Worldwide, since the start of the covid-19 pandemic, 755 deaths from this disease have been recorded for every million inhabitants, according to compilations made by Johns Hopkins University. If in one country the death rate from the pandemic was only a quarter of the world record, it would be a model for the whole planet to emulate what it has done.
Such a country has a name: Venezuela. According to the compilations of the American university, it suffered only 196 covid deaths per million inhabitants, which, in principle, would indicate exceptional health management and social behavior… Or that the accounts he carries do not correspond to reality.
There are different countries that show dubious statistics, and Dmitry Kobak, a researcher at the University of Tübingen, Germany, set out to test what these recordings say. This scientist from the Berens laboratory – which applies data science to the dynamics of health – reviewed all the reports made daily since the first contagion and the first death reported by 237 countries and territories to the World Health Organization (WHO) for two years.
After making the expensive trip, he discovered that there are 18 countries whose death data raises doubts, and three other countries where the information that raises suspicion is not only that of deaths but also that of infections. .
Which countries are these? How can we determine that the accounts of any country have problems? There are different instruments that statisticians use, such as the variance or the Poisson distribution (after the 19th century French mathematician and physicist Siméon Denis Poisson). These tools help to assess whether country reports are reliable.
Summarizing as simply as possible, in a reality like infections or deaths in an epidemic, each day’s cases respond to a trend but in the middle of it there are sudden changes, because factors predictable influence as well as other random factors. those.
Pandemic data must vary and appear erratic from day to day because, as Kobak explains, “people are infected randomly and the disease progresses independently in each patient”.
It is for this reason that it is customary to soften the presentation of data, showing averages of one or two weeks.
When the day-to-day data is very similar, that’s not the normal way daily infections or deaths behave. Rather, it seems, says Kobak, “a statistical anomaly… strong indicator of data manipulation and suggesting deliberate deception”.
The researcher also found that when the numbers of deaths or infections are very similar from day to day, this goes hand in hand with a record, both of deaths and infections, lower than what is happening in reality.
“Most of the countries on this list,” he says, “are known, based on excess mortality data, to greatly underestimate covid deaths”.
Anomalous data in Russia
In Russia, for example, Kobak’s study found that in the first week of September last year, recorded deaths from the pandemic ranged from 790 to 799 per day, a very narrow range.
In contrast, over the same period, deaths in the United States ranged from 1,185 to 2,010, with stronger day-to-day variations. And although the Kobak report does not mention it, we can say that in Colombia, also that week, deaths fluctuated between 47 and 81, also a wide range.
Thus, throughout this period, the death records of the United States and Colombia have experienced large variations, of 28 and 18% on average respectively, compared to the daily average of deaths.
On the contrary, in Russia, the variations were only 0.3% compared to the daily average of deaths. According to Kobak’s tests, such a result has 1 in 143,000 chance of happening, indicating there is something fishy about the Russian accounts.
Kobak even reviews, separately, the accounts of each of the 85 federal regions of Russia, and finds the same anomaly in 82 of them. “This suggests – he says – that the manipulation of data that occurs at the regional level may become invisible at the country level and, also, that a separate mechanism may have been set up in Russia in August and September 2021 for keep the number of reported deaths just below 800 at the federal level”.
Other countries with anomalies
The possible maneuvers to balance the covid accounts do not seem to be a generalized fact in the world. In fact, in all of the 237 countries and territories observed, the anomalies are only observed in 18 of them for deaths and in three others for infections.
In addition to Russia, the countries that arouse suspicion in their death certificates are Albania, Saudi Arabia, Algeria, Azerbaijan, Belarus, Cambodia, Egypt, El Salvador, Kyrgyzstan, Lebanon, Mongolia, Nicaragua, Serbia, Syria, Turkey, Uzbekistan and Venezuela in which mortality is said to be a quarter of that of the world.
Nicaragua is a particularly telling case, reporting exactly one death per week since early 2021. Our test correctly detects this obviously implausible lack of variation.
Also, in the way infections are counted, Qatar, the United Arab Emirates and Tajikistan are added as suspects.
Besides, the registration of cases or deaths that seems to be done to maintain a pre-agreed figure goes hand in hand with the under-registration of deaths. Kobak discovered this by comparing the results of his study with a follow-up of excess deaths during the pandemic in 118 countries carried out by him with economist and statistician Ariel Karlinsky.
To estimate how many covid deaths countries are not recording, cases confirmed by authorities are compared to excess mortality compared to normal times.
With this procedure, Kobak and Karlinsky calculated that only one in 105 covid deaths was officially registered in Tajikistan; in Nicaragua, one in 50; in Albania, one in 18; in Belarus, one in 14; in Egypt, one in 13, and in Russia only one in 3 or 4 deaths from covid.
And, precisely, these countries are part of the group for which there is an indication that the files are manipulated. Other countries in this group, such as Saudi Arabia, Cambodia, the United Arab Emirates or Venezuela, do not even have data to see how much of the excess mortality was due to the pandemic.
Colombia is one of the countries where deaths officially registered as covid coincide more with excess mortality. Studies estimate that 100 out of 128 deaths from covid have been recorded, a level similar to that of Italy and even lower than that of the Netherlands.
Among the countries in which the registration of deaths by covid seems accurate, since it corresponds precisely to excess mortality, are Malta, Chile, Switzerland or Panama.
The under-reporting, Kobak says, may be due to insufficient testing capacity. But the high undercount proportions suggest deliberate misreporting or misdiagnosis of Covid deaths, “which is consistent with the idea that the insufficient dispersion can also indicate data manipulation”.
The studies of Kobak and Karlinsky have thus made it possible to find signs of manipulation in the way in which deaths due to covid are recorded in around fifteen countries. Likewise, they establish the relationship in this way to hold the number of deaths with a lower record than what happens in reality. And then, like the British weekly The Economistthe anomaly occurs almost exclusively in countries “without functioning democracy or a free press”.