No, exactly the opposite - not enough data.
A. because he’s an idiot who cares more about how closing down businesses would affect his buddies in the business world than he cares about the people in his state
B. because he’s an idiot who cares more about doing what is most popular at that moment to increase his chances of reelection, and
C. because he’s an idiot.
Ohio, California, Illinois, and Kentucky have been decently successful at “flattening the curve” so far, by implementing strict measures in relatively early stages of the virus. Ohio has not only reduced the maximum daily projected new hospitalizations by 85%, it has also delayed the peak by a month, allowing the state to prepare for the present worst case scenario.
Unfortunately, Florida will likely follow a trajectory more similar to that of NY than than that of Ohio, California, or Illinois. NY is on its trajectory because A, density of population in NYC, B, DeBlasio is an idiot, and C, the culture in NYC is such that they don’t listen. We are seeing similar behavior in Louisiana, as many megachurches keep on holding services to large crowds.
West Virginia is likely going to be a case in which relatively few are infected, but relatively large numbers of these will be serious and death rate will be relatively high. Early social distancing, the rural nature of the state, and low travel rates overseas and out of the state reduced initial spread and will continue to slow spread, but an older and unhealthier population, and more limited resources like ICU facilities, means that the fatality rate will be high. That is likely why the Murray models predict that WV won’t outstrip it’s regular hospital beds but will likely not have enough ICU units (though they also have enough time to prepare.
Looking at the Murray models, I am skeptical of a number of them, because of their assumptions, and I am especially skeptical of the nationwide model, since the variability in the trajectories of different states mean that the their model is way too smooth and neat. I also think that they are trying to use the same model on states with different levels of data quality, and different methods of data collection.
When you have data which is all biased in the same way, you can incorporate that into your model. However, when the data is biased in different ways, using a model which takes into consideration a different bias than the actual bias in your data produces highly unreliable results. If you are using data with a large variation in biases means that the confidence limits on your results are extremely large.
That is what is happening with the Murray model. The number of beds that they calculate may be needed at the peak in the USA is anywhere between about 100,000 and 460,000, the peak is anywhere between April 10th and April 21st, and it ends anywhere between June 13th and July 23rd.
Of course, this also makes assumptions that are based on preliminary understanding of the coronavirus.
