MENU USA versus COVID-19: Who’s Winning?
This is truly the battle of the century- not since 1918 have we faced such an opponent. Since a picture is worth a thousand words, we’ll let them tell the story. The recently completed series of myths regarding the data raised several important interpretation obstacles; the myths will be referred to as needed, but not rehashed.
Case Count
The case count took off in March and peaked around April 1. Here’s the daily number of new cases since then:
There are eleven days in a row below the centerline at the end. Assuming data accuracy, that is a statistically significant finding and evidence that case count country-wide is dropping. Good news? Well, it certainly isn’t bad news, but we need to take a stab at understanding why it is happening. Here is a ranking of cases by state:
Couple of things: New York state has more cases than 35 other states combined, and the New York / New Jersey area accounts for about a third of the overall cases. That would suggest that if the case count is truly dropping, then it should be driven by this area. Take a look:
As expected, there’s a clear trend in the downward direction for the two highest-volume states- definitely a positive sign. It’s important not to assume these results apply to the entire country, however. If we remove NY/NJ and look at the rest of the country:
No trend in either direction for the “other 48” as a group. Breaking the data down by state is the typical next step. Here are two geographically dispersed examples:
Looking west coast- California does not show any signs of dropping. If anything, the case count is slightly increasing. Looking southeast, Florida appears to be in a relatively steady state for the last month or so. This isn’t criticism toward either state- simply a recognition that we can’t assume every location is at the same point in its journey to combat the virus. There’s no one-size-fits-all assessment to be made.
So what does it all mean?
- The nationwide case count is dropping- but there are still 20,000+ new cases / day.
- New York and New Jersey dwarf the other states in volume; as they go, so goes the overall trend.
- The United States is too big to view as one entity; case count progression needs to be evaluated on a state level to determine the different stages of progress.
- Not every state is past its peak; some are rising and some appear to have stabilized.
Virus-Related Deaths
The metric we’ll use for this one is the number of deaths versus mortality rate, in an effort to reduce data inaccuracy as much as possible (see Myth #2). Unfortunately, there still appears to be data collection issues. The graph on the left shows daily numbers. The points circled in red correspond to weekend reporting days, which clearly have a different data collection process. To mitigate that, we calculated a daily average for each week, shown on the right. The good news is that there appears to be a downward trend over the last several weeks.
What states are the deaths coming from? The breakdown is as follows:
New York and New Jersey once again have the highest volumes, which is not surprising since they have the most cases. Note that 42 states combined do not have the death count of New York alone. The New York deaths per day chart below mirrors the overall, which is logical. We definitely have statistical evidence that New York is past its peak.
Note the weekend dip isn’t pronounced in New York, suggesting a more consistent reporting process. A quick look at the data reveals that other high-count states like New Jersey, Massachusetts, Pennsylvania, and California clearly are not reporting deaths consistently on the weekends.
Looking Ahead- TOUGH Decisions to be Made
Can the United States learn from the success other countries have had in dealing with the virus? Sure, to a point. Studying what has worked in other countries like Italy and Spain is smart, but only a first step. The graph below illustrates the size differential between the United States and European countries- Italy is about the same size as Arizona. So the individual states have a certain degree of autonomy to make their own decisions.
How are the states doing? It would be impractical to go through each one by one, so we’ll use Florida as an example because it raises a lot of key issues. Myth #5 illustrated the challenge of balancing medical prudence with economic realities. Florida grosses over $100 billion dollars per year in tourism revenue, and tourism keeps over 750,000 people employed. Closing down the state has significant financial repercussions, to say the least. With that in mind, the state made the decision to re-open more aggressively on May 18th. Meanwhile, here’s the data on case count and deaths per day (adjusted to get an average per week to compensate for lack of weekend reporting):
Interpretation: there’s no statistical evidence that anything has changed appreciably in the last month or so that would support Florida making their move to reopen now.
Does that make it a bad decision? Didn’t say that. Florida is averaging about 820 new cases and 40 COVID deaths per day. There are two important and horrible questions to answer:
- Is that “okay”? Obviously, Florida would rather have zero cases and zero deaths. But is it worth keeping 750,000 people unemployed and losing an average of $300M per day in tourism revenue to try to reduce those numbers?
- How much of an increase in cases / deaths will occur when the doors are opened?
The first question has been answered by the tough decision to reopen. The second is the real wild card. I chose Florida in part because of the article “Florida throws open its doors — and holds its breath.” That title perfectly summarizes the situation that the individual states and the nation are facing. Potential outcomes range from “Why did we shut down in the first place?” to “Oh crap, what have we done?”
I’m holding my breath, too.
Data and graphs current through May 18-19, 2020.