Month: November 2023

Private citizens have always been allowed to file discrimination suits regarding voting rights. But no longer, at least in the upper Midwest:

The Eighth U.S. Circuit Court of Appeals, in a 2-to-1 ruling, held that private plaintiffs can’t bring legal claims under Section 2 of the Voting Rights Act.... [The ruling] said the Arkansas branch of the NAACP and another organization couldn’t challenge the district lines drawn for the Arkansas House of Representatives after the 2020 census.

Interesting! And who was responsible for this change of heart?

Circuit Judge David Stras, a Trump appointee writing for the majority.... Joining him in the majority was Circuit Judge Raymond W. Gruender, a George W. Bush appointee.

What a coincidence. A couple of conservative judges overturned decades of precedent to shut down claims of racism from the NAACP. Who would have guessed?

The OpenAI affair just keeps getting weirder. The ringleader of the decision to fire CEO Sam Altman has always seemed to be Ilya Sutskever, one of OpenAI's founders and its chief scientist. But now he has regrets:

More than 500 employees of OpenAI have threatened to leave the company if the current board doesn’t resign and reinstate former Chief Executive Sam Altman and former President Greg Brockman. The threat was made in a letter to the board of directors and among the signees was Ilya Sutskever.

....On Monday morning, Sutskever posted on X trying to repair the damage. “I deeply regret my participation in the board’s actions,” he wrote. “I never intended to harm OpenAI. I love everything we’ve built together and I will do everything I can to reunite the company.”

What a complete shitshow.

According to a study reported by the Wall Street Journal, everyone would like more money. But how much more money? They displayed the answer in absolute dollars, but I converted it to percentages for a clearer look:

Assuming this is meaningful—which I'm not sure of—this is an interesting U shape. People become more satisfied with their incomes all the way to the $100K level, where people are so satisfied they figure paradise would only require a 14% raise. But then, as you get into the top 10% club, it goes up. The richer you are, the more you need to be happy.

Personally, I think these folks are all thinking too small. I figure it would take a billion dollars to truly make me satisfied, which comes to about a one million percent raise, give or take. And I still probably wouldn't be happy because then everyone would always be hitting me up for stuff. You just can't please some people.

In any case, the moral of the story is that the ideal income is about $125,000. For some reason, at that level people live pretty comfortable lives and don't feel like they need much more. Presumably it's because the next level above this is yachts and ritzy golf clubs, and those are so expensive they aren't worth thinking about. But move up a little and suddenly these big ticket items start to seem achievable.

Or maybe people who earn lots of money are the kind of people who are motivated by money and are always chasing the brass ring. The more they have, the more they want.

What. The. Fuck?

Sam Altman won’t return as CEO of OpenAI, despite efforts by the company’s executives to bring him back, according to co-founder and board director Ilya Sutskever....Emmett Shear, co-founder of Amazon-owned video streaming site Twitch, will take over as interim CEO, Sutskever said.

This is from The Information, but it's all over the place. I guess on Monday we'll find out if the entire staff of OpenAI is going to quit, as gossip had it yesterday. Either way, this is a helluva soap opera. The CEO of Twitch? On an interim basis? This makes even less sense than firing Altman in the first place.

The OpenAI board is truly mind boggling. There's no way any group of ordinary humans would ever have done this.

UPDATE: And the other shoe drops:

Sam Altman and Greg Brockman and many of their former OpenAI colleagues will join Microsoft, the software conglomerate’s chief Satya Nadella said Monday, capping an intense three days of discussion following the sudden exit of the AI startup’s executives.

However, Microsoft says it "remains committed" to its $10 billion partnership with OpenAI.

The Washington Post describes Argentina's recent past:

For only the second time in its history, Argentina has seen 10 years without economic growth. During that decade, poverty rates shot up from 28 percent to more than 40 percent. Now, for the first time ever, even formal workers in Argentina’s economy are below the poverty line. Inflation is nearing 150 percent. The peso has plummeted, prices change nearly weekly, and Argentines are forced to carry around large wads of cash just to buy groceries.

If you think this spells bad news for the party in power, you'd be right. Today, lunatic insurgent candidate Javier Milei—a self-described libertarian anarcho-capitalist—won Argentina's presidential election. Here's what he plans to do:

Wielding chain saws on the campaign trail, the wild-haired Milei vowed to slash public spending in a country heavily dependent on government subsidies. He pledged to dollarize the economy, shut down the central bank and cut the number of government ministries from 18 to eight.... He has branded Pope Francis, an Argentine, an “evil” leftist. Climate change, he says, is a “socialist lie.” He would hold a referendum to undo the three-year-old law that legalized abortion.

Most of the election discourse has revolved around Milei's plan to adopt the US dollar as Argentina's national currency, but Argentina has flirted with dollarization before. In 1989, after inflation hit 5,000%, the peso was pegged to the dollar, which worked well for a while. The peg slipped a decade later and plans were made for full dollarization. But then Argentina fell into a deep depression and plans were shelved. Even the dollar peg was abandoned.

It's easy to see why Argentinians might want to outsource monetary policy to the US, but it's a loser idea. It's always going to fail eventually when the economies of the two countries get too far out of sync. In any case, Argentina has defaulted on its foreign debt three times in the past couple of decades and understandably has a hard time attracting dollars these days. This means its domestic stock of dollars is low—which might not make dollarization literally impossible but would sure make it painful and tricky.

There's also the fact that Peronists still control Congress, which presumably means that Milei won't be able to simply bulldoze his ideas into law. Big talk doesn't necessarily translate into big action. In that sense, he might be even more Trumplike than people think.

