Page:Stabilizing the dollar, Fisher, 1920.djvu/72

18 Statistics, only one group (house-furnishings) fell in price between 1896 and 1913, the year before the war. Assume that the nine groups, like the nine coins, are independent of one another,—for instance that "clothes and clothing," when they rise, do not prevent "drugs and chemicals" from falling; assume further that, for any one group among the nine, the chances of rise or fall are even; then the chances that eight out of the nine would rise coincidently would (as in the case of the coins) be exactly 10 in 512.

In actual fact the chances are less; for the assumption that a rise is as likely as a fall is not true of any ordinary commodity. A fall is really what we would, in most cases, expect because of improvements in methods of production. Taking this fact into consideration the chances that eight groups would rise coincidently are therefore less than 10 in 512—doubtless less than 1 in 100.

Of the 243 commodities recorded under the nine groups only 27 fell in price. It is true, of course, that not all of the 243 commodities are independent. Many commodities like bread and flour, or pig iron and iron products, move necessarily in sympathy with one another; but, even so, we may, I believe, safely put the chance of such an accidental rise simultaneously in 216 commodities out of 243 at less than one in a thousand.

This all corresponds with common sense. We seldom have world-feasts or world-famines. If the corn crop is short in some places, it is usually abundant in other places. If it is short in all places, the crop of wheat or barley or some other staple food is practically certain to be at least normal. If there is