It is said that the first casualty of war is truth. In the purported war on coronavirus, Rod Liddle rightly asks ‘how reliable are the coronavirus figures?’ and comes to the conclusion the answer is 'not very'. He is right, for at least two important reasons.
Firstly, not all states are carrying out the same number of tests per million people. For Germany, it is 2,023; Britain, 960; and France, 560. That has a consequence on the number of positive infected cases. Given that the most widely-accessed world rankings of total cases per country used by the media, such as Worldometers (updated in live-time), are based on that figure, the results are not hugely helpful.
Some might say this shows why we should use total number of deaths per country to determine the impact of coronavirus. Here, though, is the second problem: do countries record deaths in the same way?.
Take France, for instance. Every evening at 7.15pm, news channels are interrupted to hear the director general of health, Professor Salomon, gravely deliver from his lectern the days coronavirus statistics for France. He has been doing so for several weeks. First, the details the number of infected cases; second, the number of deaths.
These are very precise in the way the French admirably are in anything to do with mathematics. No sloppy rounding up or down as might be the case in Britain. Mathematics is a science for the French, which explains why they have more Fields medals per capita than any country in the world. It was therefore disturbing to note that earlier this week, Prof Salomon’s reference to ‘total number of deaths in France’ suddenly changed to ‘total number of deaths in hospitals’.
An owl-eared journalist picked this up in the questions. He asked about the deaths in old peoples’ homes, retirement homes, nursing homes and individual homes. Prof. Salomon, utterly imperturbed, carefully explained that France only recorded deaths in hospitals because it was too complicated to record the others scientifically. Those deaths would eventually be integrated to the official statistics at a later date by the national statistical agency. He admitted, however, that hospital deaths ‘only represent a small part of deaths’. And he added that the ‘two principal places of death are the hospital and nursing homes.’
This was jaw-dropping news for me at least, although it has not provoked much interest in France. But the implications are astounding. It means that the real deaths in France are not 1,696, as recorded and announced by Prof Salomon last night, but far greater. If could quite easily be double that number: if so, France would be the worst affected state in the world, just after Italy and about to overtake Spain.
Now it is quite possible that France is not the only country to collect and present data in this way. But the point is that this contrivance makes the figures for the number of deaths (as well as all the international tables) not comparable. It also gives an artificially low mortality rate. I do not suggest any sleight of hand, but it is surprising that no one in the director general of health’s department (or the health ministry) understood how misleading such statistics could be.
There is yet another statistic that has gone largely unnoticed during France’s remarkably severe lockdown regime. Last night, the French interior minister told France 2 TV news that since the beginning of the national lockdown on 17 March 3.7 million police checks have been carried out nationally to enforce the coronavirus lockdown and 225,000 fines issued. These strike me as large numbers for one of the world’s prominent democracies.
This evening, the French prime minister announced to the nation that the lockdown will last until 15 April. Who knows what statistics will have been amassed by then, other than the most tragic? But as the celebrated British statistician George Udny Yule warned: ‘In our lust for measurement, we frequently measure that which we can rather than that which we wish to measure…and forget that there is a difference’.