A STATISTICAL METHOD FOR MEASURING ‘MARGINAL UTILITY’ 179 
obligations, and time even to learn of what has happened. More- 
over, moving from house to house costs money and trouble, which 
deter tenants from making complete adjustments. Finally, hous- 
ing accommodation is not as finely graduated or subdivisible 
as are food rations and other branches of the family budget. Food 
as a whole—and even individual foods, such as bread, milk, meat, 
etc.,—are almost infinitely subdivisible so that the adjustment 
can be made to the limit of the power of man to discriminate. 
But a family which is house hunting sometimes has to choose 
between a tenement which has too many or too big rooms and 
one which is too small, since the ideal intermediate size is not 
available. For all these reasons rent adjustments are less perfect 
than other consumption adjustments. Nevertheless, even as to 
rent, when two countries are compared, it seems fair to assume 
that, for the average or typical family, and “in the long run,” the 
adjustments are made with considerable precision. 
Comparability of Wants of Different People 
There is one other assumption, or group of assumptions, still 
to be mentioned, the assumption of comparability of wants among 
different people; for, in practice, we have no such convenient 
family as one which remains invariable in its wants and lends 
itself to study under successive episodes. But we do have, avail- 
able, thousands of workingmen’s budgets in the United States. 
England, ete. 
The simplest case of measuring one want against another is 
where we have only one particular individual, say a housewife, at 
one particular time, say January 1, 1900, under one particular 
set of circumstances, in the act, say, of buying eggs. At that 
moment when, after balancing her want for eggs against her 
want fo. dollars, she decides how many eggs she will buy, we may 
say definitely that one want is being measured directly against 
another in the same mind. But can we properly compare her 
particular want for eggs or dollars with that of another woman 
by her side who is going through the same process? Can we even 
compare her own individual wants at two different times? 
Finally, are we justified in taking her market decisions as repre- 
sentative of the wants of other members of her family? 
To all these questions I would answer “yes”—approximately 
at least. But the only, or only important, reason I can give for