Social distancing really does flatten the curve. (See Italian data below.) This will save lives. Are we prepared for the inevitable backlash?
Part I: The Good News: Social Distancing Works
I have been tracking the effect of social distancing on specific countries and states that implemented social distancing and I am seeing that the data is shifting from an exponential fit to a sigmoid fit.
For example, Italy. Data taken daily from Al Jazeera’s updates and Wikipedia’s coronavirus page. Graph is through 31/March.
In this figure, I’ve plotted deaths per day in the dots, and fit three curves to it. The yellow line shows an exponential to the first 14 days, which you can see fits well for the first 14 days, but then deviates from the data (R2=0.75). The red line shows an overall exponential fit to all of the data, which does not fit well (R2 less than 0). The purple line shows a sigmoid fit. The sigmoid fit explains the data well (R2 =0.93).
Italy locked itself down on 09/March. It took two weeks, but they have successfully flattened the curve. I hope that they can keep it up. I hope that this is the peak and data will soon be better fit by a Gaussian.
When I do similar analyses for various US states or for the US as a whole, we are generally still too early to differentiate the exponential and sigmoid fits, but they look like they are starting to flatten as we approach two weeks. As you can see in the Italy data, it was only in the third week that a clear effect was seen. But, in general, the lockdowns seem to be working.
These data line up with the excellent work by the Institute for Health Metrics and Evaluation at the University of Washington. (That’s a great site, by well-respected researchers, with state-by-state projections of expected hospital resource use and expected death tolls.)
So, stay home, stay safe, and stay physically distant. (And support your health professionals!)
Part II: The bad news. Y2K
My numbers are predicting that with this social distancing, we are not looking at large death tolls. (That’s great news!) The IHME sees similar numbers. (They predict a death toll around 80,000. Every death is terrible, and I would like to have seen a lower death toll than that. If we had done this right from the beginning, maybe it could have been much lower.) But Trump is now creating a public expectation of 100k-200k+. (Presumably, this is because of a model from the CDC. I hope the IHME model is right and the CDC model is wrong.)
A digression for the younger readers among us (and the forgetful older readers): In the waning years of the 20th century, computer programmers started to raise a concern that people were still using old code in critical infrastructure. This old code had a problem. It stored dates as decimals, with the year as only the last two digits of the year. (So 1985 was “85” and 1990 was “90”.) This would be a problem come 2000, because the programs would assume that the year was 1900 and would error out. In 1999, people were scared and hoarded supplies for the inevitable crisis… that never came. The clock ticked over from 31 December 1999 to 01 January 2000 and nothing happened. A lot of people to this day think the Y2K bug was a hoax. But it was not! We fixed it! We fixed it completely. I was just a young pup programming (I worked at IBM for summer internships) in the late 80’s and early 90’s but I knew a lot of old programmers (including some old COBOL and FORTRAN programmers) who were called out of retirement to fix their code. The programmers fixed the code and the crisis was averted.
After the Y2K non-disaster averted disaster, I started noticing a lot of disbelief about government investment in disaster relief and a lot of disbelief in scientists’ warnings about future disasters (q.v. global warming / climate change). Yes, I know that both of these ideas preceded Y2K, but they gained a lot of traction from the Y2K non-disaster averted disaster. There was a definite sense that one could only “cry wolf” so many times.
So the purpose of this diary is to put out to the DailyKos community:
- How are we going to prepare for Trump to say “See what a great job I did? I kept the death toll below 100k. Everyone said that it couldn’t be done. Everyone said there would be millions of deaths….” How are we going to prepare for the news media to fall in line, as they often do?
- How are we going to respond when Trump claims that the pandemic was a hoax to hurt his election chances? (Perhaps by disrupting the economy?) How are we going to prepare when the news media presents both sides as if they have equal merit?
- How are we going to prepare for the inevitable questions about whether we really needed to crash the economy? (Imagine we do this really well and we end up with 50k deaths. That’s a lot. That’s terrible. But that’s a typical flu season. What are we going to say when the market media asks whether the cure was worse than the disease?)
- How are we going to prepare for the right wing to argue that this was a ploy to gather support for a progressive agenda (better unemployment, including unemployment for gig workers, Medicare-for-all, stimulus checks, universal basic income)?
The problem is no one can see the counterfactual.
Yes, let’s save the world. But let’s be ready for the inevitable political counter-attack when we do.