Games with Models, 5/21/20 Data

COVID19-hospitalizationsandprojections-052120-featured

In her daily press conference, today, Governor Gina Raimondo apparently stated that if she’d known then what she knows now, she would have shutdown the state’s economy sooner.  This as the state Department of Labor and Training announces April’s unemployment rate as 17%, with nearly 100,000 RI-based jobs simply evaporating (that’s about 20%).  That is the highest unemployment rate (by far) since the Bureau of Labor Statistics began tracking numbers in the ’70s.

Let’s hope there aren’t any more pandemics — or even particularly bad flu seasons — while this governor remains in office.  She’ll have a hair-trigger standard for shutting down, even as some of us believe the data and medical findings are suggesting that, if we knew then what we know now, we never needed to shut down at all.

Today’s data release was good news.  The number of new cases of COVID-19 dropped, with the 14-day infection rate falling below 0.3 (where 1.0 is the threshold between expanding and receding).  The number of reported deaths went up quite a bit, but that included revisions to numbers for prior days going back to April 15. New deaths were six and nine for the prior to days.

The best news came with hospitalizations.  The number of newly admitted patients went down again, while the number of discharged patients went up quite a bit, although not to the two-week average, for some reason.

COVID19-hospitalizationsandprojections-052120

(See here for my original methodology and here for a subsequent modification I made.)

Projections versus actuals (date of report).

  • Cases:
    • Projection for 5/21: 13,492
    • Actual for 5/21: 13,571
    • Projection for 5/22: 13,712
  • Hospitalizations:
    • Projection for 5/21: 241
    • Actual for 5/21: 254
    • Projection for 5/22: 245
  • Deaths:
    • Projection for 5/21: 543
    • Actual for 5/21: 556
    • Projection for 5/22: 562


  • Mario

    The hospital discharges are legitimately strange, but everything is basically continuously improving. I still don’t see anything actually making it into July, although the expected number of deaths varies quite a bit each day. Only on the unexpectedly bad day a little while ago did it go above 700, so that, I think, is as bad as I could see it getting.

    She is right about shutting down earlier, though. She was at least four or five days late based only on what we knew at the time. If there is any case to be made that the shutdown is unnecessary, it is only that the people are apparently capable of adjusting their behavior enough to suffice on their own, which is great, but we can’t really assume that the counterfactual where the people act to reduce their risk even if the government doesn’t take the danger seriously, and the fact that the people reducing their risk would cause the same economic disruption as the shutdown did makes the debate more or less moot.