OH. 7

THE DIFFUSION OF PRICE-LEVELS

91

have shown a considerable measure of agreementwith one another in spite of their different composi-tions. Furthermore, most of the traditional index-numbers have not only been of the wholesale typebut have also, by chance, approximated pretty closelyin their composition to that of International Standards.This has supplied another batch of inductive “ verifica-tions ”, tending to the conclusion that all index-numbers “ come to the same thing in the end ”,since index-numbers of international type have, asthey ought, tended to agree with one another notonly inside a given country but as between differentcountries also. Apart, however, from the fact thatthe measure of agreement has not really been anythingremarkable considering how many of the commoditiesare the same in all of them, it is obviously illogical toconclude that, because different indexes of the Whole-sale Standard or of International Standards agree*with one another, therefore they are good indicators ofthe Consumption Standard. On the contrary, thetables given above (pp. 60, 62) supply strong pre-sumptive evidence that over long periods as well asover short periods the movements of the Wholesaleand of the Consumption Standards respectively arecapable of being widely divergent.

The prevalence of these ideas has been particularlyunfortunate in its effect on the study of short-period,fluctuations ; for the failure of different price-levels to!move in the same way is the essence of such fluctua-'tions, so that to take a wholesale index as measuringall of them alike has been to assume away the veryproblem under investigation.

In the notion that a change in any one price-leveltends to be diffused, there is, of course, an importantelement of truth, especially when the initial disturb-ance has been of a monetary character. There is apressure on other price-levels in the same direction.Given sufficient time, and assuming that no non-