12 BULLETIN 1382, U. S. DEPARTMENT OF AGRICULTURF 
Houses occupied increase in size from 2.7 rooms per family to 
7.8 rooms per family with the rise in average total expenditures. 
This represents an increase in rooms from 0.9 rooms per person to 
1.5 rooms per person, when the average number of rooms per house 
is divided by the average size of familj 7 or of household (Table 5). 
The average of the inventory values of furniture and furnishings in 
the home for the year of the study is closely related to the average 
amount of total expenditures. 
The percentage of all family living provided by direct purchase 
(Table 5) shows a marked and a fairly regular increase from 41.7 
per cent to 70.8 per cent. Roughly, only two-fifths of the family 
living is provided by direct purchase with families using less than 
$300 worth of goods as compared with about seven-tenths purchased 
with families using $3,000 or more than $3,000 worth of goods per 
year. 
To some degree at least, some of the relationships evident in 
Table 5 may be attributable to increase in size of family or house- 
hold, and to varying ages of different persons composing the house- 
holds. Although no correlation between the average total expendi- 
tures and the average ages of individuals or between the distribution 
of the average expenditures and ages of individuals is shown, such 
correlation might become evident if the effect of the size of family 
or household were eliminated. At any rate, it seems probable that 
"living cycles" — that is, periods when there are no children, when 
children are growing and developing, and when children are self- 
supporting — may bear a definite relation to expenditures for all 
purposes or to the distribution of expenditures for various purposes. 
COST-CONSUMPTION UNIT AND HOUSEHOLD-SIZE INDEX 
The farm family, owing to extremely variable composition, is not 
regarded as a satisfactory basis for determining the relation of the 
ability to pay to the standard of living, as stated previously. The 
famity fails to take account of the fact that the number, sex, and age 
of individuals composing the family or household make a difference 
in needs for food, rent, clothing, and other articles. Again, the 
term " family " makes no allowance for certain initial costs or fixed 
costs which must be borne by all families or households, regardless 
of the number and ages of the individuals which compose them. 
These initial costs are not the same for the different groups of eco- 
nomic goods. Further, the supplementary costs required to meet the 
demands of the third or fourth or other additional member of the 
family vary with the kind of goods used. Especially is this true 
with clothing and food, if the third or fourth or other additional 
member of the family be a son or daughter in the late teens. On an 
average, the supplementary or added costs for clothing for the son 
or daughter of this age are one and one-half times those for either 
parent, while the added costs for food are apparently about 80 per 
cent as much. In the same way costs meeting the demands made by 
the third or other additional member of the family on the other 
goods, as rent, maintenance of health, and education facilities, vary 
with little or no regard to the sum total of all the goods which this 
member of the family uses. 
The task of finding a common unit of comparison has been avoided 
in some cost-of-living studies by selecting the " standard " family, 
