Which States Are Doing the Best or the Worst With Coroanvirus?
It may be that you have been checking those “coronavirus trackers” on line, wondering what’s up with that virus you may have heard a little bit about.
Although interesting, this raw data has some limitations.
For one thing, it’s a tally of every assumed/known/reported/estimated case ever, so it’s never going to go down, unless you have access to a time machine somehow. (And you can ignore that silly “recovered” statistic, as no one is tracking that. If a person didn’t die, they recovered in 1-2 weeks, although a bad case of respiratory illness can leave you feeling pretty worn out for a few weeks as well.) So, it does not tell us what is happening now, which would be much more useful information.
Also, the raw data doesn’t take into account population figures. A country with 40 million people might have, by the raw data, 2,000 total cases over the past few months and another country has 10,000. But, whoops, that second country has 200 million inhabitants, so it’s actually doing better.
We thought it might be interesting to take the raw data for the US individual states, and see if we can make something out of it. Again, a limitation of the following is that these reported figures are just total cases since the beginning of tracking in February 2020, so they don’t tell us how a state is doing right now.
What we did was first note a state’s (and the District of Columbia) population rank in the United States. For example, California has the most people of any state in the US and thus they are number one. Wyoming has the least people and ranks at number 51.
We then noted down the ranking for number of cases per state off of the Bing search engine “tracker.” For example, New York has the most reported cases since the beginning of this issue, and therefore ranks number one on the case list. Wyoming has the least and ranks 51st.
We then took these numbers and compared them. We subtracted the case ranking from the population ranking. For example, Arkansas ranks 34th in population and 36th in rank in terms of number of cases, giving them a score of -2. Iowa ranks 32th in population and 29th in numbers of cases giving them a score of +3.
If everything were equal and this virus was equally distributed across the US, every state should score a zero. Their rank in terms of population should match up with their rank in terms of cases. Of course, that did not happen.
The higher the score in the positive direction, the more number of cases the state has relative to their population ranking versus the expected result if everything were equal.
The lower the score, or a negative score, means lower numbers of cases than expected for pure population rank, which is better than the "positive" number in this case..
Let’s look at the numbers, and then we can discuss or speculate on what they might be telling us.
Number are from the morning of April 23, 2020.
Here are the numbers for the calculation of State rank minus case reported rank equals?
First listed by state population. Remember the more negative or lower the number, the better.
4.New York +3
13.Wash St -2
36.N Mex -1
41.N Hamp -1
43.Rhode Is +23
45.S Dakota +7
46.N Dakota +2
What this initially shows us is what we already knew, that cases reported are not evenly distributed across the US.
Let’s look at the “top ten” for more cases than expected.
1.Rhode Island +23
8.New Jersey +9
9.South Dakota +7
Again, this list tells us what we already knew. The major epicenter of this infectious agent is in the Northeast and along what is sometimes called “Bo-Wash,” the conurbation that runs from Boston to DC. There is a secondary epicenter in the deep South centered on Louisiana and Mississippi. And there is something going on in South Dakota?
This data, of course, does have limitations. We don’t know what’s going on right now from this data, nor do we know the rate of testing in each state. And if a state has a smaller population, it reduces the statistical power of the calculator, because it more possible for a few cases to move the needle in one direction or the other.
What are the top ten“best” states, the ones with the much lower rates of infection that expected?
3.N Carolina -11
And a shout out to the congenial states of Florida, Kentucky and West Virginia which all come in at -5.
A word about New York, which you may have noticed is not on either of these lists, despite all the drama. New York is fourth is population rank and first in cases reported, given them a score of +3. This means their reported cases are fairly close to what you would expect for their population.
Why is this? Aren’t they supposed to be the worst of the worst? Two factors. First, New York state is famously two separate places, New York City and Upstate. Probably most of the cases have been concentrated in the city area. Second, New York City is sharing the pain with the surrounding communities nearest to it, many of which are located in Connecticut, New Jersey and other nearby states. If you look back at the list of all states, you will also see that of the top ten most populous states, New York is tied with Michigan for the “worst” score.
Otherwise, what you first notice from these best and worst lists, is, as the real estate agents say, “location, location, location,” or as physicians say “common things are common.” Places where the virus is common have more cases. Places that are located on the edges of the country, that are harder to reach, such as Minnesota, Maine, New Hampshire, and Oregon have comparatively less, as does the mid-atlantic “south-light” are of North Carolina, Virginia, Tennessee and Kentucky. Places with wide open spaces, such as Arizona and Texas are also doing well.
It also shoes two things may be having less impact than people think: urban areas and restrictive measures.
Despite a wide variation in implementation, there is little to suggest that global “lock-downs” or harsher restrictions have had much of an effect. Most states, outside of the spared or more afflicted areas are pretty much statistically the same in population to case ratio—that is, a variation of a few “points” that is seen cannot be shown to be to anything but chance. That makes sense as well. These measures, by necessity have been driven by human, rather than viral agendas. A virus does not know whether you are buying bread or a basketball, and certainly doesn’t care if your job is considered “essential.”
There doesn’t seem to be as much association with urban areas per se as one might think either. New York and Michigan have urban areas, but so do Texas and Arizona, not to mention California and Missouri, both of which clock in at -3.
This needs further study, but one suggestion would be that this may be related to the penetrance of public transportation in a particular urban area. New York and DC have very busy and much used mass public transport, and in particular underground trains. Phoenix, Houston, Dallas, Los Angeles and, surprisingly, the Bay area of California, do not have such well-developed public mass transport. If further analysis gives support to this speculation, then forward planning for controlling infection spread should include measures to mitigate this threat.