Nate Silver

Nate Silver @NateSilver538

That is, it's probably the combination of social distancing + nontrivial % immunity that's producing the relatively steep declines in NYC and Italy. So those declines might be unrealistic to expect elsewhere in places that apply the same policies but don't have as much immunity. — PolitiTweet.org

Nate Silver @NateSilver538

I do kind of worry that failure to account for this property may also create confusion. If 20% of people in NYC are immune, or 30% in Stockholm, that has a material effect on why the numbers look different in those places than others (e.g. Wash. DC) where immunity might be <5%. — PolitiTweet.org

Nate Silver @NateSilver538

I've seen "herd immunity" used in different ways by different folks. I'm trying to communicate the idea that the more people are immune, the slower the disease will spread, other things held equal. I'll avoid "herd immunity" going forward if something different is coming across. — PolitiTweet.org

Nate Silver @NateSilver538

BTW, Trump has figured this out! By focusing on case counts, the media creates disincentives to do more testing because it makes the numbers look superficially worse. One reason (not the only one) why we're not pushing for testing as much as we should. https://t.co/lDxP8CMueL — PolitiTweet.org

Nate Silver @NateSilver538

New York's serology studies have actually found slightly *lower* rates of COVID-19 among healthcare workers than in the general population. Suggests that PPE and masks help a lot! https://t.co/I6RXC4oIdP — PolitiTweet.org

Nate Silver @NateSilver538

New York's serology studies have actually found slightly *lower* rates of COVID-19 than in the general population. Suggests that PPE and masks help a lot! https://t.co/46kBt4IqTO — PolitiTweet.org

Nate Silver @NateSilver538

@mattyglesias The lags are a big issue, but a 9 week cycle is too long. Maybe 2-4 weeks, with states having more hope of coming in at the shorter end of the range if they're doing more testing (ideally either very widespread or random) and carefully monitoring hospital data. — PolitiTweet.org

Nate Silver @NateSilver538

One thing that might be hard to tease out from the data... but I wonder if the *tone* of messaging matters a lot. "Certain business are opening, but we'd urge you to be very cautious" could produce a different reaction than "Go wild and get that economy revving up, folks!". — PolitiTweet.org

Nate Silver @NateSilver538

Important... implementation and relaxation of formal stay-at-home orders is only loosely related to actual changes in behavior. People engaging in a bit more economic activity everywhere (maybe warm weather/social distancing fatigue) but not much more so in states to "open" up. — PolitiTweet.org

Nate Silver @NateSilver538

That doesn't mean it's just making stuff up or engaging in fake news. On the contrary, the facts it relays are generally accurate in isolation. But the problems are in how facts are strung together and emphasized. Often there are sins of omission (e.g. no context on testing). — PolitiTweet.org

Nate Silver @NateSilver538

Not providing context on the increase in testing is such a basic error, and has been so widespread, that it's revealing about the media's goals. It's more interested in telling plausibly-true stories ("narratives") that sound smart to its audience than in accuracy/truth per se. — PolitiTweet.org

Nate Silver @NateSilver538

@jbouie I think even Mnuchin believes this, which is why he wrote "what have you done for the country LATELY" (emphasis mine). — PolitiTweet.org

Nate Silver @NateSilver538

There are uncertainties around immunity especially in the long run but people's priors have been screwed up by bad media coverage and poor messaging. Here's a good overview that doesn't sugarcoat the messy details: https://t.co/VKIyTMK2Eu — PolitiTweet.org

Nate Silver @NateSilver538

I think Sweden's approach might be a disaster if applied in the US, but deaths and ICU cases have been *decreasing* in Sweden, especially in Stockholm, which tends to support the view that there's at least fairly robust short-term immunity. https://t.co/nUMhGfcfvU — PolitiTweet.org