During the Republican debate a couple of weeks ago, I had a question about a peculiar commercial halfway through: what is "Biden's whale EZ pass lane"? Today, Politico finally—finally!—answers me:

A new television commercial from a GOP-linked conservative group accuses Biden of raising gasoline prices by creating an “E-ZPass lane for whales in the Gulf.” The ad aligns with GOP lawmakers’ attempts to block the administration from limiting oil and gas exploration in an endangered whale’s habitat in the Gulf of Mexico.

OK. It turns out that a species known as the Rice's whale was discovered two years ago in the Gulf of Mexico. However, there were so few that it was immediately designated an endangered species. In August, to protect the whale, the Biden administration removed six million acres from a sale of oil leases in the gulf and also proposed restrictions on oil vessels in the area. A bunch of the usual suspects instantly sued them and the case is now in court.

Fine. But why did anyone think it was a good idea to call this an "EZ pass lane" for whales? That's a bizarrely nonsensical in-joke. Much better is this:

[The] ad debuted the same day Republicans in the House Natural Resources Committee approved Louisiana Republican Rep. Garret Graves‘ “Requiring Integrity in Conservation Efforts (RICE’s Whale) Act,” H.R. 6008, which aims to undo the administration’s efforts to protect the species from offshore oil and natural gas drilling.

Now we're talking. More here if you're interested.

The Office of Management and Budget issued a new set of guidelines for regulatory analysis a couple of weeks ago, and it's surprisingly important. It's also pretty technical, so hang on.

The rules are set out in Circular A-4, one of OMB's 136 circulars, and the first thing it addresses is the discount rate that agencies should use. This is a measure of how important the future is. A high discount rate means you're discounting the future and mostly focusing on immediate effects. A low discount rate means you consider future effects important.

To some extent, the value you should use is a matter of opinion. But there is a quantitative method of approximating it: take a look at financial markets and figure out the actual discount rate that's predominant when investors buy things like long-term bonds. The OMB did that, and decided that the discount rate had dropped over the past couple of decades from 3% to 2%.

Why is this important? Because small changes in the discount rate can have big effects on regulation. In the area of climate change, for example, a lower discount rate means we're taking future effects more seriously, and that in turn means that regulations now (which have present-day costs) should be stiffer. Here, for example, is an estimate of how much we should tax carbon based on the future effects of global warming:

A discount rate of 3% implies that the social cost of carbon is only $80 per ton. A discount rate of 2% implies a cost of $185. The OMB's new 2% default rate means that cost-benefit analysis is more likely to come out positive for regulations on things like carbon and pollution, which have effects mostly in the future.

The second thing that Circular A-4 does is something I wrote about a few months ago when it was in draft form: it changes the way we treat different groups when we do cost-benefit analysis. In a nutshell, the new treatment assumes that a dollar is worth more to a poor person than a billionaire. So if some new regulation costs a rich person ten dollars but benefits a poor person ten dollars, it will now pass a cost-benefit analysis because that ten bucks means a lot more to the poor person than the rich person. Here's how much a dollar is worth to different income groups:

A dollar of benefits to the poorest group is now worth more than $700 in cost to the richest group.

There's more to Circular A-4 (it's 93 pages long) but these are the two most important. Between them, they make it much more likely that cost-benefit analyses coming out of federal agencies will be both fairer and more reasonable.

The Washington Post reports today on the proliferation of junk fees:

The fees together may cost Americans at least $64 billion annually, according to a rough White House estimate.

This may sound like a lot, but it actually amounts to only 0.34% of all consumer spending.

However, that understates things because there are huge swaths of life that have no fees at all: grocery shopping, gasoline, medical care, and so forth. Junk fees seem to be truly visible in three big sectors: banking, airlines, and ticket brokers.

Airlines earn about 10% of their revenue from "ancillary services." Ticket vendors add an average of 27% in "convenience fees." The banking industry charges total fees equal to about 13% of operating revenue.

The Biden administration's approach has been to cap fees in some cases (credit card late fees, for example) and to demand transparency in others (like airline baggage fees). Or, rather, their approach is to try to do this, since they've been far from universally successful so far. As the Post article documents, vendors are fighting fee restrictions like crazed weasels, promising (at the least) to tie this stuff up in court for years. Don't expect a lot of relief anytime soon.

Here's a look at online shopping:

E-commerce spiked during the pandemic but then subsided until mid-2022. Since then it's been growing again. It looks like the pandemic gave e-commerce a permanent 2% bump relative to its long-term growth rate.

I haven't posted an update on my M-protein level for a while, which I know is the highlight of the month for many of you. The reason is that I don't know what it is.

You may recall that in September I got no M-protein number and my doctor told me it was because the lab doesn't report anything if it's zero. Hooray! But no, a couple of days later the lab did report a result: 0.12.

In October it happened again, and this time I never got a number. However, my doctor was on vacation so I figured I'd just wait until the next month for a result and not worry about it.

Now it's November, and once again I got no M-protein number. But then I noticed something on a different test called a serum immunofixation. In the notes it said:

Huh. Coelution. What's that? It is, as you'd guess, when two things elute at the same time. To elute something is to extract it from something else by washing with a solvent. In my case, apparently, the M-protein and something else are being extracted simultaneously and can't be resolved, so there's no way to measure either of them.

Generally speaking, this is probably good news. It means that the amount of M-protein is less than the resolution of the test, which in turn means it must be pretty low. It's not zero, because the test also says it still detects IgG Kappa monoclonal gammopathy, which is a fancy term for the precise version of multiple myeloma that I have. But at least it's low.