Counterfactual Deaths and Games with Models, 6/16/20 Data


An interesting point came up in the comments to Monique’s post, yesterday.  I asked what the COVID-19 death rate would have been had we managed to protect vulnerable Rhode Islanders (particularly those in senior care) while not worrying as much about everybody else.

To this, Mario suggested:

The death rate would have been lower, sure, but the number of deaths would have been significantly higher. The idea that the virus could be kept out of nursing homes while allowed free access to the rest of the state is just absurd. The only chance to keep those people safe was with a lockdown of everyone else. The fact that they failed doesn’t negate that it was the only option.

Click over to see the full discussion, but let’s just say I don’t agree that the state couldn’t have better chewed its gum if it had stopped trying to walk at the same time.  Those in senior care have been 17% of all cases in Rhode Island but account for 78% of all deaths.  This is of people who tested positive; the number of non-seniors who’ve had the disease without being tested positive is probably higher.

Even with this ratio, though, Rhode Island would have to have 49,000 cases outside of senior care for the number of deaths outside to equal what we’ve seen inside.  That’s about three-and-a-half times what we’ve had.  If Rhode Island had taken an approach of masks, hygiene, and social distancing — adding in those who would freely have chosen to reduce their out-of-home activity without its being ordered — it isn’t absurd to think the toll of the disease could have been reduced, while also limiting the amount of collateral damage to our economy and to our lives, which includes deprivations around major life events.

As for today’s COVID-19 report, things continue along. The number of new cases was up, although less than 100 and with a significant increase in testing.  Hospitalizations held pretty much steady, even though intensive care unit patients were down, largely because only four people were discharged while six people were admitted.  (Remember, this is people in the hospital for any reason who happen to test positive for COVID-19.)

Also as per usual, the double-digit number of newly reported deaths is not mostly recent deaths.  The numbers have been revised back to mid-April.  So, while the headline number of “new fatalities” is 14, we’ve been under five per day for more than a week.


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

Projections versus actuals (date of report).

  • Cases:
    • Projection for 6/16: 16,140
    • Actual for 6/16: 16,164
    • Projection for 6/17: 16,214
  • Hospitalizations:
    • Projection for 6/16: 122
    • Actual for 6/16: 129
    • Projection for 6/17: 123
  • Deaths:
    • Projections for 6/16: 853
    • Actual for 6/16: 865
    • Projection for 6/17: 868

  • John

    This state will be bankrupt. Very Soon.

  • Lou

    Fascinating revisionist history from someone who was first in denial, and then became known for touting Burrillville as a hot spot as a diversion from where the real “vulnerable ” were/are.