They went to great trouble to sample many thousands, (good), but then used multivariate analysis to reach a general conclusion about the country as a whole while burying the local information in a national "average".
Even my limited encounter with statistics reminds me that samples with high standard deviation CANT be used to force data into a "normal distribution". That's the exact same error that lead to the liquidity crisis ten years ago when the "quants" constructed their derivatives.
It's not 5% in Spain, it's 14.2% in Soria, 11.3% in Madrid, etc.
Still much lower than one might have expected,but it tells the same story that we're seeing in New York: densely settled places are hit hardest and first. It may be possible that more rural and less dense areas simply get hit later, but that's speculation at this point.
First question? What is the rate of false positives and false negatives in these tests?
I live in Spain. They will probably use this study as an excuse to extend the lockdown even longer.
So? Only 4% in US have
Only 0.4% have contracted covid-19 in the U.S.
Part of whats going on in Spain, and elsewhere, is there are regions and countries where it just has not taken. And its not just a urban/rural split either.
In Spain it has for instance largely avoided Andalucia, with its big cities of Sevilla and Malaga, Valencia, Murcia, Galicia, the Canaries, Mallorca - most of “tourist” Spain in fact. This is confirmed by the excess mortality figures from civil records.
The same can be seen in Italy, with an even more distinct north-south split, and globally.