Kevin Drum

The Guardian's front page is currently promoting a story that says antibiotic resistance is associated with air pollution—in particular, with small particulates called PM2.5, which are produced mostly by fossil fuels in richer countries and by dust and residential coal burning in poorer ones. Unfortunately, when I clicked through to the study this was based on, I found this:

The misuse and overuse of antibiotics are the main drivers of antibiotic resistance...

This is perfectly reasonable. It's surely the modern, industrial-scale use of broad-spectrum antibiotics that are both the first and main cause of increasing antibiotic resistance. But later on there was this:

We found that the magnitude of the contribution of PM2·5 to aggregate antibiotic resistance is greater than that of antibiotic use...

Now they're saying that air pollution is the primary cause of antibiotic resistance. So which is it?

Data in the study strongly supports the PM2.5 hypothesis. In fact, they report that PM2.5 contributes 10.9% of variation in antibiotic resistance, compared to only 2.4% for antibiotic use.

That's four times as much, which makes this whole result seems unlikely—especially since the study doesn't even hint at an underlying mechanism for the hypothesized PM2.5 dynamic. I'd take this one with a grain of salt for now.

Here is a chart from Semafor that's been making the rounds:

Zounds! A 40-point gap has opened up between Democrats and Republicans over the past two decades. What happened?!?

To see the simple and obvious answer, all you have to do is look at more than three data points:

Republicans generally thought the federal government was OK when George W. Bush was president. Then they suddenly decided it had too much power when Barack Obama was elected. They changed their minds when Donald Trump was elected. Then the federal government suddenly became too powerful again when Joe Biden took office. Democrats display the same trend in reverse.

Moral of the story: stop trying to deceive us with charts based on transparently ridiculous cherry picking. Alternatively, if you genuinely didn't realize this was cherry picking, please find another job.

Here is your super short guide to tomorrow's election in Ohio:

December: Ohio abolishes August elections because they are wasteful and attract low turnouts.

February: Two pro-choice groups submit language for a referendum in November to enshrine abortion rights in the state constitution.

May: In a turnaround, Republicans vote for one last August election this year. After 111 years of referendums needing 50% of the vote to pass, they suddenly decide it's urgent to ask voters to increase this to 60%.

Tuesday's vote is explicitly about making it harder for the November abortion referendum to pass. It almost certainly has majority support in the state, so Republicans are looking for a way to allow a minority to keep abortion restrictions in place.

If they win, and the abortion measure then loses, don't be surprised if Republicans submit a new referendum to change things back to 50%. That's how they roll these days.

This is the altar at the Basilica of Sacré-Coeur in the Montmartre district of Paris.

The Washington Post says that American companies are trying to reduce their exposure to China. The result is that we're buying less and less stuff from them:

U.S. companies are accelerating efforts to reduce their dependence upon Chinese suppliers, even as officials in Washington and Beijing labor to put a floor under their sour relationship.

Through the first five months of this year, U.S. imports from China were down 24 percent from the same period one year ago, according to the Census Bureau....Foreign investors, meanwhile, are building fewer new Chinese factories, suggesting that other Asian countries will keep increasing their share of U.S. imports at China’s expense.

This stuff is like catnip to me. Usually about five minutes of basic research suggests that these kind of trend stories are mostly nonsense. But this time was different. I checked out import levels and it turns out the Post is right:

In mid-2022, imports from China had grown at roughly the same rate at imports in general. But from August to the start of this year Chinese imports plummeted even as imports from everywhere else stayed steady.

So yes, imports from China are indeed down. This still leaves a bit of a mystery, though: what happened? That is, what happened specifically in the second half of 2022 to account for a sudden and severe decline in imports from China? Donald Trump raised tariffs years ago. Joe Biden announced a ban on selling advanced chips to China last year, but there's no reason that would affect imports. And I don't recall China doing anything in particular to stifle exports to the US. The Post says that various actions on both sides have "chilled commercial ties," but that doesn't really mean anything.

Perhaps this is all a response to pandemic supply chain problems. Companies are starting to reshore their operations outside of China, but that takes a while and it's only showing up now. Maybe. Or it could just be noise in a series that's always been fairly volatile.

Ideas?

Yesterday I wrote about new regulations limiting sulphur content in the bunker fuel used by oceangoing vessels. I concluded that the effect on climate change of these limits was minuscule, but today Alex Tabarrok links to an article in Science that suggests the effect might have been more substantial.

But there are reasons for skepticism. On average, solar radiation amounts to about 1000 watts per square meter. According to a study quoted in the article, this compares to an increase of 0.1 w/m² due to the sulphur limits. That's not a lot.

