Ep. 30: How western '#datajournalism' on #wuhanvirus deaths in India is a crock

I think western media has been less than honest in its ‘data analysis’ of Wuhan virus cases and deaths in India.

One newspaper said the numbers were an undercount by 30x. Several others have confidently assumed similar large undercounts. After all, India must be hiding the true numbers — for what exactly, they didn’t say.

I tried to understand how they concluded this, just as they confidently said there were ‘250 million “farmers” protesting’, ‘the largest protest ever in the world’, etc, when the actual numbers were probably 2500.

That is an overcount of 100,000x! This is ‘data journalism’?

Even if you assume 25,000 ‘farmers’ (in reality kulaks) were involved, this is a gross exaggeration by a factor of 10,000. This is Soviet-level or CCP-level propaganda. Slate and Foreign Policy ought to be tarred and feathered.

Here’s my hypothesis. The Wuhan virus ‘data analysis’ looked at the obituary pages of local newspapers, added up the deaths from all causes. They did not take the incremental death toll from the Wuhan virus. The base or normal death rate was not considered.

Is this innumeracy, or confirmation bias?

My local paper in Trivandrum has an entire page+ of obituaries every single day, year in and year out: that’s much bigger numbers than the Wuhan virus death toll. Here is the page from May 3: I count 64 obituaries (one is a relative of mine who died of a brain tumor). One person, it says, died of Wuhan virus.

There are three major local papers, so let’s say they will carry notices of 50x3 = 150 deaths, eliminating duplicates.

Now here’s news about deaths in Trivandrum district (population around 3.5 million) from the unquestionably leftist newspaper, The Hindu.

The actual number of Wuhan virus deaths in Trivandrum during the recent spike? Roughly 5 a day.

Ergo, by the impeccable logic of the ‘data journalists’, Trivandrum is under-reporting by a factor of 30, ie 150/5. There you go! Quad Erat Demonstrandum.

I bet this FT fellow (“I collated local news reports”) did precisely what I did above: took all deaths from all causes reported in the media, and pretended they were all from the Wuhan virus. Out of 1.3 billion, some number of people die every day from other causes: heart attacks, accidents, strokes, diabetes, asthma,….. but the Brit and his native informant simply added up all the deaths in the obituaries.

Anecdotally speaking, just think. How many people in your circle of acquaintances fell ill with the Wuhan virus? How many died? How many died of other causes? My hunch is that the all-causes deaths would dwarf the Wuhan virus deaths by a factor of 10-20x.

So much for ‘data journalism’. Lies, damned lies, and statistics. Pandemic porn narrative wins!

3 May 2021, updated 4 May 2021