William Stanley Jevons, a self-claimed revolutionary fighting against “the noxious influence of authority”, brought various innovations to economic science. The marginal utility theory of exchange, of course, was among them and has been cited the most. His persistent promotion of mathematical economics has also been noticed. But there was another line of economic inquiries which occupied Jevons for the whole period of his career. It may be called statistical economics. Although Jevons has been referred to as a pioneer in the use of index numbers, it is only in a few recent studies that he is examined exclusively as a statistician (Stigler 1982; Aldrich 1987).
My examination of Jevons’s statistical works will differ in some ways from these and other previous studies. Focusing on the statistical methods as he applied them to economic investigations, the examination will illuminate what he thought of such methods and what uses he made of them. In particular it will be shown that he trusted and employed statistical methods to ascertain the exact forms of economic relations. The present chapter has a further theme; his statistical works are examined in relation to the view of John Elliot Cairnes that we could never have “an exposition of economic principles drawn up in quantitative formulas”. There was nothing unconventional in this view, but Cairnes advanced it in an unusually straightforward manner. And that is exactly why his argument is reviewed in the next section. Then the penultimate section will examine how Jevons employed statistics to surmount what Cairnes called an insuperable obstacle in the way of erecting economics into an exact science.
In his lectures on the Character and Logical Method of Political Economy,1 Cairnes made a comparison between economics and the natural sciences. In astronomy or chemistry, for instance, “the discovery of a law of nature is never considered complete till…an exact numerical expression [of the law]