1 
COST OF LIVING IN THE UNITED STATES 
did not measure changes in the cost of living for the ultimate 
consumer;' and that, since food required only about two- 
fifths of the average family’s annual outlay, changes in retail 
food prices alone did not adequately measure changes in the 
total cost of living? What was needed was an estimate of 
the change in the total cost of living, related to local condi- 
tions and to a definite standard of living. 
It had generally been assumed up to this time that the 
creation of an index of the total cost of living, based on 
changes in retail prices of the major component items, 
weighted and combined according to consumption, was a 
task fraught with almost insuperable difficulties because 
of the operations involved. The establishment of a basic 
budget which would properly represent complete require- 
ments and the maintenance of comparable standards from 
period to period? required a choice of typical articles and a 
knowledge of their use by average families, which could be 
obtained only through considerable effort in careful investi- 
gation. The selection of retailers who would furnish quota- 
tions, and the enlistment of their co-operation in numbers? 
sufficient to afford a broad enough basis for generalizing, 
implied familiarity with a wide variety of marketing condi- 
tions. The collection of the data and the clerical work 
{A comparison of the trend of wholesale prices and of retail prices is given in 
Table A (pp. 162ff.) and is illustrated in Chart 2. This shows clearly that wholesale 
prices often rise sooner and to a higher level than retail prices and that the reverse 
movement is also more rapid and more abrupt. 
? Comparisons of the figures showing changes in food prices and in the total cost 
of living all indicate that food prices rose much more rapidly and to a higher point 
and fell much more rapidly and to a lower point than did the total cost of living, 
See, for example, Table 1 (p. 30), Table 4 (pp. 66-67), Tables C-1 and C-2 (p. 182). 
* A cost of living index series has frequently been started through collecting retail 
prices of food, since these articles are for the most part easily standardized and com- 
parability of standard is readily maintained. Coal, matches and kerosene may be 
added; and then house rents. When clothing and sundries are reached, however, 
difficulties arise, This is because of the well-nigh limitless choice of samples and 
the fact that choice necessarily rests on individual preference or judgment; and 
because of complications due to changes in styles and changes of season and the 
problems of standardization. The Canadian number has grown in that way, as has 
the index number computed by the federal statistical office in Germany, which 
did not include sundries until February, 1925. 
* An index of retail prices is necessarily based on an average of several quotations, 
whereas for an index based on wholesale prices only representative market quota- 
ONS are necessary.