Pandemic by the Numbers – Maybe

Lies, damn lies and statistics
Mark Twain

The above phrase is often attribute to Mark Twain but has probably been around as long as statistics have been accepted as valid “facts.”  However, that validity comes with a very strong caveat.  Statistics can be very misleading unless you, in fact, are a statistician and prepared to ask all of the appropriate underlying questions. 
A simple example is instructive.  The Standard & Poor’s Index dropped a little over thirty-five percent (35%) from its high to its low during the COVID pandemic.  The market has grown thirty-one percent (31%) from that low to Friday’s close.  So the market lost 35% and has regained 31%.  No problem, your 401K or IRA should be in great shape – but it isn’t.  Despite gaining 31%, the S&P is still down about 15%.   And the reason is quite simple.  When measuring the decline of a market, the highest number becomes the denominator while when measuring the growth in the market the lowest number becomes the denominator.  The point is when you read a statistic – particularly one put forward by a bias media – you need to ask about the underlying data.

Statistical modeling is an even more fragile commodity.  When you ask a six-year-old what two plus two is she will say “four” and smile about how well she has done.  If you ask the same question of a statistical modeler she will respond, “What would you like it to be?
The fact of the matter is that modeling relies on “assumptions” and based on those assumptions predicts outcomes – usually in ranges.  The modeling is basically a set of equations while the assumptions that populate those equations are usually- but not always – based upon observable events.  An article in Monday’s Wall Street Journal by Max Colchester noted:
“Some leading modelers say their disciplines is being asked to provide a level of certainty that it is unrealistic to expect, especially given how little is known about the new coronavirus.  And they fear that they could become scapegoats for politically unpopular policies.
“’Any model that gets within 50% of the actual result has done well,’ says Keith Neal, a professor in the epidemiology of infectious diseases at the University of Nottingham.  ‘It is not an exact science.’”
So, for example, the modeling used by the CDC for projecting the effects of the coronavirus (COVID-19) is based on observable effects primarily in South Korea and Italy.  The effects included the number of cases, the speed of the spread, the morbidity, and the recovery rates.  And after that, the impacts of each of those elements on the number of beds, ventilators, protective clothing, testing regimens, etc. that would be needed. The initial projections from these models suggested that the number of deaths from COVID 19 would be between 100,000 and 240,000 and those figures assumed a degree of success in mitigation efforts.  Now – a short 30 or so days later – the model suggests that the total number of deaths will be less than 60,000.  The count as of Sunday was about 40,500 and the ratio between diagnosed cases and morbidity has likewise been dropping.  Supporters of the modeling note that the projections will change as more data is gathered.
Which raises as huge question about the validity of the data and just as important the manipulation of the data by the media.  The mainstream media seems to have taken great joy in using the reported data in the United States to suggest that America is wrong, that we are the epicenter of the virus and that despite Herculean efforts we have largely failed – failed in stemming the spread, failed in treating those infected and failed in saving the lives of those infected – and that the worst is yet to come.  And in order to bolster their agenda, they use selective pieces of data and projection.  It is beyond me why the mainstream media would be cheering for our country’s failure.
So let’s start taking some of this apart.  And it all comes down to my first admonition: When you read a statistic – particularly one put forward by a bias media – you need to ask about the underlying data.
Is the United States the epicenter of the coronavirus?  Many members of the mainstream media – those who are routinely wrong about everything – have claimed so with great relish using the data on confirmed cases and deaths.
And even those who don’t make the claim directly, including the normally staid Wall Street Journal, suggest the same by the routine comparison of available numbers for the United States vs. the world.  But in order to make the claim that the United States is the worst in the world you have to have the same accuracy in reporting as that which is being done in America – and we don’t, not by a long shot.
So let’s use the Wall Street Journal to demonstrate the point.  Each day, usually on page A6, the Journal publishes a comparison of United States data versus worldwide data.  For instance, on Monday, April 20, it noted the following:
U.S. Cases – 755.533
Worldwide cases – 2,394,291
U.S. Deaths – 40,461
Worldwide deaths –164,938
U.S. Recovered –67,172
Worldwide Recovered – 611,880
Of those six numbers only one is verifiable – U.S. Deaths.  Neither the Wall Street Journal nor any member of the mainstream media, nor any members of the Center for Disease Control, nor the White House Coronavirus Task Force know whether any of the other numbers are accurate within a magnitude.  In fact we do know that within a closed society (China, Russia, Iran, North Korea, etc.) where the government controls information that the numbers produced by them are largely suspect since they routinely lie about everything.  And yet there they sit without any disclaimers leading readers to conclude that the United States is indeed the epicenter of COVID-19.
Not only do they not know the accuracy of the figures on a worldwide basis, they do not know the accuracy of the figures for United State cases since the numbers reported represent only those cases confirmed by physicians.  We know that in many cases those contracting the coronavirus have few if any symptoms and that those symptoms pass quickly.  In point of fact, ABC – not necessarily a beacon of accuracy – reported on a study conducted by Stanford University of 3,300 residents of Santa Clara County (CA) that indicated that the antibodies produced by people who have had the coronavirus were present in between forty-one and eighty times as many people as reported in the “official” count.
 Now, if those figures reported by ABC are accurate it would suggest that not only the actual number of people infected has increased dramatically, but that the number of people “recovered” would increase even more dramatically. Which in turn would suggest that there is a virtual army of people who have been infected, recovered and are now presumptively immune.  An army so sufficiently large the nations economy could be restarted without much risk to the general public. (I caution about whether the figures are accurate both because of the news organization that reported them and because we have no idea how the study was conducted and which test was used to determine the presence of the antibodies – the accuracy of these tests vary widely with some resulting in as much as a fifty percent error rate.  That conforms to my initial concern that before you accept any of these numbers and projection you need to know the underlying data and algorithms.)
What it would not show – contrary to what I suspect is the motive for reporting the study in the first place – is that the United States has even more cases proportionally than the rest of the world.  Unless a similar study is conducted worldwide any conclusions drawn from the Stanford study as compared to the rest of the world are illusionary.  It is important to verify the methodology, results and conclusions from the Stanford study because it holds promise, if accurate.
When you step into the world of statistics and statistical modeling you are stepping on ice.  Worse yet, if you use that information to propel a political agenda you are doing so at the expense of the unsuspecting public’s health and economic well being.  A better solution is to put all of the information out there – including the lack of information from other nations – and let the American public start making their own decisions about the degree of risk they are willing to assume.  The idea that we are allowing politicians who have never accomplished anything in life other than getting elected to make decisions regarding our health and economic safety is absurd.  Particularly where many of those decisions intrude on our constitutional rights – freedomof speech, freedom of religion, the right to keep and bear arms, the prohibition against a taking without due process or just compensation, and the right to move freely about the country.
And so we return to the beginning.  There are liars, damn liars and statisticians.  Figures lie and liars figure.  And all the while the American public suffers.