Month: April 2021

Congratulations, Mr. President:

President Biden will withdraw all American troops from Afghanistan over the coming months, people familiar with the plans said, completing the military exit by the 20th anniversary of the Sept. 11, 2001, attacks that first drew the United States into its longest war.

....Biden’s decision comes after an administration review of U.S. options in Afghanistan, where U.S.-midwived peace talks have failed to advance as hoped and the Taliban remains a potent force despite two decades of effort by the United States to defeat the militants and establish stable, democratic governance.

....“This is the immediate, practical reality that our policy review discovered,” the person familiar with the deliberations said. “If we break the May 1st deadline negotiated by the previous administration with no clear plan to exit, we will be back at war with the Taliban, and that was not something President Biden believed was in the national interest.”

“We’re going to zero troops by September.”

There has never been a good solution to the problem of Afghanistan. For whatever reason—and we will be studying it for years—we cannot defeat the Taliban using only the resources the American public will tolerate. It's even possible we can't defeat the Taliban, period. A stalemate is the best we can do, and a stalemate will last forever since, after 20 years, it's obvious that the establishment Afghan regime will never be able to produce either a consensus government or a standalone military capable of standing up to the Taliban.

This has been clear for a long time. Obama knew it. Trump knew it. But neither had the courage of their convictions. If the US pulls out completely, the Taliban will overrun Afghanistan in a year or so. In other words, the United States will have definitively lost a war it spent 20 years fighting. No president is willing to be the guy who approved that.

But now Biden says he's going to do it. If he follows through on this, it will be a mark of singular courage. He sure as hell has my support for it.

Nate Silver tweets about something that's been making the rounds on Twitter this morning:

Interesting that people in the public health sphere seem to be pretty agnostic on what effect the J&J pause will have on vaccine hesitancy (maybe leaning toward it being a slight positive) while the sentiment of people who cover politics/media is toward it being a clear negative. https://t.co/ZiUsDvj2lD

— Nate Silver (@NateSilver538) April 13, 2021

Basically, the question is what effect the CDC's halt of the J&J vaccine will have on the public. There are two options:

• It's bad because it will produce an increase in the general distrust of vaccines.
• It's good because it sends the message that the CDC is sensitive to even tiny risks and will only approve vaccines that are absolutely safe.

I'm in the first camp, but the second one absolutely has merit. The problem is that there doesn't seem to be any actual evidence either way. We're all just guessing.

For what it's worth, here's how my thought process goes:

1. First, keep in mind that "the public" is mostly very average and non-rational and doesn't follow the news closely. They are not NOT NOT like you and your friends and the folks you follow on Twitter.
2. The CDC's announcement will probably have no effect on people who are already vaccine boosters. They'll just shrug and figure it's no big deal.
3. Nor will it have an effect on confirmed vaccine skeptics. It will give them one more reason to be skeptics, but that's all.
4. The effect will be solely on people at the margins. Keeping point #1 firmly in mind, they will not think hard about this. They will not read newspaper articles about it. At most they'll see headlines or hear brief radio reports or something like that. Their only takeaway will be a vague version of "vaccine has problems."
5. In a few days, the CDC will presumably declare the J&J vaccine safe. Or maybe safe except for middle-aged women. Or something.
6. What will the marginal person think of this? My guess is that they'll remember only that there's been some "controversy" about whether vaccines are safe. On the margins, then, a few of them will move toward the vaccine skeptic camp.

Maybe I'm wrong! But in the political world, at least, there's a lot of experience demonstrating that all you have to do is set up something as "controversial" and most people will never remember anything else.

How could we test this? Survey evidence about vaccine hesitancy might be useful, but it would have to be very precise. What's more, if anyone is doing this the results won't be available for months.

When the J&J vaccine is re-approved, we could measure how many people start saying they don't want it.

Or sometime next week we could deploy a survey that just asks people what they think about the J&J vaccine. Unfortunately, this wouldn't be very useful unless we have a previous survey on the same subject.

Any other ideas?

You will recall (won't you?) my warning from yesterday that the next few months would bring reports of artificially high inflation rates. Today we got the first one, a headline CPI rate of 2.6% in March, up from 1.7% last month.

