Sep 29Liked by Nate Silver

One point I would like to see incorporated into this discussion, but is hard because it’s so politicized, is that we should accept that mistakes were made but also be very forgiving of those mistakes because this was an incredibly hard problem that forced us to take our best guess a lot of those times.

Of course, many liberals don’t want to admit mistakes while many conservatives don’t want to forgive mistakes.

Expand full comment

The vaccine was an effective response to the pandemic and is one of the only good things Trump could justifiably take credit for but is the one thing his cult won’t give him credit for.

What a world.

Expand full comment
Sep 29·edited Sep 29

Here's my question/pushback Nate. I seem to recall that the Northeast (and if i'm remembering correctly, Louisiana and Michigan) were hit super hard in the initial weeks of the pandemic and that the high death rates in those mostly blue states, which is the main reason the death rates in 2020 seem to be fairly evenly split across political lines, were for that reason.

Also, if I'm remembering correctly, I think California and Oregon were the first places to have cases but escaped the fate of the Northeast in part because of more aggressive social distancing measures in the first few weeks (although I'm guessing weather played a role too).

Long story short, while I would agree that some of the social distancing measures (most notably the continued school closings into the fall of 2020) went too far, I'm less convinced that social distancing didn't play an important role in minimizing deaths pre-vaccine.

Expand full comment

How does voter age demographics split between Dem and Rep. Given that the vast majority of deaths were in the over 65 population, age certainly should be folded into this. Vaxxed or not, a person under 35 had a much greater than 99% chance of survival.

Expand full comment

I agree with you.

Expand full comment

I'm not at all sure this is a contrarian position across America.

As long as we had neither a vaccine nor any reasonable way to know how bad the pandemic could get, it was reasonable to promote policies to minimize the spread of the disease. The death toll was appalling, and we did not know how the story was going to turn out.

Once the vaccines were available, the odds changed significantly, both for individuals and for the population as a whole.

Ordinary (that is to say not on-line partisans) people that I know, both liberal and conservative, would agree with this in general.

This is not to say that there is no blame to hand out.

One group decided to actively discourage people from doing the safe thing, while the other side went over the top in trying to force people to choose the safe thing. Politicians on both sides of the aisle egged this on, and once it reached a certain point, were at the mercy of highly agitated mobs. Worse, each mob became more strident and unreasonable because of the antics of the opposing mob. All helpfully amplified in the media outlets.

Going forward, I am reasonably confident that next time we would handle a pandemic better, but I have zero confidence that we will do better with the next great polarizing event.

Expand full comment

Absolutely, to me the clear consensus evidence-based position on covid-19 = stay current on your vaccines and move on with your life (unless you have extenuating circumstances, ie severe immunosuppression, live/work in a nursing home, etc). I think masks and testing made sense early on before vaccines, but as you said the data is very murky on how much that helped, if at all. However, once the Delta and Omicron variants came out it was so highly contagious it was probably impossible to ever make a dent in with NPIs

Expand full comment

> Other than maybe the fall of the Soviet Union, COVID has been the most consequential event of my lifetime.

I think 9/11 wins that contest by a mile

Expand full comment

Don't differing characteristics of blue states and red states make it exceptionally misleading to look at only state-level results to assess COVID policies? I'm certainly not an expert in COVID or statistics. But, with identical NPIs, I would think blue states--on average, more urban and dense--would have had far worse outcomes than red states. If that's true, then it would follow that having roughly equally outcomes means that blue states had, on average, quite effective NPIs.

Expand full comment

Some observations on school closures as NPI (you can find the data over five articles at my blog):

1. Schools remained in remote if enough parents wanted it that way. Had nothing to do with CDC.

2. The single biggest factor in parental preference for remote vs in-person was race.

3. The single biggest factor in whether in-person instruction was full-time or hybrid (edited, I originally wrote remote in error) was the state politics. Dem governors pledged allegiance to CDC, GOP governors didn't. CDC guidelines made it impossible to offer full-time in-person.

4. Parents who wanted remote were guaranteed remote for the 2021 school year. Parents who wanted in-person were dependent on the majority parental preference in their district.

5. Since most (over 50% revealed prefernece) non-white parents wanted remote, they were generally pleased with schools, as they were always able to choose remote. Since most (over 75% polled and revealed preference) whites wanted in-person AND most whites go to majority white schools, they were generally pleased with schools as well.

6. Parents in non-majority white districts (most major cities, lots of high immigration suburbs) who wanted remote were generally screwed. This isn't a big group--about 15%, according to polls. The previous three data points suggest these parents were mostly white. They punched way above their weight level, since "mediafolks with kids" live in cities and high immigration suburbs and are mostly white.

So given that fact base, there are two ways of looking at the "damage" caused by school closures (again, important to remember that polls throughout and after showed high degrees of parental satisfaction).

First, there's no question that parents who wanted in-person instruction (regardless of race) were dependent on the will of their neighbors and this was completely unfair. Parents who wanted in-person instruction were not given equal rights (except in Florida and Texas and perhaps some other GOP states). No one ever approached the issue from this viewpoint, in large part because advocates for opening schools simply couldn't comprehend that they weren't speaking for all parents. But the anger levels among parents and non-parents who aren't aware that most parents got what they wanted would have been alleviated if they had been guaranteed in-person instruction. However, this would have been really really expensive and for all the talk of how great Florida was, the parents, kids, and teachers who had to deal with five-ten kids in person and 20 on zoom pretty unanimously say it sucked.

