Kevin Drum

Over at Vox they're writing about "A Crisis on American Roads." Part 1 is mostly about a mother whose daughter was killed in a crosswalk in Washington DC by the driver of a van. The rest of the piece claims that the increase in vans and big SUVs is responsible for our recent increase in pedestrian deaths.

I don't find that implausible at all, but I nonetheless wanted some evidence. Finally, after plowing through 2,000 words (!) I got it. But my spidey sense began to tingle when I saw this:

There are two weird things here. First, were SUVs really only 10% of all vehicles on the road in 2000? That seems awfully low compared to my memory of that period. Second, there's no line even trying to compare the growth rate of SUVs to the growth of pedestrian deaths. That seems like an odd omission.

You know the rest of the story: I decided to collect the best statistics I could find and draw a similar chart. See the footnote for details.¹ Here's what I came up with:

What you see here is, first, a huge drop in pedestrian fatalities from 1990-2010. (The decline actually began around 1970, but I didn't use those figures in the chart.) Second, there's a slow rise from 2010 to the present.

The share of big cars, however, shows nothing like this. It just goes up slowly and steadily for the entire period from 1990-2020. There's really no correlation here.

Why did pedestrian deaths go down from 1970-2010? Safety enhancement in cars certainly played a role in reducing fatalities among drivers and passengers, but not among pedestrians. Nor has anything new happened since 2010 that would cause pedestrian fatalities to go up significantly (about 50% over a single decade).

None of this is positive proof that big cars aren't at fault in recently rising pedestrian deaths. But it does mean we should be skeptical and demand better evidence. The Vox article provides two studies on the subject. The first is 20 years old and shows that light trucks and vans are more dangerous than passenger cars. No argument there. I think we can all agree on that much.

The second study is very recent and compares pedestrian death rates in different cities. It concludes, "If all [emphasis mine] light trucks were replaced with cars, over 8,000 pedestrian deaths would have been averted between 2000 and 2019." That's about 400 deaths per year. In a real-world situation in which we didn't eliminate every single light truck on the road, but instead produced a 12% reduction in order to return to the 2000 average, this comes to about 50 deaths per year, or a reduction of around 1%. That's not a meaningful result.

So: are big SUVs and light trucks responsible for our tsunami of pedestrian deaths? I doubt it. But pedestrian fatalities did suddenly spike up starting around 2010. Something else must have changed then, but it's not obvious to me what it is. However, I do have a guess:

This is not great data. It comes from the National Household Travel Survey, which is conducted about once a decade. It doesn't tell us total miles walked or anything useful like that, and the fit is not impressive. Still it does show that we're walking more: about 35% more since 2000.² Don't take this as more than a suggestion, but it certainly makes sense that if people are walking more, there are going to be more walking accidents. It might not have anything to do with cars and drivers at all.

Then again, there's also this:

Forget big cars and small cars. Forget about how much people are walking. Instead just look at something plain and simple: the number of pedestrian deaths per million vehicle miles. All vehicle miles. It turns out that our streets have gotten steadily more crowded (by about 20% since 2000), and when you adjust for that crowding the number of pedestrian fatalities has gone down, not up.

Maybe we're overthinking this. Maybe there was no crisis to begin with.

---

¹The NHTSA doesn't track "SUVs." The closest it comes is "Light duty vehicle, long wheel-base." which includes large passenger cars, vans, pickup trucks, and sport/utility vehicles with wheelbases longer than 121 inches.

So that's what I used. Unfortunately, this category was invented in 2007 and produced a huge discontinuity in the data. For example, vehicle miles driven dropped from 1.1 billion in 2006 to 0.6 billion in 2007 because certain vehicle categories were switched into the short-wheelbase group. I had no choice but to do some sketchy interpolation, but I doubt it introduced more than a modest amount of error.

As for the data itself, vehicle miles driven is here. Pedestrian fatalities from 1927-1989 are here. Fatalities from 1990-2005 are here. Fatalities from 2006-2020 are from the FARS database. The middle set of years is from the Governor's Highway Safety Association. The other two are from the NHTSA.

²This is all well and good, but does the amount of walking account for the drop in pedestrian deaths from 1970-2010? Did we walk a lot less during those years? Sadly, we don't have any data for that. For now it remains a mystery.

Last week a restaurant in Virginia refused service to a group called the Family Foundation, which opposes gay marriage and abortion rights:

“We have always refused service to anyone for making our staff uncomfortable or unsafe and this was the driving force behind our decision,” read an Instagram post from Metzger Bar and Butchery....“Many of our staff are women and/or members of the LGBTQ+ community. All of our staff are people with rights who deserve dignity and a safe work environment. We respect our staff’s established rights as humans and strive to create a work environment where they can do their jobs with dignity, comfort and safety.”

Just to rub it in, the restaurant called the Family Foundation and canceled the reservation for their dessert reception 90 minutes before it was scheduled to start.

There's an outside chance that this is a strongly-held conviction on the part of the restaurant and the timing is just a coincidence. But what are the odds? It happened three days before the Supreme Court was scheduled to hear arguments in the case of a website designer who didn't want to design a site for a gay wedding.

So this is most likely intended as a warning shot. You don't want to create websites  for gay couples? Then we don't want to bake desserts for groups that oppose gay couples.

I suppose this was inevitable, and legally I don't know if the case of the restaurant and its desserts is comparable to the website case. But there's no question that it escalates the war over civil rights vs. religious and free speech rights.

I have periodically suggested that liberals could rein in a few policy views in order to gain votes from centrists. I usually get scoffed at, with the implied question being "Who do you want to throw under the bus?" No one, of course, but I confess this whole fiasco is an issue I wish everyone had left alone. If some Christian hardass doesn't want to make a website or bake a cake for a same-sex wedding, then fuck 'em. Just hire someone else. There's no need for random revenge. And it's not as if this kind of provincialism is so widespread that it makes it all but impossible for gay couples to buy a wedding website.

