Summary: The information superhighway provides us with data on a scale unavailable to anyone before the modern era. Unfortunately, it appears to be making us dumber. Seeing our responses to COVID-19 shows how. We can’t afford this.
Lost in the weeds
One of the great oddities with many major events is that the reporters outnumber the participants. There is not only a massive overcapacity of news media vs. their audiences – the competition making them often desperate for clicks – but there are too much news media and vs. the available news even during the crisis, making them desperate for something to say.
So the major media floods us with trivia about COVID-19. Every case is reported. Countless people give analysis and make predictions. Some are experts in the field. Some are experts in somewhat related fields. Some are absurdly confident amateurs. All get reported like certainties. This produces a cacophony, confusing more people than it illuminates.
Worse, too many people pick the views that match their political biases, making effective leadership more difficult – and fracturing our social cohesion (people cannot act together if they see different realities).
Building big conclusions in the air
Perhaps the most common result of the flood of data about COVID-19 is the construction of castles in the air. Much of the data is of low quality and rapidly changing (a core reality in both wars and epidemics). In today’s America, competition is fierce in almost every field. The temptation to experts is great to use this kaleidoscope of data to produce exciting stories and get their 15 minutes of fame.
We see this with the exciting news stories of models’ forecasts about COVID-19. The models are valid but their assumptions are guesses. Modeling the spread of COVID-19 in an unprepared population is relatively easy, assuming one can accurately predict the deterioration in fatality rates as infection rates rise. But how to model the spread in a population taking large-scale but varying protection measures – from more washing of hands to social distancing to sheltering-in-place? The modeling quickly becomes garbage-in, garbage-out. Worse, journalists often ignore the experts’ caveats and report them like prophecies.
Amateur experts are a bigger problem. They take the numbers and produce big bold conclusions, mostly gibberish since they do not understand what the numbers mean. They usually report their conclusions with mad confidence.
Their most common error is assuming that key epidemiological factors are constants – and crunching the numbers to declare that the experts’ numbers are wrong. But these numbers represent conditions only of a specific time and place. R0 is affected by the population’s density, age distribution, health, and social behaviors (see the CDC page explaining how it is “easily misrepresented, misinterpreted, and misapplied.”). The case fatality rate is affected by the population’s age and health plus the effectiveness of its health care system – and by the definition and methods of identifying a “case” (e.g., clinical criteria, testing of the ill, mass serological screening – all give different counts of cases and so different fatality rates).
Both R0 and fatality rates are affected by measures to defend against the virus. Calculation of these numbers is not like counting apples.
Now for the bad news
“Bad money drives out good.”
— Gresham’s Law (1860).
All of these dynamics generate “content” that is reported, misreported, and exaggerated across the internet.
Now for the bad news. This mess of misinformation often displaces authoritative information in the public spaces, just as bad money displaces good money in the marketplace. People travel the information superhighway to become better informed. They often become less well-informed than when they began.
The saddest aspect of all this is that the essentials of the COVID-19 epic are reported daily in a clear and concise form.