TUCSON - The media herd has always committed an astonishing degree of statistical malpractice, but it has become even more acute in their coverage of COVID-19.  Something is clearly amiss with journalism schools.  Or is it a problem of reporters being too lazy to conduct fifteen minutes of research on the Internet before embarrassing themselves?  Or are they just not intellectually curious?

Even City Journal might be guilty of malpractice, although the magazine is usually stellar in its use of statistics.  An example is a recent article that applauded Germany for its handling of the virus and compared the country's infection and mortality rates to those in other countries.  Other media have reported the same about Germany.

But neither City Journal nor the other media clarified whether Germany uses a different data collection method that makes its results seem better than other countries.  In the absence of such a qualifying statement, the reader doesn't know if methodology was taken into consideration and, thus, if the reported conclusions can be trusted.  

A seemingly informed reader of the City Journal article posted a comment on-line that the conclusions cannot be trusted, because Germany indeed uses a different methodology.  The comment is pasted at the end of this commentary.

Of course, it's possible that the reader is wrong.  Further research would be necessary to know who is right and who is wrong. 

The point of this commentary isn't to determine who is right or wrong on this particular issue, but, instead, to raise the question of why the media is so accepting of published statistics.

Of course, being human, reporters are susceptible to confirmation bias, as all of us are.  Confirmation bias is the tendency to seek sources that match one's own beliefs and preconceived notions.   This bias is even more pronounced with editorialists, commentators and talk-radio hosts than with news reporters, whether on the right or left. 

Country-to-country comparisons are fraught with not only differences in data collection methods but also with differences in demographics and culture.  For example, Sweden is often held up as a paragon of social justice and income equality.  But the very same media that sing the glories of racial diversity while stereotyping whites as privileged don't point out that Sweden is predominately white.

Speaking of race, the errors and omissions in stories in the American media about race rival the errors and omissions that were in agricultural production statistics reported by Pravda.  The problem has progressed way beyond political correctness to propaganda or even agitprop.

An interesting academic study would be to determine the impact of newsroom diversity on news stories.  Has it made reporting richer, less biased, and more accurate and openminded, as the conventional wisdom says, and as I would hope?  Or has it resulted in reporters engaging in confirmation bias about their own race, by cherry picking statistics that portray their race in a sympathetic light?

Don't look for academia to undertake such a study.

With that segue, let's return to City Journal.

In another recent article, the magazine took New York City Mayor Bill de Blasio to task for citing the oft-repeated statistic that blacks suffer more than other races from COVID-19, due to disparities in income and less access to healthcare.  The article said: 

The uninsured rate among black New Yorkers is only slightly higher than the white rate; Latino New Yorkers, including many illegal aliens, have much higher uninsured rates but a slightly lower death rate. Meantime, Asians in New York City, with higher poverty rates than any other group, show the lowest incidence of Covid-19 deaths, by a significant margin.

Is this true?  Without further study, I don't know.  But I do know that most reported statistics on racial matters and now on COVID-19 can't be trusted, even when they match my beliefs.  It's amazing that many reporters haven't come to the same realization.

* * *

Addendum:  Reader Comment about City Journal Article on Germany's COVID-19 Infection and Mortality Rates   

There is one problem with this rosy story. It's not actually true. If Germany used the same COVID disease classification and testing regime as France, its number would be quite like France's. If France used the same disease classification and testing regime as Germany its numbers would be pretty close to the German ones. The Italian numbers add all possible deaths, not actual ones. The real numbers won't become apparent until the excess mortality numbers become clear in a few years' time.

The German RKI COVID test rules was symptoms plus no prior medical condition or symptoms plus history of repository problems. So any patient who had symptoms plus any of the co-morbidities that actually have very high mortality rates with COVID, pulmonary etc, did not get tested. So, the German mortality numbers are only a good indication of the magnitude of those who did from COVID, not with COVID. The base criteria used by all other European countries.

Then there is testing numbers. Huge numbers of tests in Germany but the actual testing rate per 100,000 population ranged from EU average in some German states to very low in others. Some German states have very low COVID mortality rates because they have done very low per capita testing. Under German rules you have nt died of COVID unless you have the test done. Fewer test, fewer deaths.

All the above has been well documented in the German language media over the last few months. Like in FAZ. Given that Germany has a huge over supply of hospital beds and a well-documented very wasteful system that keeps people in hospital beds from 3 to 4 times longer than any other western country

I dont think there are many lessons to be learned from Germany. Now Taiwan and Hong Kong, that's a very different story. In fact, Taiwan's handling of the SARs CoV 2 outbreak has been peerless and shows how it can be done.