ARE STATISTICIANS LIARS ?
[To the Editor of THE SPECTATOR] SIR,—Mr. Worsnop, far outdoing in scientific caution the Medical Research Council, to whose report he refers, dis- misses Mr. R. M. Moore's table of comparative death-rates as " valueless " because it is not " supported by data as to environmental situation, habits of life, anthropological charac- teristics, &c., &c., of each of the 628,000 individuals con- cerned." In this I think he is over-cautious, but he certainly shows a critical spirit that commands respect : which makes it all the more surprising that he should be so ready to draw conclusions from the discredited B.M.A. inquiry, which cer- tainly fulfils none of the conditions he mentions, and which also commits the " howler " of comparing the death-rates of non-comparable age-groups.
As regards Mr. Moore's investigation: both the Medical Research Council and Newsholme in Vital Statistics agree that, when all allowance is made for the special difficulties of interpreting insurance figures, and for further complicating factors that are discussed by Moore himself and by others, there remains strong evidence that, when a simple division is made between abstainers and non-abstainers, the former have a lower mortality rate. Whether this relation between absti- nence and longevity indicates a direct causal connexion is, of course, quite another matter, and if Mr. Worsnop will re-read my letter he will see that it is one on which I did not attempt to pronounce.
I have no quarrel with Mr. Worsnop's anti-teetotal prefer- ences, but only with his arguments : and I certainly do not want to dispute the pronouncement of the Medical Research Council that the statistical evidence so far available does not suggest that strictly moderate drinking shortens life.—Yours
Previous page
