In the Spectator Dalrymple reviews a tendentious book, Why Some Politicians Are More Dangerous Than Others, in which New York University psychiatry professor James Gilligan claims that Republican presidencies inevitably lead to a much higher death rate. (H/t Michael P.)
Some of his statistical manipulations to arrive at his conclusions — which, despite his repeated protestations to the contrary, I suspect were foregone — seem to me to be doubtful. For example, on page 28 he states that the difference in violent death rates between Republican and Democratic presidencies ‘amounts to a difference… representing roughly 114,600 fewer violent deaths per year under Democrats than under Republicans’.A man who can look at that figure and not see immediately that it must be wrong is no more to be trusted with statistics than an alcoholic in a wine-merchant’s. There are approximately 16,000 murders a year in the United States, and 30,000 suicides. Even Gilligan does not maintain that all of those are caused by Republican presidencies, or the memory or possibility thereof; but even if he did maintain it, it would still not amount to 114,600 per year. He writes what cannot be true.