Nate Silver @NateSilver538

@PatrickRuffini The dynamics have become surprisingly similar to election polling, where the media narrative is only loosely correlated with the actual flow of data and usually lags it by 2 weeks. https://t.co/7CxfmiXPWZ — PolitiTweet.org

Nate Silver @NateSilver538

RT @edyong209: 🚨I wrote about the issue of coronavirus “strains”. Despite the recent overblown stories, there’s no strong evidence that mor… — PolitiTweet.org

Nate Silver @NateSilver538

@joshtpm They've had considerably fewer tests per capita than NY, which could be part of the problem. If you're not able to test some of your more marginal cases, there may be more people walking around sick. — PolitiTweet.org

Nate Silver @NateSilver538

@joshtpm Yeah, this has been weird. The CT & NJ data generally hasn't been great, whereas there have big declines in Westchester & Long Island. — PolitiTweet.org

Nate Silver @NateSilver538

I think we need to think/worry about the large group of states (and countries?) that get stuck in the middle. Too many cases to do good contact tracing, too few to sustain as much support for distancing (or get any appreciable benefit from immunity) and no clear exit strategy. — PolitiTweet.org

Nate Silver @NateSilver538

Conversely, if you're a state like Hawaii that never had many cases to begin with, it probably means you have some intrinsic geographic, etc. advantages that keeps transmission low. Also, with a low absolute # of cases you can do much better contact tracing, which helps a lot. — PolitiTweet.org

Nate Silver @NateSilver538

Why is this? If you're a state that had LOTS of cases then two things happen: 1) People are scared so social distancing is stricter and people abide by it more. 2) Some nontrivial fraction of your population is likely immune, at least for some period, which helps at the margin. — PolitiTweet.org

Nate Silver @NateSilver538

One observation related to this: States that have seen clear declines in new cases tend *either* to have had either a LOT of cases early on (e.g. NY, LA) or very FEW cases (e.g. MT, HI). The states in the middle (e.g. MD/VA) aren't seeing much, if any, decline, conversely. — PolitiTweet.org

Nate Silver @NateSilver538

Have no idea if the improvement is sustainable. It may not be—especially if states relax restrictions. But a bit ironic we've had an optimistic 3-day period for new cases data (especially as a % of tests) at the same time the media narrative is about how new cases are exploding. — PolitiTweet.org

Nate Silver @NateSilver538

But the *non* New York numbers have also not been so bad over the past three days, with non-NY cases around ~20K per day (down from a peak of ~23K) and a positive test rate of 8-10% (down from a peak of ~16%). — PolitiTweet.org

Nate Silver @NateSilver538

New York did *not* report data on cases and tests today. (It *did* report data on deaths.) That matters less than it once did because NY's cases have fallen a lot and its positive test rate is now similar to that of the country overall. But it's worth nothing. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths: Today: 1,949 Yesterday: 2,527 One week ago (4/29): 2,700 Newly reported cases: T: 22K Y: 22K 4/29: 27K Newly reported tests: T: 215K Y: 259K 4/29: 230K Positive test rate: T: 10% Y: 9% 4/29: 12% — PolitiTweet.org

Nate Silver @NateSilver538

OMG!!!! — PolitiTweet.org

Nate Silver @NateSilver538

If cases are roughly flat, conversely, it may feel like you're in Groundhog Day and the endgame is less clear (at least until vaccines/herd immunity). So people may begin to say "well, we can't do this *forever*…" and want to relax restrictions at some arbitrary point. — PolitiTweet.org

Nate Silver @NateSilver538

In some ways, if you DO have a sustained decline, it's *easier* to maintain support for distancing measures. First, morale is higher. Second, you can target specific goals, e.g. wait until cases are below X key threshold, and have reasonable confidence about meeting them. — PolitiTweet.org

Nate Silver @NateSilver538

I agree with all the very smart expert plans that use a sustained decline in cases as one of the criteria for re-opening. But I wonder about how that's playing out in practice for decision-makers like governors as a matter of political psychology. — PolitiTweet.org