The University of Washington’s Institute for Health Metrics and Evaluation produces the coronavirus model that currently seems to be getting the most attention. For now, IHME is sticking with its projection of 61,545 COVID-19 deaths in the U.S. by August 4. This is down from former predictions of higher levels.
The IHME model is embarrassing to the authorities here in Minnesota. Our governor’s model projected 74,000 deaths in Minnesota alone without social distancing, and 50,000 with the draconian measures our governor has implemented, to devastating effect. IHME projects 442 deaths in Minnesota. Obviously, someone is badly off the mark. Given that Minnesota currently has only 64 deaths, both models are probably high, with the governor’s model being absurd. Scott wrote about it here.
In this post, I want to focus on how the IHME model treats the five states of the Upper Midwest: Minnesota, Wisconsin, North Dakota, South Dakota and Iowa. Importantly, IHME says clearly that all of its projections are “assuming full social distancing through May 2020.” For the reasons stated below, I don’t think that assertion can be true. Like others, the IHME model is opaque and apparently inconsistent.
Minnesota and Wisconsin have both implemented stay-home orders and closed most businesses, as the IHME site recognizes. IHME currently projects 442 fatalities in Minnesota and a similar number, 357, in Wisconsin. But now take a look at Iowa, North Dakota and South Dakota.
IHME projects 743 COVID-19 deaths in Iowa, considerably more than either Minnesota or Wisconsin. Iowa has around one-half the population of Minnesota and is more rural. It projects 369 deaths in North Dakota and 356 in South Dakota. South Dakota’s population is one-seventh that of Minnesota and the state is even more rural than Iowa.
What is going on here? IHME correctly notes, on its pages dedicated to Iowa, North Dakota and South Dakota, that those states have not implemented stay-home orders and have not closed “non-essential” businesses. One can only assume that the much higher per capita death rates that IHME assigns to these states reflects the absence of such measures. Yet the IHME site explicitly contradicts that conclusion. See, for example, this screen shot of a portion of the organization’s South Dakota page:
With respect to Iowa, South Dakota and North Dakota, IHME’s projections appear to be entirely speculative, because those states have experienced hardly any fatalities, let alone the relatively elevated rate that IHME predicts. IHME’s theory is that the fatalities are going to begin any day now. This, for example, is its projection for South Dakota, which so far has a grand total of six deaths, per IHME:
See what I mean? The epidemic will begin tomorrow.
To be fair, IHME could be right. It is possible that COVID-19 is coming later to more rural states like Iowa, North Dakota and South Dakota. Maybe if we wait another week, deaths will begin to appear at the rate predicted by this model.
This will be an interesting development to watch. South Dakota’s governor, Kristi Noem, is one of the more impressive people in public life. On April 1, she conducted a press conference which you can see here. While encouraging social distancing and other voluntary measures, Noem said that Americans are a free people who value their independence and are fully competent to take appropriate measures to protect themselves. She therefore declined to close down businesses or issue a stay-home order.
That attitude is all too rare among America’s governors, and is the exact opposite of the one held by Minnesota’s Governor Tim Walz. While, as I said, IHME’s model is opaque and contradictory, that organization apparently believes that South Dakota will pay a price for standing up for freedom, along with North Dakota and Iowa. Let’s check back in a week or two and see whether that prediction turns out to be true.