108 TWELFTH REPORT. 



worthy when they show to the world at large w-hat coimiioii observiatiou 

 shows to those personally familiar with the conditions described."^ 



It is very jirojierly my task then, thongh at the risk of some oircumlo- 

 ciition, to confirm common observation by showing- statistically the 

 changes which the prices of household goods are undergoing, and this un- 

 dertaking is all the more readily assumed through the fact that commod- 

 ity prices have been very carefully arranged to form the well known price 

 tables or index numbers. The elaboration of these tables has been the 

 work of years under such names as those of Mulhall, Palgrave, Mitchell, 

 Bradstreet, Dunn, ^^auerbeck and Sootbier and a resi>ectable literature 

 has been developed concerning the choice of averages, the commodities 

 to be chosen, the use of weights and the selection of base years. 



The avowed pur|jose of these price tables or index numbers is to 

 reflect the movement of jtrices — to show the averages of the marginal 

 comparisons between money and commodities which have been made 

 within a given time, and. for this purpose, many considerations are in- 

 volved which only remotely affect the cost of living. It is true indeed 

 that in a broad sense all the products of man's labor are intended for 

 consumption and, in an immediate or remote way, enter into the ex- 

 penses of living, but, as we have seen, the term "cost of living'' to have 

 a useful meaning must refer to the cost of real things to real house- 

 holds and the cost of living to a people would be the aggi'egated ex- 

 penses of these budgets. Under these circumstances production goods 

 of all sorts and lands and mines, forests and fisheries since they never 

 are found in family budgets should be excluded in a technical sense 

 from consideration. The problem of selection which separates the com- 

 modities which enter into cost of production from those which do not 

 is also an, important problem in the construction of price tables or index 

 numbers. It is obvious at once that in arranging a price table, which 

 shall show the average of prices among a people during a given period, 

 that observation of all the ratios of exchange which have been made 

 is humanly impossible. Xo single commodity can have all its measure- 

 ments in terms of money recorded nor can all the commodities even in a 

 single place. Under these circumstances certain typical commodities 

 must be chosen and their most typical prices made use of in arranging 

 the average. 



It is this matter of the commodities which have been chosen and the 

 prices which have been utilized which makes some of these tables more 

 acceptable in showing the high prices of household goods than are 

 others. This difference may be easily illustrated. Sootbier for example 

 found the statistics available in the Hamburg customs house and board 

 of trade records for ascertaining the prices of nearly one hundred and 

 forty-four articles over a period of thirty-eight years from which to com- 

 pute his index numbers. In 1885 Hamburg became a free port and his 

 records ceased. It is apjiarent that availability alone determined largely 

 in this instance the commodities which were chosen. 



In contrast with this rather superficial basis for price collection, the 

 United States Senate table of relative prices for the 1840-1891 period 

 grouped and weighted commodities according to the proportions in which 

 they entered into household expenses and greatly enlarged the number 

 of commodities upon which observations were made. 



^Distribution of Wealth, Chas. Spahr, preface. 



