost his 1980 reelection bid against Ronald Reagan thanks in part to the long-running Iranian hostage drama. It's long been suspected that Reagan's team actively tried to persuade Iran not to release the hostages so that Carter wouldn't get a victory bump, but there's never been any firm proof.

Until now. In a story that's getting surprisingly little play, the New York Times reports that Ben Barnes, a major player in Texas politics back in the day, says he accompanied John Connally on a mid-election tour of the Middle East:

What happened next Mr. Barnes has largely kept secret for nearly 43 years. Mr. Connally, he said, took him to one Middle Eastern capital after another that summer, meeting with a host of regional leaders to deliver a blunt message to be passed to Iran: Don’t release the hostages before the election. Mr. Reagan will win and give you a better deal.

Then shortly after returning home, Mr. Barnes said, Mr. Connally reported to William J. Casey, the chairman of Mr. Reagan’s campaign and later director of the Central Intelligence Agency, briefing him about the trip in an airport lounge.

The tour kicked off on July 18. Here's what Gallup polling looked like on that date:¹

Carter plummeted in the polls during the first months of the hostage crisis, but recovered in April and stayed ahead of Reagan for the next couple of months. Reagan opened up a small lead on Carter in July, but it was narrowing. It was precisely the time that the Reagan campaign was probably most worried.²

So that fits. Still, how do we know that Barnes is telling the truth?

Mr. Barnes identified four living people he said he had confided in over the years....All four of them confirmed in recent days that Mr. Barnes shared the story with them years ago....Records at the Lyndon Baines Johnson Library and Museum confirm part of Mr. Barnes’s story. An itinerary found this past week in Mr. Connally’s files indicated that he did, in fact, leave Houston on July 18, 1980, for a trip that would take him to Jordan, Syria, Lebanon, Saudi Arabia, Egypt and Israel before returning to Houston on Aug. 11. Mr. Barnes was listed as accompanying him.

Did the Reagan campaign ask Connally to do this? Did they even know he was doing it?

Mr. Barnes said he was certain the point of Mr. Connally’s trip was to get a message to the Iranians to hold the hostages until after the election. “I’ll go to my grave believing that it was the purpose of the trip,” he said. “It wasn’t freelancing because Casey was so interested in hearing as soon as we got back to the United States.” Mr. Casey, he added, wanted to know whether “they were going to hold the hostages.”

There's no way to know for sure. But we know that Richard Nixon tried to sabotage a Vietnam peace deal before the 1968 election. This would be right in character.

¹I've erased the Republican convention bump in late August in order to provide a clearer view of what the polling trends looked like at the time.

²There's an extensive literature about the 1980 polls and whether they were accurate. But that doesn't really matter here. What matters is what people thought at the time, and the Gallup poll is a pretty reliable indicator of that.

With the release of ChatGPT, the chattering classes began to chatter about the safety of artificial intelligence. Partly this was about things ChatGPT could do right now (help kids cheat in school, for example) but mostly it was about the possibility of AI destroying humanity at some point in the future.

But there's an odd thing about these discussions: they almost all turn on the possibility of future AIs getting mad at us; or formulating goals that (sadly) require wiping the planet of humans; or going nuts and destroying the entire solar system in a rampage.

And sure, maybe. Anything is possible. But all of these things assume that AIs are like humans. They get angry. They have desires. They develop mental illnesses. They fear death and will protect themselves from it.

But none of that is remotely likely. These are things that developed in humans via evolution, and there's no reason for a computer to have any of them. They don't have lizard brains that still control their destinies because they never had lizard brains in the first place.

But you know who does still have lizard brains? Humans. The real danger of AI is not that it will spontaneously or mistakenly destroy us all. The real danger is that humans will deliberately deploy them for our favorite activity: killing each other. If you're really concerned about the safety of AI, this is what to focus on. I'd say it's about 99% of the real danger.

If you want to project the state of the economy, one of the key metrics is personal saving. But this is harder to estimate than it seems. Here's one crack at it:

The dashed line is the "normal" quarterly saving rate and the orange line is the actual saving rate. As you can see, it spiked when the government mailed out checks at the start of the pandemic and then again when Joe Biden's rescue plan passed. Then it went down as the government programs ended and consumers started drawing down their savings accounts. Around mid-2021, actual quarterly saving dropped below the historical trend.

The purple line is just arithmetic: it shows the total amount of savings above normal. This peaks after Biden's bill, and then declines as people spend their savings. All this extra government money is now gone.

Or is it? Here's the same chart, but this time it suggests that total savings are still above normal by about $700 billion. Can you tell what the difference is?

In the bottom chart I changed the normal saving line slightly. That's all. But that slight change was enough to change the amount of excess saving by quite a bit.

The top chart is my best guess because I didn't mess around with the trendline. It's just a simple least-squares fit of the pre-pandemic trend. But "normal" is nonetheless a guess, and the amount of savings still left to be spent is extremely sensitive to this guess.

This is why predictions are hard, especially about the future. I think savings are essentially gone now and this is going to start hitting the economy very soon. But I could easily be wrong enough to put off the day of reckoning for another six months.

And so could anyone else. No matter how sophisticated your technique, you're still stuck trying to predict human behavior based on a projection of "normal" over the course of three years. There's an inherent lack of precision to that, and therefore to your overall economic predictions as well.

We'll know the answer before too much longer. But only in hindsight.

You should check out Kiera Butler's piece at Mother Jones about Sarasota Memorial Hospital in Sarasota, Florida (naturally). A couple of months ago the "medical freedom" movement suddenly decided to target them:

With an unprecedented 300 people at the February meeting, attendees lit into Sarasota Hospital physicians, accusing them of killing Covid patients....That’s when things got really chaotic. Mike Flynn, a resident of Sarasota, former national security advisor to former President Donald Trump, and conspiracy theorist extraordinaire, tweeted after the February board meeting, “I attended this meeting and the Sarasota Memorial Hospital took what could have been a rebuilding of trust and further damaged this institution with a ‘fox inside the henhouse investigation.’” He added, “Their little report is not the end of the investigation. More to follow.”

Dr. Jane Ruby, another live streamer and right-wing political pundit who has a show on Peters’ network, devoted a February episode to Sarasota Memorial, titled “Killing Hospital Begins Cover-Up.” Ruby referred to a doctor at Sarasota Hospital as “an arrogant, condescending piece of crap,” and criticized the staff. “You enabled that hospital to not only murder innocent people but to make millions of dollars doing it,” she thundered. “They’re coming for all of you. You’re all the targets now.” Stew Peters posted the episode to the 270,000 followers on his Telegram channel.

And what was Sarasota Memorial's sin? Apparently they had declined to treat COVID patients with ivermectin or hydroxychloroquine, both of which are still MAGA obsessions. As usual, the frightened MAGAnauts quickly presented an excellent opportunity for conservative hustlers to stoke the flames and make a little money on the side.

Read it all, but do it on an empty stomach.