Abstract
Recently, in general, it has been evident that a rigid interpretation of Engel's Law is not true for a particular family, for particular times, and under particular conditions. So I attempted to investigate positively the application of Engel's Law on food expenditure and the pure Engel coefficient. Thus, I analyzed the family expenditure data from "Annual Report of the Family Income and Expenditure Survey" and "National Survey of Family Income and Expenditure for all Japan", into the Engel curve by 16 income groups. This investigation was undertaken upon using both (a) Engel coefficient and (b) Engel coefficient. They are as follows : (a)=Food expenditure/Living expenditure (b)=Food expenditure/Expenditure Secondly, I analyzed these data into the systematic material that living expenditure of which the total size of per head is nearly equal have revealed classifying the household size. The results obtained are summarized as follows : 1. The Engel's Law operate in the reverse direction over the definite range of income group and the Engel curve representing each income group in using both (a) and (b) Engel coefficient above mentioned. The Engel curve such as (a) and (b); (a') and (b') [Fig.5] show almost similiarity in style. 2. The pure Engel coefficient curve that eliminate household size effect shows contradiction to average Engel's Law with increase in household size.
