On July 16th, Robert Macky wrote an Intercept article ripping Trump for false reporting about Corona virus deaths. Trump says death rates are going down. Macky says he’s lying to say such a thing. However, Macky ends the article with a demonstration of an ugly bias against Trump.
After slamming the White House for misusing graphs, Macky does worse himself. He ends the article with his version of truth – a graph of TOTAL deaths integrated over all time labeled “per capita mortality rate”.
If the death rate in America dropped to ZERO Covid-19 deaths for a week, month, or year, do you know what the Mackey graph would do? Mathematically it’s impossible for the graph he displays to ever drop down unless the dead rise from the dead. That’s a convenient way to claim bad things about Trump and say he’s lying whenever he says death rate is decreasing.
Now, in my book, THAT is lying in order to call Trump a liar.
Using a proper graph, a very different story emerges. If you make a tiny assumption that the overall population stays about constant for a day or week, the shape of a deaths per day graph would be the same shape as a per capita death RATE graph. Or, see “Daily confirmed deaths: are we bending the curve” at ourworldindata.org.
As of today, Florida and other state rates are going down. The U.S. weekly death rate has dropped about 79% since the April 22nd maximum from 19,006 to 4,070 . Macky could have shown this graph, but it would have supported Trump’s claim that death rates are decreasing. Death rates have now risen back up to 5,329 as of Jul 20th, then they will go down again, which matches the oscillating nature exhibited by every nation. If it’s true in every nation, you’re wrong to blame Trump.
It’s interesting to me that the deaths seem to rise and fall on a weekly pattern. See “What is the daily number of confirmed deaths” at orwoldindata.org. To any statistician, this is a huge and obvious residual. There is no virus that knows what day of the week it is. Makes me wonder how the data is reported.