But that wouldn't have solved much anyway, because the pandemic damage was not, in the main, experienced by the kids of the angry white parents denied in-person instruction, but the kids of black and Hispanic who actively demanded remote instruction and indeed, would have preferred to still have their local school in remote in the 21-22 school year. The learning loss wasn't imposed on these parents. They actively chose it. This is why it shouldn't surprise anyone that California actually slightly outscored Florida and Texas on the NAEP (less learning loss).

The only way to end the learning loss was to end the parents' right to choose remote instruction, which is what most state legislatures did in summer 2021, when we lucked into a window of time when it looked like the vaccines would end covid entirely. Had we not had that window, we might still now be dealing with parents able to legally insist on remote education at their local school.

As evidence of the change, consider Lori Lightfoot's reaction to her union demands in January 2021 and January 2022. The first year, the union refused to return to the classroom and she was pissed but *didn't* shut off remote access. Why? Because the overwhelming majority of her parents wanted remote. The next year, when the union refused to go back to school, she shut down remote, to their considerable shock. But she still had a majority of parents who were keeping their kids out of school and demanding remote. The difference: the Illinois legislature had set a mandate on when she could enable remote education. Had she submitted to parent (not union) wishes in 2022, she'd have lost a lot of money for the schools. https://www.reuters.com/world/us/chicago-public-schools-cancel-classes-again-covid-19-teacher-walkout-2022-01-07/

Once the schools closed in March 2020, most states changed the laws around attendence and school funding for a year and most governors--including DeSantis--refused to even consider forcing parents into virtual academies instead of guaranteeing them the right to remote ed at their local school. The next school year, they all flipped and forced parents to choose virtual academies and banned schools from offering remote except in the case of really high case numbers and then only temporarily.

As someone who opposed school closure in March 2020, I had a lot of time to think about this and I'm pretty sure once the schools closed everything we did was baked in. No one anticipated that parents would actively resist sending their kids back to school in demographically unfortunate patterns (well, I did, but then I spend a lot of time thinking about race and education and I'm nobody, anyway.)

I realize this is only marginally ontopic, but since the enormous costs Nate mentioned very much include the learning loss, I thought people might be interested.

Expand full comment

I broadly agree, particularly when it comes to deaths. And certainly blue states went overboard on some NPIs.

But the idea that NPIs had no effect on *infection frequency* doesn't pass the smell test. If I avoid crowds and if my family members isolate when infected, it's hard to believe that won't reduce my infection count by ~30% or more.

Deaths are a saturating non-linear function of infection frequency, so if everyone gets infected at least once, high NPI states could fewer infections but a roughly similar death count.

Not saying you'd claim otherwise, but some people might have that takeaway.

Expand full comment

Why are the deaths overall so much higher in the right column, period. Illinois is good example...they are high pre vaccine, then post vaccine it is basically unchanged. Many states went up in deaths...despite falling in relative state rankings. Something else is surely going on. The largest spike in death happened in the US post vaccine. Why?

Also, were vaccine uptake numbers really that different across states amongst the 65 and older? Most evidence I have seen suggests no. There was high uptake there across states. This was the cohort dying. And they were highly vaccinated.

Expand full comment

While there is a substantial disparity rate among vaccination rates among the young based on political affiliation that disparity is much less pronounced among the elderly.

Put simply the group most at risk from the virus, the old, got vaccinated regardless of how they voted.

Expand full comment

At least one major NPI would have a big effect and require negligible lifestyle costs. That would be to improve air quality by installing HEPA filters or Corsi-Rosenthal boxes in schools and other major indoor public spaces. They've been shown to reduce all airborne disease by ~50%-ish even in real practice. The dollar cost is negligible for the benefit. These are off-the-shelf items. It's a disgrace that major practical steps like this have been lost in the political squabbling.

Expand full comment

1) Comparing COVID-19 death data is not reliable at all, as it has been demonstrated that COVID-19 deaths and all-cause excess mortality does not correlate well: https://twitter.com/orwell2022/status/1701981998200291438

2) There's no significant difference in the divergence of mortality right after the vaccines became available. Only in maybe August/September 2021 it starts to diverge, which is also the point in time when most coercions & mandates started! In my opinion, it's much more likely that many people (at least temporarily) moved to the 'free states', to escape the mandates. I've done so myself, spending many months in the southern states, during the height of the covid totalitarianism in my then home state of Washington.

There's no robust data by vaccination status of comparable cohorts, that allows to conclude that vaccines have saved lives or brought any clinical advantage.

Relying on voter registrations, or 'Trump voters', is not a very reliable scientific approach, as it misses to adjust for underlying confounders such as health status.

Expand full comment

My memory was that for the first few months (March through about June of 2020) urban areas that tended to lean democratic had higher COVID death rates. My interpretation of the data was that urban areas had better transportation connections to the international locations that covid came from, no one had immunity, and initially both blue states and red states locked down. After June 2020, blue states had stricter policies and also lower covid death rates. These effects cancelled out so overall death rates were similar by the time the vaccines came out.

This being said, I agree with Nate's point that the costs of many NPI's (mainly school closures) exceeded the benefits.

Expand full comment