This uptick is artificial because it's based on a comparison with March 2020, which was super low. I promised to offer a different view, namely a comparison with March 2019 that's then divided by half to get an annual rate. Here it is:

The two-year average is basically a way of smoothing the data, and there are other ways of doing this. However, this quick-and-dirty approach is enough to let you see what's going on. With the exception of the pandemic dip from last year, inflation has been running at about 2% and it still is. There's nothing to be concerned about.

The FDA and the CDC have called for a halt to the Johnson & Johnson single-shot coronavirus vaccine because of six reported cases of a rare blood clot:

All six were women between the ages of 18 and 48 and all developed the illness within one to three weeks of vaccination. One woman died and a second woman in Nebraska has been hospitalized in critical condition....Nearly seven million people in the United States have received Johnson & Johnson shots so far.

The US population of women aged 18-48 is 65 million. About 2,000 have died of COVID-19 over the course of 12 months, or roughly 300 every two months. That's right around 5 per million.

The population of women who have received the J&J vaccine is presumably about 3.5 million. One has died of a blood clot over the two months the vaccine has been available. That's 0.3 per million. If the woman in Nebraska also dies, that goes up to 0.6 per million.

I just don't understand this. Even if it turns out that the J&J vaccine does cause blood clots, the mortality rate is less than a tenth of the mortality rate from COVID-19 itself. Why would you risk undermining public confidence in a vaccine for such a small danger? Especially when firm data will be available in just a few days?

Is there more here than meets the eye? Something I don't understand about these statistics? Or is it a belief among regulators that any risk, no matter how small, has to be made public immediately lest they be accused of being non-transparent? What's going on here?

Two recent police encounters, the first with Caron Nazario and the second with Daunte Wright, which resulted in his death, have prompted a question: Why do police officers so frequently seem to approach encounters with the public as if they might be their last? Just how dangerous are police encounters, anyway?

There are lots of ways to measure this, but the simplest—and the one with reliable data going back the farthest—is to calculate officer fatalities per contact. Here it is:

With the exception of 9/11 and a mysterious spike in 2007, officer fatalities averaged about 4 per million encounters through 2008. Then it dropped to 3 per million in 2009 and has stayed there ever since.

This counts all fatalities and all encounters with police. Data on traffic stops alone is more difficult to estimate, but a study by Jordan Blair Woods of traffic stops in Florida over a ten-year period puts the number at 0.15 per million for "routine" stops (i.e., those initiated solely because of traffic violations).

Overall, according to data from the Department of Labor, police officers are killed on the job at the rate of 140 per million. This is much higher than the national average of 34 per million, but still places them in only 22nd place, behind airline pilots, roofers, farmers, and crossing guards.

POSTSCRIPT: This is all aggregate data, and it's worth noting that there's a limit to what it can tell us. It can't, for example, tell us how dangerous night stops are. Or stops of men vs. women. Or stops of different kinds of vehicles, or of individuals with outstanding warrants. The aggregate data is worthwhile, but it's not the whole story.

Here’s the officially reported coronavirus death toll through April 12. The raw data from Johns Hopkins is here.

"Replacement theory" has a long and undistinguished pedigree. In its original form, starting about a century ago, it was all about the fear of minority cultures outbreeding dominant cultures. Catholics, for example, had more children than Protestants and would therefore soon take over the country by sheer numbers. Ditto for Eastern European and Black communities.

Needless to say, none of this happened. Despite that—or maybe because of it—it has morphed in recent years into an argument that Hispanics will eventually take over the country because they are outbreeding the white population and swarming into the country illegally. As with the original version of replacement theory, it is mostly the preserve of hardcore white supremacists.

This is what makes Tucker Carlson's recent approval of replacement theory so toxic:

I know that the left and all the little gatekeepers on Twitter become literally hysterical if you use the term "replacement," if you suggest that the Democratic Party is trying to replace the current electorate, the voters now casting ballots, with new people, more obedient voters from the Third World. But they become hysterical because that's what's happening actually. Let's just say it: That's true.

....Everyone wants to make a racial issue out of it. Oh, you know, the white replacement theory? No, no, no. This is a voting right question. I have less political power because they are importing a brand new electorate. Why should I sit back and take that?

Carlson's rant has gotten some attention, but surely not as much as it deserves. What it deserves is for Carlson to have been fired immediately. Carlson knew exactly what kind of history he was drawing on, and he gleefully went ahead and did it anyway.

So why hasn't there been a stronger reaction? Part of the reason, I suppose, is that a lot of people write it off as just more of the usual drivel from Fox News.