But I'm not gay and I'm not getting married, so that's easy for me to say. I get it. At the same time, it sure seems like it amps up the culture wars a lot in return for only a tiny, symbolic gain.

UPDATE: I initially wrote that the current Supreme Court case was about a baker and a cake. Needless to say, that was a case a few years ago. The case at hand today is about a website designer who doesn't want to create a website for a same-sex wedding.

---

I see that more classified documents have been found at Mar-a-Lago. By itself that's probably not a big deal, but you should read the whole story anyway. It's bonkers:

At one point, as Mr. Trump sought National Archives documents related to the investigation into whether his 2016 campaign conspired with Russian officials, he proposed that his lawyers suggest a trade with the agency: what he sought in exchange for the documents he had.

wtfwtfwtfwtfwtfwtfwtfwtfwtfwtfwtfwtfwtf. Maybe we should just hang Donald Trump upside down by his ankles and shake him to see what comes out.

---

Finally, about that attempted coup in Germany:

Prince Heinrich XIII of Reuss, a descendant of a 700-year-old noble family that once reigned over a tiny state in eastern Germany, was a relatively obscure figure — until Wednesday, when he was named as one of the leaders of a group accused of plotting to overthrow the German government.

So it wasn't just a hapless attempt at a coup, it was a hapless attempt at a monarchist coup. Of course it was. In fairness, though, I prefer weird farces like this to serious attempts at imposing fascism. Prince Heinrich is now my hero.

Today I have another triptych for you. Tomorrow too.

This one is an homage to Orange County's tallest skyscrapers, the twin towers of the Irvine Spectrum. They break the tape measure at 20 stories and 323 feet, outclassing the tallest buildings in such California cities as Fresno, Bakersfield, and San Jose.

The comment section has been overwhelmed with spam lately, so I've added a new spam filter. In addition to Akismet, I'm now running Anti-Spam Bee. We'll see how that goes for a day or two and then reevaluate.

"What does ChatGPT mean for _______ ?" You can fill in the blank with law, news writing, college essays, or a million other things. Generally speaking, though, the answer is nothing. I mean no disrespect to the remarkable output of ChatGPT, but Bob Carpenter gets it right:

I’m getting that same uneasy feeling I felt watching the lifelike movement and reactions of the Boston Dynamics robots. It’s the uncanny valley effect of watching something that’s almost, but not quite fully lifelike.

As good as it is, ChatGPT right now is only a curiosity and a warning. It's a curiosity because even a modest effort exposes it as an idiot savant, full of on-point facts but not really able to draw sophisticated conclusions from them. It's a warning because it's probably only a few years away from having the knowledge and verbal abilities of a PhD student.

How do we respond when that happens? I'm not sure, but I'll say this: we currently live in a world full of lawyers and professors and journalists who are able to calmly accept the prospect of millions of unemployed truck drivers when AI fills the world with self-driving trucks. But they will probably be a wee bit more upset at the prospect of millions of unemployed lawyers, professors, and journalists.

Maybe this is a good thing. The only way we'll get a serious response to AI is if either (a) it affects the working class in numbers so big it creates riots, or (b) it reduces the income of the ruling class by 1% or so. Both would be considered problems of about the same magnitude and would provoke roughly the same energy toward finding a solution.

Take this as you will, but not one word of this post is a joke.

Today Democrats won a 51st Senate seat and the Trump Organization was found guilty of tax fraud. That's a pretty good day! I wonder what Wednesday will bring?

The book is Valley of the Dolls. The year is 1945. Theatrical attorney Henry Bellamy wants his secretary to find an apartment for a younger colleague returning from the war:

"Anne will come up with something," Henry insisted. "Try for the East Side. Living room, bedroom, bath, and kitchen, furnished, around a hundred and fifty a month. Go to one-seventy-five if you have to."

This is a stretch. No one thinks our heroine will be able to find anything at that price—and she doesn't. There are a lot of soldiers returning from the war, after all. In reality, this kind of apartment probably runs $200 or more.

In today's dollars that comes to about $3,300. "East Side" is a little vague, and furnished apartments are no longer common, so let's search Zillow for one-bedroom unfurnished apartments on the Upper East Side that rent for less than $3,000:

I don't live in New York and maybe I'm being badly fooled by Zillow. So help me out, New Yorkers. Can you rent a one-bedroom apartment on the East Side for prices like this? Or is Zillow way off base?

Today you get a triptych of honeybees. The one at the bottom was taken with flash.

Chris Hayes thinks we've all calmed down a bit this year:

There are *some* ways of measuring this. For instance I'm curious what airline data says about number of disruptive passenger incidents this year v last year.

— Chris Hayes (@chrislhayes) December 6, 2022

I am officially bored today, so here are the numbers:

The numbers for 2022 are extrapolated in different ways,¹ but I think it provides approximately the right answer. The absolute number of unruly passenger incidents is also down.

But not down by much! We are still way short of the halcyon years of 2000-19, when nearly all American adults were apparently able to sit still for a few hours without erupting in fury.

¹Unruly passenger data is available through October, so I multiplied by 1.2 to get an annual figure. This may be wrong if, say, the holiday season produces more angry people than the rest of the year. Passenger miles data was extrapolated using the linear trend through August, the most recent month available.

Another admission from the ex-president, who apparently thinks he can take government records because of something to do with @BillClinton’s cat (who denied any wrongdoing) pic.twitter.com/raF0t1NhBA

— Andrew Feinberg (@AndrewFeinberg) December 6, 2022