The 1991 Pinatubo volcano, for example, reduced solar radiation by about 4 w/m², which led to a global temperature drop of 0.5°C. This suggests that a change of 0.1 w/m² would produce a global increase of about 0.012°C, a very small amount.¹

But what about the North Atlantic specifically, where ship traffic is heavy? The study estimates that the effect on solar radiation is about ten times stronger, which might lead to a sea temperature increase of 0.12°C—though this is probably an upper limit thanks to the inertia of ocean temps. This compares to a recorded increase of 0.91°C so far this year. In other words, the sulphur limits are responsible for at most one-eighth of the total warming.

There are other reasons for skepticism. Certain regions of the world are designated as special Emission Control Areas. Sulphur content in these areas was limited to 1.5% in 2005. One such area is the Baltic Sea, but nothing happened there when the new limits took effect:

We're currently seeing a big sea temperature rise in the North Atlantic three years after new sulphur limits took effect, but if you take a look at 2008, which is three years after the ECA limits took effect, sea temperatures start to drop in the Baltic Sea. And in the 17 years since the limits started, Baltic temps have just followed their previous trendline.

Put all this together and there's still little reason to think that the sulphur limits have had much effect on either global warming in general or on North Atlantic sea temperatures specifically. Without all the other factors in play,² it's unlikely we'd even notice anything different.

¹Yes, I'm assuming a linear response. I think that's safe at low levels.

²The eruption last year of an undersea volcano near Tonga; a strong El Niño this year; and an unusual lack of Saharan dust over the ocean.

The New York Times suggests that the war against woke isn't doing any better than the war against Christmas. Even Republican voters are tired of it:

When presented with the choice between two hypothetical Republican candidates, only 24 percent of national Republican voters opted for a “a candidate who focuses on defeating radical ‘woke’ ideology in our schools, media and culture” over “a candidate who focuses on restoring law and order in our streets and at the border.”

Vivek Ramaswamy may be a no-hope candidate for the Republican nomination, but he made his name as an anti-woke doomsayer. That was only two years ago, but today he says he's already moved on:

In an interview, Mr. Ramaswamy said the evolving views of the electorate were important, and he had adapted to them....“At the time I came to be focused on this issue, no one knew what the word was,” he said. “Now that they have caught up, the puck has moved. It’s in my rearview mirror as well.”

Law and order and border security have become stand-ins for “fortitude,” he said, and that is clearly what Republican voters are craving.

Among Republicans, the border is pretty clearly the issue with the most legs. It's not always the top concern of the moment, but year in and year out it's always near the top of the list. And it's at record highs this year:

Donald Trump might not be the smartest guy in the room, but he has a native shrewdness about people. Immigration was his top issue on the day he descended from Trump Tower onto our TV screens, and it's still his top issue today.¹

¹Aside from complaining about how badly everyone treats him, of course.

Here are two measures of consumer confidence, one from an OECD survey and one from the University of Michigan:

Both measures agree that consumer confidence has been rising over the past year. However, it's still way below normal. It's no wonder that the economy hasn't yet helped improve Joe Biden's approval rating.

Paul Krugman has a column today about interest on the national debt. It's going up thanks to the Fed's rate hikes, but he says it's not likely to be a long-term problem. His explanation gets a little complicated, though, so I'd like to simplify it. Here is every significant Fed interest rate hike from the past 60 years:

With one exception, interest rate hikes tend to last 3-4 years before reverting to their previous level—and while you might want to argue with my precise figures, there's no point in that. All I'm saying is simple: the Fed raises rates and the Fed lowers rates. Our current rate hike will likely be over by 2025 or so. And when that happens, interest on the national debt will go down.

Krugman says the same thing with a formula: on average, interest rates tend to be lower than GDP growth, so over the long term they aren't a problem. In our present case, I'd say the longer term is another couple of years, and that's all.

August 3rd: Donald Trump receives a warning from the judge overseeing his arraignment.

August 4th: Donald Trump responds.

A "spokesman" later "clarified" that Trump was referring only to "RINO, China-loving, dishonest special interest groups and Super PACs." Approximately no one believes this, least of all Jack Smith, who's prosecuting Trump:

Federal prosecutors asked a judge on Friday for a protective order limiting what Donald Trump and his team can say about the criminal case against him....Prosecutors’ request for the protective order came a day after Trump pleaded not guilty to charges that he orchestrated a criminal conspiracy to forcibly overturn his 2020 election loss to Joe Biden.

You have to admit that Trump keeps things lively.