But I suspect something else as well. Several years ago I made the argument that using "white supremacy" as an all-purpose synonym for "racism" was a mistake. There's a wide range of racist behavior, and it's useful to be able to make distinctions between, say, a bigoted cop and Adolf Hitler. The former is a racist; the latter is a white supremacist. One is much worse than the other.

Obviously I lost that argument and white supremacy has become firmly ensconced as a routine descriptor for any kind of anti-Black behavior. So now, when Carlson adopts an argument that is really and truly a product of white supremacist culture, the phrase has lost its sting. It sounds as if we're merely accusing him of racism, something that we quite correctly accuse Fox News of all the time. So people just shrug and figure that Fox is gonna Fox and it's not that big a deal.

I'm not pointing this out as some kind of gotcha. It's just an example of something that I think could benefit from a much stronger than usual description. Carlson isn't merely tiptoeing around his usual appeal to white racism, he's explicitly adopting an argument promoted by the likes of the KKK and the Nazi party. We need a word that elicits the kind of outrage this deserves, but we no longer have one.

Here's yet another thing I discovered recently while looking for something else: an index of partisan conflict maintained by the Philadelphia Fed. Why does the Fed track this? I don't know. But here it is:

If this index is worth anything, it suggests that partisan conflict was actually on a slight downward slope from the Reagan era all the way through the end of the Bush presidency. When Obama was elected it suddenly shot up, and it shot up again when Trump was elected. It peaked during the first few months of the Trump presidency and then plummeted for the next four years, eventually ending up at the old pre-Obama level.

Needless to say, this is not the conventional wisdom. Those of us who follow politics would guess that partisan conflict rose sharply when Newt Gingrich came to power; rose again during the Clinton impeachment; fell after 9/11; and then rose during the Iraq War. But according to the Philadelphia Fed, none of those things had any impact at all. It was only Obama and Trump who had any lasting effect, with brief surges during the fiscal cliff episode of 2013¹ and the attempt to repeal Obamacare in 2017. And conflict has been calming down for the past few years despite two impeachments, a pandemic, an insurrection, and an endless presidential Twitter stream of outrage.

(In case you're interested, the lowest level of partisan conflict ever recorded was in April 2020, when everyone sang kumbaya and passed the first COVID rescue bill. But it didn't last.)

Anyway, this index is based on the frequency of newspaper articles reporting disagreement in a given month. Is this legit? Beats me. But it's certainly an interesting pushback to the conventional wisdom about the era from 1984-2008.

¹This is only barely visible on my smoothed version of the chart, but it's quite noticeable on the Fed's version.

With spring in full bloom, our oriole has made a triumphant return to the front garden. This year I got a much better picture of him, thanks to a tripod, a remote shutter release, and a couple of days of obsessive watching.

Our orioles (there's also a female around) like to drink the sugar water from our hummingbird feeder. That means my camera was pointed at the feeder and got lots of shots of our hummingbirds as well. I've included one as a bonus photo today.

From the Wall Street Journal:

With Economy Poised for Best Growth Since 1983, Inflation Lurks

That damn inflation! It's always lurking.

But as long as we're on the subject of inflation, something is about to happen that you'll probably hear about many, many times over the next few months. So you might as well hear about it now.

It's this: thanks to unusually low inflation at the start of the pandemic last year, the headline inflation rate for the next few months is going to be unusually high. This is not because inflation is actually high. It's just a mechanical result of the arithmetic. It's as if you were comparing your electricity usage to a year ago when you were on vacation for a week. It would look like your usage was up 25%, but it's not. It's just a mechanical result of the previous year being artificially low.

Here's a chart that will either confuse you or make things clearer. Inflation is actually calculated as an index in which the average price between 1982-84 is set to 100. Here's what the CPI index looks like over the past few years:

The point here is simple: Even if the inflation index in April 2021 is precisely on the trendline of 1.6% growth, it will artificially look higher because the year-over-year comparison is with April 2020, which is in the red dip.

There's not much to do about this except to ignore the inflation numbers for the next few months. One alternative, I suppose, is to compare the current index to the one two years ago, and then divide by two to get an annual rate. That's unofficial, but probably gives a better idea of what the real inflation rate is. I'll try to remember to report that whenever I post about inflation.

One other thing: all the various forecast metrics suggest that inflationary expectations are still restrained. There's bound to be some upward pressure as the economy opens up this year, but nothing substantial.